June 2013

Financial literacy and inclusion

RESULTS OF OECD/INFE SURVEY ACROSS COUNTRIES AND BY GENDER

FINANCIAL LITERACY AND INCLUSION: RESULTS OF OECD/INFE SURVEY ACROSS COUNTRIES AND BY GENDER

FOREWORD

Governments around the world are recognising the benefits to individuals and national economies of having a financially literate population that has access to appropriate financial products with relevant consumer protection in place. In recent years, the G20 has endorsed three sets of principles in this regard on financial consumer protection, financial inclusion and national strategies for financial education, indicating firm commitment towards full and safe financial integration1. Measurement and analysis play important roles in designing and implementing such policies. A measure of financial literacy can be used to indicate the level of need for financial education across the population. More detailed analysis can be used to identify the aspects of financial literacy most in need of work, and the groups of population that need targeted support. Demand side measures of financial inclusion provide similar insights, indicating where the policy focus should be placed. The process of measurement can be complex and expensive. Developing a questionnaire, designing a survey method, collecting data and analysing that data to provide meaningful results is time consuming and costly, and requires expert inputs. It is for this reason that the OECD International Network on Financial Education (INFE), with the support of the Russian/World Bank/OECD Trust Fund on Financial Literacy, agreed to develop such a questionnaire and method for use across countries thus cutting costs to individual countries, reducing the time needed to undertake a survey and providing the opportunity of comparing data across countries2. The resulting OECD/INFE Financial Literacy Core Questionnaire has proved a huge success. This publication shows how it has been successfully employed in 14 countries3. The countries participating in this first ever cross-national study of financial literacy and financial inclusion (comprising indicators of product awareness, use and choice, as well as financial knowledge, attitudes, behaviour) must be commended for their willingness to undertake the study. Many of them have already taken their first step towards improving financial literacy levels as a result of using the questionnaire to create a financial literacy survey; some have even repeated the measure in order to track progress and identify changes. The success of the questionnaire is also evidenced by the fact that countries as widespread as Korea, Iceland and Jamaica are now benefiting from the tried and tested questions to capture levels of financial literacy amongst their adult populations. As more countries use the questionnaire, the added value from cross-national comparisons will grow rapidly.
1

G20 Principles for Innovative Financial Inclusion (2010) developed by Global Partnership for Financial Inclusion; G20 High-level Principles on Financial Consumer Protection developed by the OECD Task Force on Financial Consumer Protection (2011); OECD/INFE High-level Principles on National Strategies for Financial Education developed by the OECD International Network on Financial Education (2012). The OECD/INFE has developed a detailed toolkit comprising a core questionnaire and methodological guidance as well as a set of supplementary questions: see OECD/INFE (2013a). Albania, Armenia, British Virgin Islands, Czech Republic, Estonia, Germany, Hungary, Ireland, Malaysia, Norway, Peru, Poland, South Africa and UK. 3

The resulting survey data are rich and provide numerous insights into levels of financial literacy in the population and the needs of different groups (Chapter 1). The same data are further analysed in Chapter 2 to explore levels of financial inclusion, and the potential association between exclusion and low levels of financial literacy. The final chapter (Chapter 3) reports analysis by gender using the OECD/INFE data and other sources. Annexes provide additional detail, including information about the lessons learned from countries applying the Core Questionnaire and discussion about the approach taken to creating measures of financial literacy. This publication, drafted by the OECD Secretariat, gives an indication of the research possibilities and policy relevance of the data collected but much more can be learned from in-depth analysis, whether focusing on certain target groups (this publication includes a chapter on the needs of women, but future work may look at young adults or those low levels of education) or particular issues of interest (such as indicators of credit use or financial strain).

TABLE OF CONTENTS

FOREWORD......................................................................................................................................................3 EXECUTIVE SUMMARY...................................................................................................................................11 Financial knowledge...................................................................................................................................12 Financial behaviour ....................................................................................................................................13 Attitudes ....................................................................................................................................................13 Overall indicators of financial literacy .......................................................................................................14 Variations in financial literacy by socio-demographics..............................................................................15 Indicators of inclusion and exclusion .........................................................................................................15 Further analysis of variations in financial literacy levels by gender ..........................................................18 Conclusion ..................................................................................................................................................19 CHAPTER 1: THE FINANCIAL LITERACY OF ADULTS IN 14 COUNTRIES ................................................... 21 INTRODUCTION .............................................................................................................................................23 Participating countries and data collection ...............................................................................................24 The structure and content of this chapter.................................................................................................24 FINANCIAL KNOWLEDGE ...............................................................................................................................26 Survey questions designed to test knowledge ..........................................................................................26 A financial knowledge score ......................................................................................................................29 FINANCIAL BEHAVIOUR .................................................................................................................................34 Financial behaviour across different domains ...........................................................................................34 A score for financial behaviours ................................................................................................................40 FINANCIAL ATTITUDES ...................................................................................................................................45 Attitudes and preferences: short term or longer term? ............................................................................45 Combining the various attitudes................................................................................................................46 RELATIONSHIP BETWEEN BEHAVIOUR AND THE OTHER SCORES .................................................................48 COMBINED MEASURES OF FINANCIAL LITERACY ..........................................................................................51 Segmenting the population .......................................................................................................................51 Developing an overall measure of financial literacy ..................................................................................52 VARIATIONS BY SOCIO-DEMOGRAPHICS.......................................................................................................54 Gender .......................................................................................................................................................54 Age .............................................................................................................................................................58 Income .......................................................................................................................................................60 Education level ...........................................................................................................................................61 Attitude to risk ...........................................................................................................................................63 Multivariate analysis ..................................................................................................................................63 CONCLUSION .................................................................................................................................................67

CHAPTER 2: FINANCIAL INCLUSION AND FINANCIAL LITERACY. ANALYSIS OF DATA FROM THE OECD FINANCIAL LITERACY MEASUREMENT EXERCISE IN 14 COUNTRIES ...................................................... 69 INTRODUCTION .............................................................................................................................................71 Financial Education, Financial Inclusion and the Need for Data................................................................71 PRODUCT AWARENESS..................................................................................................................................73 The relationship between awareness and financial knowledge................................................................75 PRODUCT HOLDING .......................................................................................................................................77 The relationship between awareness and holding ....................................................................................77 Indicators of product holding ....................................................................................................................77 Payment products ......................................................................................................................................78 Savings and investment products ..............................................................................................................79 Insurance....................................................................................................................................................79 Credit..........................................................................................................................................................79 Variations in financial literacy by product holding ....................................................................................79 ACTIVE PRODUCT CHOICE .............................................................................................................................84 RELYING ON FRIENDS AND FAMILY ...............................................................................................................86 FINANCIAL INCLUSION AND SOCIO-DEMOGRAPHICS ...................................................................................88 Variations in awareness by socio-demographics .......................................................................................88 Variations in payment product holding by socio-demographics ...............................................................89 Variations in savings product holding by socio-demographics..................................................................91 Variations in credit and insurance product holding by socio-demographics ............................................93 Variations in active product choice by socio-demographics .....................................................................94 CONCLUSION .................................................................................................................................................96 Product awareness ....................................................................................................................................96 Products held .............................................................................................................................................97 Product choice ...........................................................................................................................................99 Reliance on friends and family...................................................................................................................99 CHAPTER 3: GENDER DIFFERENCES IN FINANCIAL LITERACY ...............................................................101 INTRODUCTION ...........................................................................................................................................103 The policy interest in the financial literacy of women and girls ..............................................................103 FINANCIAL KNOWLEDGE OF WOMEN .........................................................................................................104 Gender differences in financial knowledge at young ages ......................................................................105 Less well-educated and low-income women have the lowest financial knowledge ...............................106 Gender differences are smaller but still significant after controlling for socio-demographic factors ....107 WOMENS FINANCIAL ATTITUDES ...............................................................................................................109 Women appear to be aware of their lack of financial knowledge ..........................................................109 Women have lower confidence than men in their financial knowledge and skills .................................109 Gender differences in interest for financial matters ...............................................................................110 Women are more risk-averse than men ..................................................................................................111 GENDER DIFFERENCES IN FINANCIAL BEHAVIOUR AND STRATEGIES .........................................................112 Women are more likely to have a budget ...............................................................................................112 Making ends meet: women tend to cut down on spending while men try to earn extra money...........115 Gender differences in product holding ....................................................................................................117
6

Gender differences in saving behaviour ..................................................................................................119 Women are less likely to choose financial products appropriately .........................................................123 FACTORS AFFECTING GENDER DIFFERENCES IN FINANCIAL LITERACY .......................................................127 Gender roles in household financial decision making have limited impact ............................................127 Womens lower financial inclusion and access to finance.......................................................................127 CONCLUSION ...............................................................................................................................................129 REFERENCES ................................................................................................................................................131 ANNEX 1: ANNEX 2: ANNEX 3: ANNEX 4: Tables Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. Table 12. Table 13. Table 14. Table 15. Table 16. Table 17. Table 18. Table 19. Table 20. Table 21. Table 22. Table 23. Table 24. Table 25. Table 26. Table 27. Table 28. Table 29. Table 30. Table 31. Table 32. The 8 knowledge questions ...................................................................................................27 Correct responses to knowledge questions ..........................................................................28 Creating a knowledge score ..................................................................................................30 Positive financial behaviours by country ...............................................................................35 Actively saving or buying investments in the past 12 months ..............................................38 Creating a behaviour score ....................................................................................................41 High score on each of the financial literacy components .....................................................51 Regression results ..................................................................................................................65 Awareness by socio-demographics .......................................................................................89 Payment products by gender, age and education level ........................................................90 Savings holding by gender, age and income .........................................................................92 Active product choice by age, income and education level ..................................................95 Creating a knowledge score .................................................................................................140 Ranking of country according to scoring method used .......................................................142 Financial Behaviour Variables..............................................................................................146 Testing the behaviour score ................................................................................................148 Details of sample .................................................................................................................153 Financial knowledge: division, time-value of money, interest paid on a loan ....................155 Financial knowledge: interest plus principle, compound interest ......................................156 Financial knowledge: risk and return, inflation, diversification ..........................................157 Behaviour: Before I buy something I carefully consider whether I can afford it ................158 Behaviour: I pay my bills on time ........................................................................................158 Behaviour: I keep a close personal watch on my financial affairs .......................................159 Behaviour: I set long term financial goals and strive to achieve them ...............................159 Average behaviour score by country ...................................................................................160 Attitude: I find it more satisfying to spend money than to save it for the long term .........160 Attitude: I tend to live for today and let tomorrow take care of itself ...............................161 Attitude: Money is there to be spent ..................................................................................161 Average combined attitude scores ......................................................................................162 Financial literacy segments by gender ................................................................................163 Financial Literacy Segments by Income ...............................................................................164 Financial knowledge across subgroups of women ..............................................................175 TESTING THE CORE QUESTIONNAIRE AND DEVELOPING A SCORE ............................................135 FURTHER DETAIL OF COUNTRY LEVEL DATA ..............................................................................153 PRODUCT HOLDING BY COUNTRY..............................................................................................167 ANALYSIS BY SUBGROUPS OF WOMEN......................................................................................175

Figures Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27. Figure 28. Figure 29. Figure 30. Figure 31. Figure 32. Figure 33. Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Figure 39. Figure 40. Figure 41. Figure 42. Figure 43. Figure 45. Figure 46. Figure 47. Country groupings by average financial knowledge scores ..................................................31 Distribution of knowledge scores ..........................................................................................32 Financial knowledge: Percentage with a high score .............................................................33 Responsible and has a household budget .............................................................................37 Shopping around for financial products ................................................................................39 Borrowing to make ends meet ..............................................................................................40 Country groupings by average financial behaviour scores ....................................................42 Distribution of financial behaviour scores.............................................................................43 Financial behaviours: Percentage scoring 6 or more ............................................................44 Distribution of financial attitude scores ................................................................................46 Percentage of respondents with average score over 3 .........................................................47 Relationship between financial knowledge and behaviour ..................................................49 Financial behaviour and attitudes .........................................................................................50 Financial Literacy Segments ..................................................................................................52 Country groupings by average overall financial literacy scores ............................................53 High knowledge score by gender ..........................................................................................55 High behaviour score by gender ............................................................................................56 High attitude score by gender ...............................................................................................57 Mean overall score by gender ...............................................................................................58 Financial literacy segments by age ........................................................................................59 Average overall score by income...........................................................................................61 Financial literacy segments by education..............................................................................62 Average overall scores by risk aversion .................................................................................63 Proportion of respondents aware of at least 5 products ......................................................74 Average financial literacy by product awareness ..................................................................75 Product awareness by financial knowledge ..........................................................................76 Product holding .....................................................................................................................78 Average financial literacy by payment product.....................................................................80 Paying bills on time with and without a payment account ...................................................81 Average financial literacy by saving and investment ............................................................81 Understands the impact of compound interest and has savings or investments .................82 Average financial literacy by credit product..........................................................................83 Average financial literacy by insurance product ...................................................................83 Made an active product choice in the last 2 years ................................................................84 Average financial literacy by recent product choice .............................................................85 Relying on friends or family for borrowing and saving .........................................................86 Average financial literacy by reliance on family and friends .................................................87 Payment product holding by income ....................................................................................91 Holds savings or investment account by education ..............................................................93 Credit products by gender .....................................................................................................93 Product choice by gender ......................................................................................................94 Average financial knowledge score by gender ....................................................................105 Average financial knowledge score by gender (young people)...........................................106 Percentage do not know replies by gender.......................................................................110 Responsibility for day-to-day money management decisions in the household ................113 Responsible for money management decision and has a household budget.....................114
8

Figure 48. Figure 49. Figure 50. Figure 51. Figure 52. Figure 53. Figure 54. Figure 55. Figure 56. Figure 57. Figure 58. Figure 59. Figure 60. Figure 62. Figure 63. Figure 64. Figure 65. Figure 66. Figure 67. Figure 68. Figure 69. Figure 70. Figure 71. Figure 72. Figure 73. Figure 74. Figure 75.

I keep a close personal watch on my financial affairs .........................................................114 Sometimes people find that their income does not quite cover their living costs .............115 Strategies for making ends meet: cutting back on spending ..............................................116 Strategies for making ends meet: earning extra money .....................................................117 Percentage of respondents who hold a payment product .................................................118 Percentage of respondents who hold an investment product............................................119 Percentage of respondents who hold a saving product ......................................................120 Covering living expenses .....................................................................................................121 Saving behaviour: Actively saving through formal products ...............................................122 Saving behaviour: Saving informally....................................................................................122 Has tried to compare financial products across providers ..................................................124 Information sources: made some attempt at taking an informed decision .......................125 Information sources: consulted independent professional info/advice .............................126 Products: Albania.................................................................................................................167 Products: Armenia ...............................................................................................................167 Products: Czech Republic ....................................................................................................168 Products: Estonia .................................................................................................................168 Products: Germany ..............................................................................................................169 Products: Hungary ...............................................................................................................169 Products: Ireland .................................................................................................................170 Products: Malaysia ..............................................................................................................171 Products: Norway ................................................................................................................171 Products: Peru .....................................................................................................................172 Products: Poland..................................................................................................................173 Products: South Africa .........................................................................................................173 Products: UK ........................................................................................................................174 Products: British Virgin Islands ............................................................................................174

EXECUTIVE SUMMARY

Many governments and international organisations recognise the importance of measuring levels of financial literacy4 and collecting information about financial inclusion5 to inform policy responses. In 2010, the OECD International Network on Financial Education (INFE) developed and fielded a questionnaire designed to create an international, broad based and robust measure of financial literacy as well as financial inclusion indicators. Reflecting the majority of national financial literacy measurements, this OECD/INFE Financial Literacy Core Questionnaire (or Core Questionnaire for brevity) focuses on those aspects of knowledge, attitudes and behaviours that are associated with the overall concept of financial literacy. The OECD/INFE measurement toolkit comprises this core questionnaire as well as methodological guidance and a set of supplementary questions (OECD/INFE, 2013a). The Core Questionnaire incorporates a series of questions on financial products chosen specifically to enable analysis of financial inclusion. These ask about product awareness, products currently held and approaches to choosing financial products. The questions are drawn from worldwide good practice, and asked of a representative sample of adults in each country, using either telephone or face-to-face interviews6. Almost all the questions refer to the individual answering the question, although some information is collected about the respondents household, including the household income and the number of people living with the respondent. The core questions have been picked because they are applicable to the vast majority of people and they are suitable across a wide range of countries. While all of the questions retain the same format in each country, some of the answer codes allow for country specific responses (such as methods of saving, or types of financial product). This enables cross country comparisons that are contextually meaningful, while maximising the potential to include all interested countries. In 2010, the OECD invited countries to use the Core Questionnaire to measure financial literacy in their country, and to submit the data to the OECD/INFE Secretariat as part of a project to test the applicability of a standard financial literacy measures across countries and across socio-demographic groups within countries. At this initial stage, 12 countries across 4 continents accepted the invitation:
4

Financial literacy is defined as: A combination of awareness, knowledge, skill, attitude and behaviour necessary to make sound financial decisions and ultimately achieve in dividual financial wellbeing. (Atkinson and Messy, 2012). The OECD/INFE has defined financial inclusion as follows: The process of promoting affordable, timely and adequate access to regulated financial products and services and broadening their use by all segments of society through the implementation of tailored, existing and innovative approaches including financial awareness and education, with a view to promote financial wellbeing as well as economic and social inclusion. The exception to this is Norway, where a subset of questions has been tested over the telephone, whilst the full questionnaire has been implemented through an online survey. 11

Armenia; the Czech Republic; Estonia; Germany; Hungary; Ireland; Malaysia; Norway; Peru; Poland; South Africa, and the United Kingdom. Albania joined the project at a later stage and the British Virgin Islands (BVI) subsequently undertook a national survey using the Core Questionnaire and shared the data with the OECD/INFE7. The survey population consisted of adults aged 18 and over8. Each country aimed to interview at least 1,000 individuals and data has been weighted to reflect the national population in terms of basic demographics. The data collection for this project began in the second half of 2010 and was completed in early 20119. Each participating country collected the results from their survey in an anonymised database which was then submitted to the OECD Secretariat for analysis. The results focus on the general pattern of financial literacy in different countries, identifying commonalities and differences. The exercise is not designed to rank countries according to their levels of financial literacy, although it is illustrative to draw certain comparisons across countries to highlight variations. Financial knowledge A financially literate person will have some basic knowledge of key financial concepts. The Core Questionnaire therefore includes 8 questions to test levels of knowledge in each10. The questions have been chosen to cover a range of financial topics and to vary in difficulty, although none of them is excessively complex and none of them requires expert knowledge. A financial knowledge score has been created by summing correct answers and reporting them as a score out of 100. The proportion of the population in each country that exhibited a relatively high level of financial knowledge (defined as getting at least 3/4 of the questions correct) has also been reported. In some countries, fewer than half of the respondents achieved this score, and no country had more than 70% of their population who could answer at least 3/4 of the questions. Of particular concern is the relatively large proportion of people who could not calculate simple interest earned on a savings account at the end of one year and then identify the impact of compounding over 5 years. In Albania and Peru, fewer than 1 in 5 people were able to apply their knowledge to this two-part question, and in every country except Norway, at least half of the population failed to identify the impact of compounding. There was also a worrying low level of awareness of the benefit of diversification, with at least a third and in some cases over half of respondents in each country being unable to answer this question.
7

In cross-country tables and figures, the data from BVI is reported last to remind the reader that the country was not formally part of the first measurement project, and has a smaller sample size. Where countries have surveyed a wider population (from 15 years of age in some cases) responses are only included from those aged 18 and over in some cases this meant recalibrating the weights. With the exception of data from Albania and BVI. Most countries have used the questions as designed, although there are some variations which mean that direct comparisons of overall scores between the other countries and Norway or Hungary should be made with caution. 12

9 10

Financial behaviour The way in which a person behaves will have a significant impact on their financial wellbeing. It is therefore important to capture evidence of behaviour within a financial literacy measure. The Core Questionnaire does this by asking a variety of questions in different styles, to find out about behaviours such as thinking before making a purchase, paying bills on time and budgeting, saving and borrowing to make ends meet. The financial behaviour score counts positive behaviours exhibited and rescales the final score to take values between 0 to 100; those exhibiting 2/3 of the behaviours are considered to have a relatively high score. As with knowledge, in some countries fewer than 50% achieved a high score. BVI residents showed the largest number of positive behaviours, with 71% exhibiting such a score. There is a wide variation in behaviours within countries, and noticeable variation across countries. However, of concern in all the countries surveyed is the lack of active, informed market participation: very few people reported that they had shopped around and sought independent information or advice to make a financial product choice in the last 2 years (UK participants were the most likely to have done so, at 16%). In some of the countries, the lack of active saving is also a concern, although here there are large variations by country. In Hungary just 27% had been saving in the previous 12 months whilst in Malaysia almost everyone had done so (97%). In all, only three countries found that more than 80% of their population were actively saving. The likelihood of setting long term goals also varies by country: more than 7 in 10 Peruvians reported that they did set long term goals, compared with just 3 in 10 Albanians. Whilst borrowing to make ends meet is not widespread, it is a problem for a large minority in certain countries. In particular, almost half (47%) of Armenians had resorted to borrowing the last time their income fell short of their expenditure; in Albania, Peru and South Africa over a quarter of respondents had also done so. There is considerable variation in behaviour within a country. For example, a large proportion of Malaysian respondents were active savers and carefully considered their purchases, yet hardly any (3%) had made a recent financial product choice after shopping around and seeking independent guidance. In Norway, almost 9 in 10 people reported that they were keeping an eye on their financial affairs yet just 1 in 4 was budgeting: showing that more people look over their recent financial activities than plan future ones. Attitudes Attitudes and preferences are considered to be an important element of financial literacy. If people have a rather negative attitude towards saving for their future, for example, it is argued that they will be less inclined to undertake such behaviour. Similarly, if they prefer to prioritise short-term wants then they are unlikely to provide themselves with emergency savings or to make longer term financial plans.

13

The Core Questionnaire includes three attitude statements to gauge respondents attitudes towards money and planning for the future11. The attitude questions ask people about whether they agree or disagree with particular statements, to capture their disposition or preferences. Very few respondents in Armenia (8%) and Poland (19%) got satisfaction from saving. In contrast, 64% of Peruvians and 61% of Albanians found saving satisfying. Albanians and Peruvians were also the most conservative with money, with almost half of respondents (45%) disagreeing that money is there to be spent. In contrast, just 2% of Armenians and 12% of Polish respondents tended to disagree with the statement. The average response to the three attitude statements provides an overall indicator of attitude. Respondents with an average attitude indicator above 3 (a high score), have attitudes that tend towards the longer term. Analysis shows that there is a very wide variation in attitudes across countries: in Armenia, just over 1 in 10 people have a positive attitude towards the longer term; compared with 71% in Peru. Overall indicators of financial literacy Several approaches have been used to describe the data in terms of overall levels of financial literacy: Summing the raw scores financial knowledge, financial behaviour and financial attitudes into an overall indicator of financial literacy and rescaling this from 0 to 100: The average of this combined score across all participating countries is 63.2. Scores in the Czech Republic, Germany, Hungary, Ireland, Norway, Malaysia, Peru, the UK and BVI are above this combined average (See Figure below). Looking at the proportion of the population achieving high scores on each of the three components of financial literacy: In 8 of the countries surveyed, a larger proportion of the population achieved a high knowledge score than a high behaviour score; indicating that levels of financial literacy in these countries are higher in terms of knowledge than behaviour. Conversely, in Germany, Malaysia, Norway, Peru, South Africa and BVI countrywide financial literacy levels are higher in terms of behaviour; in most cases this is because a larger proportion of the population exhibit 6 or more positive behaviours, rather than because knowledge is exceptionally low. Counting the number of high scores for financial knowledge, financial behaviour and financial attitudes that each respondent achieved: In all of the countries surveyed there are some people who did not achieve any high scores and others who achieved a high score on all 3 aspects of financial literacy. However, typically people tend to have 1 or 2 strengths. Germany and BVI stand out with over 30% of their populations achieving 3 high scores, indicating high levels of financial literacy.

11

Norway only used the first question. 14

Country groupings by average overall financial literacy scores

Lower average overall score Armenia South Africa Poland Estonia Albania Czech Republic Peru UK Norway Ireland Malaysia Hungary Germany BVI
Shaded boxes indicate homogenous groups of countries, computed in IBM SPSS19 using Scheffe Homogenous Subsets. Average scores range from 56.4 to 68.6.

Higher average overall score

Variations in financial literacy by socio-demographics It is clear that levels of financial literacy vary within countries and it is therefore useful to know more about how they vary across particular socio-economic groups. The findings of analysis by socio-demographics show important gender differences, which are further explored in Chapter 3. There is also a noticeable variation in financial literacy by age and income. In most countries, middle age is associated with higher levels of financial literacy, whilst the oldest and youngest respondents are more likely to have no high scores. Regression analysis confirms that higher income respondents are more likely to gain high scores than their lower income peers. Similarly, there is also a positive relationship between education and financial literacy. Higher educated individuals are more likely to exhibit positive behaviours and attitudes as well as show advanced levels of knowledge. Indicators of inclusion and exclusion Financial inclusion relates specifically to awareness, access and use of financial products. The second chapter therefore pays particular attention to the awareness (which is necessary in order to use a product), use, and recent purchase of a range of financial products. The Core Questionnaire includes a suggested list of product types covering savings, investments, pensions, current accounts, insurance and credit. This list has been edited by each of the participating countries to be contextually relevant. For each country, a unique list of product types has been created by national authorities or experts. Results from each country in this chapter should then be considered in context- it may not be appropriate to compare across very different countries.
15

The data from the Financial Literacy Measurement survey has provided the OECD/INFE with a unique opportunity to analysis the demand side of financial inclusion, and the possibility to explore the extent to which this is associated with financial literacy. Charts detailing the products listed in each questionnaire and showing percentages of respondents reporting product awareness, current product holding and recent purchases are shown in Annex 2. The indicators of financial inclusion focus on the number of products that respondents were aware of, the products they held, and recent product choice. An indicator of potential exclusion is also reported, looking at the extent to which respondents relied on friends and family for saving or borrowing. The findings show a positive association between financial literacy and financial inclusion. In all countries, people who were aware of at least 5 financial products had higher levels of financial literacy than those who were less aware. In some countries, low levels of financial knowledge were associated with low levels of financial product awareness, suggesting that financial education policies could usefully provide general financial knowledge whilst also improving awareness of financial products. Awareness of a range of products is a vital first step in removing demand side barriers to financial inclusion. At a minimum consumers need to know: That certain financial products exist; the purpose of each category of product: e.g. to receive income, make payments, protect savings from the impact of inflation or protect households from the consequences of expenditure shocks; which products are regulated or safe; and where and how to access the relevant products (who are the suppliers, what are the access requirements, etc).

Awareness may also include knowing which features should be taken into account when choosing amongst various offers, and being aware of consumer rights and responsibilities when holding a particular product. The data reveals that individuals with a payment account were more likely to make timely bill payments, an important finding that could be stressed in financial education programmes. Analysis also shows that understanding of the benefit of receiving compound interest on savings was higher amongst those with a savings or investment product than those who did not hold such a product. However, many savers did not know about this benefit, and may have made inappropriate choices as a result. Furthermore, financially literate people were more likely to hold insurance, a finding that strongly supports the argument that financial literacy leads to financial inclusion (since there is no reason to assume that holding insurance would make someone more financially literate). Financial literacy levels are also lower amongst people who turned to friends and family to borrow money or save, suggesting that financial education could help them to identify more secure alternatives.

16

The analysis discussed in this chapter looks at the first of these issues: awareness of the existence of the product. In particular, it shows that financial literacy was lower amongst respondents who were aware of fewer than 5 products in all participating countries. Financial inclusion depends on both the demand and supply of appropriate products. It is not possible to directly capture demand for a product, but it is possible to explore product holding, which indicates demand that has been met. There are many financial products available, serving different needs and providing consumers with a range of tools and services. This analysis focuses on 4 key categories: payment products; saving and investment; insurance; and credit. Whilst some consumers may not require products from each of these categories, they are nevertheless frequently used in combination, and provide a broad overview of product use at a country level. If awareness is considered to be an approximation for supply, the analysis shows that in some cases supply exists, but demand is low i.e. people are aware that certain products are available, but they are not using them. This finding is particularly striking in Armenia and Peru; and may invite further investigation. If consumers have made a recent product choice, then they are engaged in the financial market. This is a good indicator of the health of the market and the extent to which products appeal to the population at large. Whether this reflects a genuine match between the needs of consumers and the design of products, or whether the activity is triggered by advertising should be explored at the national level. The analysis presented here shows that financial literacy was higher amongst those who made a recent product choice than those who had not. However it should be remembered that 2 of the financial behaviours captured in our measure of financial literacy relate to how a product was chosen, and so some variation would be anticipated. Nevertheless the additional points available cannot explain the size of the gap in countries such as Albania, Germany and Peru. The financial literacy analysis in Chapter 1 shows that it is very unusual for consumers to make an informed product choice based on shopping around and seeking independent information or guidance. The way in which individuals choose a product may be influenced by a wide range of factors including the speed with which they (believe they) need to make a decision, the availability of alternatives, the extent to which information and advice is available and their previous experiences. However, it is clear that financial education could usefully highlight the benefit of comparing across products, seeking impartial opinions and regularly assessing the suitability of products held. The final indicator counts people who are not using financial products to meet (some, or all of) their existing needs. By identifying people who are either saving with friends and family or have resorted to borrowing from them, it is possible to identify people who could benefit from improved financial inclusion measures that ensure suitable products exist, help consumers to see the benefit of using formal products and encourage them to access such products. However, in some countries, and some groups, family and friends may be the first option for people to access financial support, and this behaviour would not be seen as a sign of financial exclusion. The analysis shows that in every country there was some reliance on friends and family. In some countries over a third of respondents had turned to such a network either to borrow money or to put money aside in the last 12 months. Furthermore, financial literacy was lower amongst those individuals.
17

Further analysis of variations in financial literacy levels by gender In many countries, women display lower financial knowledge than men and are also less confident in their financial knowledge and skills. Even though women appear to be better than men at short term money management behaviour they have a number of vulnerabilities in other aspects of financial behaviour. For instance, women are more likely to experience difficulties in making ends meet, in saving and in accumulating financial resources. Moreover, women are more risk averse and less likely than men to invest in risky assets, such as investments. Further, women show more difficulties than men in choosing financial products appropriately. In particular, men are more likely than women to shop around for financial products. A number of factors appear to be related to gender differences in financial literacy. Gender differences in financial literacy are strongly correlated with differences in socio-economic conditions of men and women, suggesting that limited access to education, employment and formal financial markets not only reduce womens financial well-being per se, but also limit the extent to which women can improve their knowledge, confidence and skills about economic and financial issues. Overall, the analysis of the available evidence on gender differences in financial literacy based on various sources, including the OECD/INFE financial literacy survey and other studies highlights a number of issues: Women display lower financial knowledge than men in most of the countries surveyed. In particular, young women, widows, less well-educated and low-income women lack financial knowledge the most. Gender differences in financial knowledge are in part, but not entirely, related to demographic and socio-economic factors. Available evidence suggests that women are less confident then men in their financial knowledge and skills, less over-confident in financial matters, and more averse to financial risk. They also appear to be less interested than men in financial matters. Women appear to be better than men at short-term money management behaviour: they are more likely than men to have a budget and to keep a close watch on their financial affairs. Women are more likely to experience difficulties in making ends meet. In addition, women and men have different coping strategies for making ends meet. If confronted with situations where their income is not sufficient to cover living costs, women tend to cut expenses, while men prefer finding ways for earning extra money. Women and men display different saving behaviour. Consistently with the evidence that they face more difficulties in making ends meet, women tend to save less and accumulate less wealth than men, typically as a result of their weaker labour market position. Moreover, men are more likely than women to be actively saving through formal financial products, while women are more likely than men to be saving cash at home or in their wallet, or to be saving in informal savings clubs. Women are also less likely than men to invest in risky assets. Women show more difficulties than men in choosing financial products appropriately. In particular, men are more likely than women to shop around for financial products. In some countries, men are also more likely than women to take informed financial decisions and to use independent advisors.
18

Conclusion This book presents the findings of analysis of a set of questions asked to very different populations around the world and provides simple, meaningful indicators of financial literacy and financial inclusion. The cross-country nature of this analysis enables countries to create partnerships to tackle particular issues at an international level, and also allows some countries to identify potential benchmark countries that have achieved a higher level of financial literacy across their population. The research highlights areas for concern. For instance, in every country there is significant room for improvement in terms of financial knowledge: understanding of some everyday financial concepts such as compound interest and diversification is lacking amongst sizeable proportions of the population in every country, and in most countries surveyed women are less knowledgeable than their male counterparts. The findings also highlight a large proportion of individuals who could benefit from initiatives designed to change their behaviour. In almost every country surveyed, at least 3 in 10 respondents exhibited fewer than 2/3 of the positive behaviours discussed. The analysis shows how financial knowledge and financial behaviour are associated in every country more knowledgeable individuals are more likely to exhibit positive financial behaviour. The data has also created a unique opportunity to explore some of the demand side factors of financial inclusion, including product awareness and holding, providing an important complement to existing international financial inclusion measurements, which focus mostly on supply-side data from financial institutions. The analysis undertaken in Chapter 2 provides indications that the population of some countries have low levels of financial inclusion on a number of indicators. The positive association between financial inclusion and financial literacy suggests an important role for financial education in the drive to increase levels of inclusion in some countries. Moreover, the various indicators of financial inclusion are significantly associated with various socio-demographic factors such as gender, age, education, and household income; suggesting that targeted financial education may reap the highest rewards. Chapter 3 identifies variations in the levels of financial literacy by gender, and indicates a pressing need for financial education targeted at women to address the gender gap. This is particularly important in terms of increasing levels of financial knowledge amongst women and girls, and improving ability to make ends meet and choose appropriate financial products. A number of factors appear to be related to gender differences in financial literacy. Gender differences in financial literacy are strongly correlated with differences in socio-economic conditions of men and women, suggesting that limited access to education, employment and formal financial markets not only reduce womens financial well-being per se, but also limit the extent to which women can improve their knowledge, confidence and skills about economic and financial issues. The data hold a great deal of potential. The OECD will continue with analysis in order to inform the work of its INFE, focusing particularly on variations in financial literacy by key socio-demographic groups, levels of financial inclusion and financial access, as well as exploring in more detail the relationship between various aspects of financial literacy.

19

Chapter 1:

THE FINANCIAL LITERACY OF ADULTS IN 14 COUNTRIES

INTRODUCTION

Financial literacy is rapidly being recognised as a core skill, essential for consumers operating in an increasingly complex financial landscape. It is therefore no surprise that governments around the world are interested in finding effective approaches to improve the level of financial literacy amongst their population and that many are in the process of creating or leading a national strategy for financial education to provide learning opportunities throughout a persons life (OECD/INFE, 2013b). The OECD defines financial education as follows: Financial education is the process by which individuals improve their understanding of financial products and concepts; and through information, instruction and/or objective advice develop the skills and confidence to become more aware of financial risks and opportunities, to make informed choices, to know where to go for help, and to take other effective actions to improve their financial well-being and protection (OECD 2005a). Financial education strategies benefit from empirical evidence to indicate the level of need amongst the population as a whole and within particular subgroups (OECD/INFE 2012). The measurement of financial literacy levels is therefore widely recognised as a priority for countries seeking to deliver financial education in an efficient manner and evaluate its impact at a national level. Such a measurement exercise allows policy makers to identify potential needs and gaps in relation to specific aspects of financial literacy and provides information about which groups of people are in need of most support. The results of the first financial literacy survey undertaken in a particular country can be taken as a baseline, and used to set benchmarks for financial education initiatives. Subsequent waves of a survey can be used to identify changes that have occurred during the interim period. National financial literacy surveys are clearly important tools, but the potential gain from a survey undertaken across a number of countries is much greater. Such an international study provides the opportunity to compare levels of financial literacy and progress across populations and financial markets, and is of huge interest to policy makers and other stakeholders seeking to understand why one country appears to be achieving more than another and which interventions are most effective. The OECD International Network on Financial Education (INFE) agreed to address the lack of internationally comparable data through the design and testing of a purpose built survey instrument. After many months of development and refinement, the OECD/INFE approved a core questionnaire, and countries were invited to utilise this OECD/INFE Financial Literacy Core Questionnaire according to an agreed methodology.

23

This instrument took as its starting point the following working definition: Financial literacy is a combination of awareness, knowledge, skill, attitude and behaviour necessary to make sound financial decisions and ultimately achieve individual financial wellbeing. The Core Questionnaire focuses on those aspects of knowledge, attitudes and behaviours that are associated with the overall concept of financial literacy. The questions cover a range of contexts, including accessing financial services, meeting immediate financial requirements and planning for the future. Almost all the questions relate directly to the individual answering the question, although some information is collected about the household, including the household income and the number of people living with the respondent. The core questions have been picked because they are applicable to the vast majority of people and they are suitable across a wide range of countries. Each question is designed to be asked in the same way in each country, but country-specific responses are possible (for example when respondents are asked about their methods of saving, or types of financial product, their responses will reflect the local financial market). This enables cross country comparisons that are contextually meaningful, while maximising the potential to include all interested countries. Participating countries and data collection This chapter reports on 14 countries: of these, 12 originally volunteered to take part in a measurement project using the Core Questionnaire (Armenia; Czech Republic; Estonia; Germany; Hungary; Ireland; Malaysia; Norway; Peru; Poland; South Africa and the United Kingdom), whilst the 13 th, Albania joined the project at a later stage with the guidance of our INFE Measurement Subgroup Expert from Italy. The 14th country, The British Virgin Islands (BVI), did not participate in the project, but undertook a national survey using the Core Questionnaire and shared the data with the OECD/INFE. Each country aimed to interview at least 1,000 individuals and data has been weighted to reflect the national population in terms of basic demographics. The data collection process began in the second half of 2010 and was completed early 201112. Each participating country undertook a nationally representative survey using the Core Questionnaire, collected the results in an anonymised database and provided a dataset to the OECD Secretariat for analysis. Additional discussion about the methods used to develop measures of financial literacy is presented in Annex 1, whilst further information about the countries that have participated and the approaches that they used can be found in the Annex 2 (Table 17). The structure and content of this chapter The following three sections focus on each of the three components of financial literacy: knowledge, behaviour and attitudes. Specific questions are reported as well as average score across several questions and the distribution of these scores.

12

With the exception of data from Albania and BVI, which were received in 2011. Note that the achieved sample size for BVI is 535. 24

These sections are followed by analysis of the relationship between knowledge and behaviour, and between attitudes and behaviour. The chapter then discusses how the population can be segmented according to their strengths and weaknesses in terms of financial literacy, and looks at overall levels of financial literacy by combining the scores from each of the three components. The final section describes how levels of financial literacy vary by key demographic factors. The analysis is particularly powerful in showing general patterns in financial literacy around the world. However, as highlighted in the text that follows, the reader should apply caution when making specific cross-country comparisons from this data, as there are some variations in the questions used and the methods employed.

25

FINANCIAL KNOWLEDGE

A financially literate person will have some basic knowledge of key financial concepts and the ability to apply numeracy skills in financial situations. The Core Questionnaire therefore asks a range of questions in relation to concepts such as simple and compound interest, risk and return, and inflation. Survey questions designed to test knowledge The Core Questionnaire includes 8 survey questions designed to test aspects of financial knowledge that are considered to be relevant across countries (Table 1). These vary in style and content in order to avoid undue biases that could be caused by different ways of processing information across certain types of people or cultural norms. Whilst some knowledge questions allow a person to give a completely free response others provide a list of possible answers, from which the respondent must choose their response. The questionnaire also encourages respondents to say if they don't know the answer to something, in order to dissuade them from guessing (this helps to ensure that the survey captures actual levels of knowledge rather than lucky guesses). In some countries questions were amended or substituted, as indicated in Table 113. To some extent, this limits our ability to make cross-country comparisons and in particular, caution should be applied when comparing results from Hungary and Norway with the other countries or with each other. The questions can only provide meaningful information about the level of financial literacy of individuals and populations if they are sufficiently varied to differentiate between high and low achievers by combining a mixture of easy and more difficult problems. The analysis of responses to each question shows that the spread of difficulty in the questionnaire is appropriate; differentiating well both within countries and across countries. There are also a sufficient number of questions to provide a good overview of a persons basic knowledge, indicate general willingness to absorb financial information and an ability to apply knowledge to particular problems14. A high score indicates that someone has a high level of financial knowledge, but does not necessarily suggest that they are financial experts.

13

All countries made essential edits to currency units. In some cases this also required changing the amount to reflect national prices. An international survey is not intended to capture country-specific knowledge, such as understanding the tax system within a county, or knowing about the retirement provision provided by the state. Countries wishing to find out more about the levels of knowledge amongst their population are encouraged to draw on the INFE Supplementary Questions: Additional, Optional Survey Questions available at www.financial-education.org. 26

14

Table 1. The 8 knowledge questions

Question as in the original version of questionnaire (and response codes with correct response in bold text) Division Imagine that five brothers are given a gift of $1000. If the brothers have to share the money equally how much does each one get? [Open response: $200] Now imagine that the brothers have to wait for one year to get their share of the X. In one years time will they be able to buy: Multiple choice: a) More, b) the same amount, or c) less than they could buy today. Interviewers also recorded 2 other responses which were considered to be correct: it depends on inflation, it depends on the types of things they want to buy You lend X to a friend one evening and he gives you X back the next day. How much interest has he paid on this loan? [Open response: 0] Changes to core questions

Time-value of money

Armenia used a previous version of the core questionnaire, which asked about a lottery prize rather than a gift. Not asked in Norway*. Norway: Imagine that you get a gift of 1000kr, and you put it in the drawer at home for 12 months. After one year how much could you buy for this money? Armenia had 4 options (excluding: It depends on the types of things)

Interest paid on a loan

Calculation of interest plus principle

Compound interest

Risk and return

Suppose you put $100 into a savings account with a guaranteed interest rate of 2% per year. You dont make any further payments into this account and you dont withdraw any money. How much would be in the account at the end of the first year, once the interest payment is made? [Open response: $102] and how much would be in the account at the end of five years? Would it be: a) More than $110 b) Exactly $110 c) Less than $110 d) Or is it impossible to tell from the information given An investment with a high return is likely to be high risk [True/False]

Not asked in Albania and Norway*. In Malaysia respondents were asked what return was earned on the loan. This question is not used in the score for Peru, due to problems with data coding. Hungary changed the interest rate to 5%

Hungary: The options given were less than simple interest, making it impossible to know whether respondent identified the impact of compound interest or just calculated simple interest RSA: If someone offers you the chance to make a lot there is a chance that you will lose a lot Not asked in Norway

Definition of inflation Diversification

High inflation means that the cost of living is increasing rapidly [True/False] It is usually possible to reduce the risk of investing in the stock market by buying a wide range of stocks and shares [True/False]

RSA: It is less likely that you will lose all of your money if you save it in more than one place (Yes) Norway and BVI: To buy a single share carries less risk than buying shares in mutual funds (No)

*Norway asked three alternative questions which have been used in their score: What is the nominal interest rate? What is meant by the effective interest rate? True/False: When you buy shares in a company you lend money to the company? Note that these questions were slightly edited as a result of feedback from the initial project.

27

The results reported in Table 215 show that most people in most of the countries could use mental arithmetic to undertake a simple division. However, despite the widespread ability across all of the countries there are still quite wide variations in the proportion of respondents who gave an incorrect answer, from just 1% in Malaysia through to 17% in the United Kingdom.
Table 2. Correct responses to knowledge questions
Proportion giving correct response (Cell percentages by country).

Division

Timevalue of money

Interest paid on loan

Calculation of interest plus principle

Compound interest and correct answer to previous question

10% 18% 32% 31% 47% 46% 29% 30% 54% 14% 27% 21% 37% 20%

Risk and return

Definition of inflation

Diversification

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway* Peru Poland South Africa UK BVI**

89% 86% 93% 93% 84% 96% 93% 93% 61% 90% 91% 79% 76% 84%

61% 83% 80% 86% 61% 78% 58% 62% 87% 63% 77% 49% 61% 74%

87% 88% 84% 88% 95% 88% 93% 61% 85% 65% 90% 99%

40% 53% 60% 64% 64% 61% 76% 54% 75% 40% 60% 44% 61% 63%

77% 67% 81% 72% 79% 86% 84% 82% 61% 69% 48% 73% 77% 83%

81% 57% 70% 85% 87% 91% 88% 74% 68% 86% 80% 78% 94% 87%

63% 59% 54% 57% 60% 61% 47% 43% 51% 51% 55% 48% 55% 41%

Empty cells have no relevant observations, including those where the response was not recorded. See Table 5 for question text. *The results reported for Norway under the Division column actually refer to an alternative questions posed: What is the nominal interest rate. Norway also slightly reworded the time value of money question, as they had not asked the previous question. Under interest for Norway, the table reports answers to: What is meant by the effective interest rate. **For diversification Norway and BVI asked Buying a single companys stock usually provides a safer return than a stock mutual fund .

Fewer respondents gave a logical answer to the follow-up question designed to identify those who understand how inflation impacts on the value of fixed cash amount. In South Africa fewer than half (49%) believed that the money would buy less in one years time despite the fact that inflation in that country was over 4% at the time of the survey. Only in Armenia, Czech Republic, Estonia and Norway did at least 4 out of 5 people give a correct response to this question. The concept of interest being paid on a loan appears to be widely understood; indeed in BVI 99% of respondents gave a correct response. The question requires the simplest possible arithmetic, but records an open ended response to minimise the possibility that people guess the correct answer. People found it harder to calculate a percentage than to undertake a division. Between 40% (Albania, Peru) and 76% (Ireland) of respondents gave a correct response to the first saving related question requiring calculation of interest plus principle (Table 19, Annex 2). The follow-up question was found to be harder still, particularly in Albania, Armenia and Peru: only 10%, 18% and 14% respectively showed that they could calculate simple interest and understand the impact of compounding.
15

Tables 18, 19, and 20 in Annex 2 provide a more detailed breakdown of responses to the knowledge questions. 28

Individuals in Hungary (86%) and Ireland (84%) were most likely to understand the basic concept of risk and return, whilst fewer than half of the Polish respondents appear to have grasped the relationship as described (48%) (Table 20, Annex 2)16. In all other countries over 60% of respondents gave a correct answer to this question. Most respondents knew that high inflation meant the cost of living was increasing, suggesting an awareness of simple economic terms. It appears that in most countries people are more likely to know the definition of inflation than know what impact it has on their spending power, but in Armenia considerably more people understood the time value of money than recognised the definition. The various diversification questions used in different countries proved to be challenging. Up to 37% of respondents claimed not to know the answer to the question used in Norway, and no more than 61% per cent of respondents in any of the participating countries gave a correct response (Hungary). The extent to which people said that they didn't know the answer to a question varies by question, and country. For example, in the Albania, South Africa and the UK around 1 in 10 respondents (10%, 10%, 8%) reported that they didn't know the answer to the division question; in Malaysia just 1 in 100 gave this response. Furthermore, almost half of Albanians (45%) said they didn't know how much money would be in a savings account at the end of the year, compared to just 2% in Poland and 3% in Norway. A financial knowledge score Analysis of the responses to each question by country indicates that the combination of knowledge questions adequately identified high and low achievers in all countries. It also shows that relatively few people refused to answer the questions17. The financial knowledge score therefore uses all 8 questions. It would be possible to create a score for each respondent from the factor analysis, and this approach is widely considered to be good practice when scoring complex data18. However, there is also a strong argument for giving each component of financial knowledge equal weighting, as each has benefits for individuals, and each has been identified as important by international experts. There is also some sense in avoiding complex statistical approaches if these are likely to be applied or interpreted in different ways in different countries or if problems with data from one country are likely to influence the way the data from other countries is analysed. The results that follow therefore report a simple count of correct answers which has been rescaled to take values from 0 to 100. This approach is in keeping with the development process: the questions within the Core Questionnaire were all chosen because they were considered to capture essential aspects of financial knowledge.

16

Note that in Albania more than one in 5 appears to have got the question wrong. However, Albania did not record don't know responses. In the case of the knowledge questions, don't know is considered to be a valid response indeed respondents were encouraged to say if they didn't know the answer - and so this will not cause a problem in analysis. This argument is based on the fact that a score that counts correct answers may be misinterpreted; people may assume it is equally difficult to gain one additional point from anywhere on the scale when in fact some questions are more difficult than others. 29

17

18

The process of counting correct answers began by assigning a value to the responses to each question (Table 3). Where countries have substituted questions, or reworded them, the replacement question has also given a value of 1 to a correct response and 0 in all other cases. In the case of a country with fewer than 8 financial knowledge questions each score has been rescaled as necessary (typically multiplying by a factor of 8/7).
Table 3. Creating a knowledge score
Question Discussion This is open response and a correct answer is therefore a good indicator of applied numeracy This is multiple response and country/context specific Value towards final score 1 for correct response. 0 in all other cases. 1 for responses c, d, e unless country tells us otherwise 1 for correct response. 0 in all other cases. 1 for correct response. 0 in all other cases. 1 for a correct response IFF the previous response was also correct. 0 in all other cases. 1 for a correct response. 0 in all other cases. 1 for a correct response. 0 in all other cases. 1 for a correct response. 0 in all other cases.

Division Time-value of money

Interest paid on a loan

This is open response and a correct answer is therefore a good indicator of understanding This is open response and a correct answer is therefore a good indicator of applied numeracy This is multiple response. Assumption is that if the respondent couldn't calculate 2% they also cannot calculate 5*2%. This is a yes/no question

Calculation of interest plus principle Compound interest

Risk and return

Definition of inflation

This is a yes/no question

Diversification

This is a yes/no question

It is important to be cautious about making firm conclusions about the similarities and differences of countries from the data because of the different questions used in some cases. This chapter therefore employs a statistically conservative approach to analysing across the countries in order to look for general patterns by grouping the countries according to whether or not their average results are significantly different. This approach does not in any way correct for the variation in questions asked, but on the assumption that they are largely equivalent, it attempts to identify groups of similar countries. Countries within the same subset do not have significantly different average scores. Countries that are present in more than one subset have scores that are between the two groups i.e. they are not significantly different from either group. Figure 1 below shows that the 14 countries fell into 7 groups based on levels of financial knowledge, although there is considerable overlap, with only 5 countries falling neatly into one group. The figure shows that average scores in South Africa put this country in the lowest scoring group, similar only to

30

Peru. At the other end of the figure, Malaysia, the UK, BVI, Czech Republic, Ireland, Germany and Estonia also had similar scores, but with a significantly higher average. Hungary appears to be uniquely high scoring, but it should be remembered that this will partly be capturing the changes made to one of the more difficult questions.
Figure 1. Country groupings by average financial knowledge scores
Lower average scores South Africa Peru Albania Armenia Norway Poland Malaysia UK BVI Czech Republic Ireland Germany Estonia Hungary Higher average scores

Shaded boxes indicate homogenous groups of countries, computed in IBM SPSS19 using Scheffe Homogenous Subsets on knowledge scores rounded to the nearest whole number. The order of the countries in this figure reflects average financial knowledge scores. Average scores range from 57.5 to 76.3.

The distribution of scores is also interesting (Figure 2). Some countries, such as Albania, Norway, and South Africa had a relatively large proportion of the population with low scores (to the left of the distributions). However, the modal (the most frequently achieved) score in most countries was more than half (Albania, Armenia, Czech Republic, Estonia, Hungary, Ireland, Malaysia, Norway, Peru, Poland, UK and BVI). Just one country, Germany, had the maximum value as the modal value. Countries such as the Czech Republic, Estonia and Germany with a negatively skewed distribution can be reassured that the majority of their population had basic financial knowledge. However, every country has some proportion of the population that achieved a low score on the knowledge test, showing that there was room for improvement in all countries.

31

Figure 2. Distribution of knowledge scores 30% 25% 20% 15% 10% 5% 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland

30% 25% 20% 15% 10% 5% 0% Malaysia Norway Peru Poland South Africa United Kingdom BVI

Scores from 0 (far left column for each country) 100 (far right column for each country).

32

The following chart (Figure 3) focuses specifically on the percentage of respondents that gained a high score. In Albania, Armenia, Peru, Poland and South Africa fewer than half of the respondents gave correct responses to at least of the questions.
Figure 3. Financial knowledge: Percentage with a high score 57% 45% 46% 61% 58% 69% 60% 57%

51% 40% 41%

49% 33%

53%

Poland

Czech Republic

Malaysia

Albania

Armenia

Estonia

Hungary

Ireland

Germany

Norway

Peru

South Africa

Base: all respondents. Lighter shaded columns indicate countries where fewer than 50% achieved a high score.

33

United Kingdom

BVI

FINANCIAL BEHAVIOUR

Behaviour is an essential element of financial literacy; and arguably the most important. The positive outcomes from being financially literate are driven by behaviour such as planning expenditure and building up a financial safety net; conversely, certain behaviours, such as over-using credit, can reduce financial wellbeing. This section therefore focuses on a wide range of behaviours, with an emphasis on those that can enhance or reduce financial wellbeing. Financial behaviour across different domains The OECD/INFE Financial Literacy Core Questionnaire asks the respondents about their behaviour using different question styles, in order to capture the maximum amount of information. From the responses to these questions, it is possible to derive information about the ways in which people manage their money, including considering carefully whether they can afford something, paying bills on time, and keeping a close watch over finances. Questions also ask whether they attempt to save and set long term goals, if they are personally (or jointly) responsible for a household budget, how they choose financial products and if they have recently borrowed to make ends meet (see Table 4 for a summary of responses; more detailed information is provided in Annex 2). Four of the questions use a qualitative scale, enabling people to provide more information about the frequency of their behaviour. These scaled questions have been asked in the same way in each of the participating countries except Norway and South Africa19. Comparisons across the majority of countries should therefore be robust. A financially literate person will always have an idea of the amount of money they can afford to spend on a purchase. The first of the behaviour statements shows that people typically did consider whether they could afford potential purchases (Table 21, Annex 2). This is especially the case in Armenia, Malaysia and Peru. However, in Norway 14% percent of respondents put themselves below the midpoint indicating that they tended not to consider their purchases; 1 in 10 UK respondents also put themselves at this end of the scale. It is also interesting that in Estonia and Poland around one in 5 respondents put themselves at the midpoint on the scale, suggesting that they were aware that they sometimes made purchases without considering affordability.

19

In Norway the questions were asked on a 7 point scale: recoded as follows for the purpose of comparisons: 1=1 (2, 3=2) (4=3) (5, 6=4) (7=5). In South Africa, a 5 point scale was used, but each of the scale points was given a verbal description. Note that Armenia labelled the scale in reverse, and asked the questions alongside some of the attitudinal questions. 34

Table 4. Positive financial behaviours by country

Cell percentages by country.

Behaviour statements
Carefully considers purchases Pays bills on time Keeps close watch on personal financial affairs Sets long term goals and strives to achieve them
30% 58% 36% 41% 61% 52% 56% 64% 59% 71% 46% 55% 43% 68%

Financial product choice

Responsible and has a household budget Has been actively saving or buying investments in the past year after gathering some info after shopping around and using independent info or advice Has not borrowed to make ends meet

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

87% 91% 75% 68% 82% 86% 83% 92% 72% 91% 70% 83% 77% 87%

77% 94% 85% 83% 96% 82% 85% 69% 79% 86% 78% 61% 89% 83%

71% 81% 76% 78% 87% 71% 85% 78% 89% 82% 81% 65% 80% 80%

59% 51% 37% 28% 22% 31% 54% 74% 25% 49% 54% 43% 43% 43%

42% 36% 72% 36% 86% 27% 53% 97% 71% 62% 51% 53% 68% 83%

49% 42% 28% 24% 52% 48% 39% 39% 57% 52% 32% 56% 29% 70%

2% 10% 8% 5% 4% 10% 3% 5% 4% 2% 3% 16% 2%

69% 53% 89% 78% 96% 86% 86% 79% 93% 73% 79% 74% 91% 87%

The first 4 columns report people putting themselves at 4 or 5 on a scale from Never=1 to Always=5. The financial product choice data is used in the final measure as follows: 1 point for gathering some information (column 7 above), 2 points for shopping around and using independent information and advice (column 8 above).

Financial literacy also requires organisational skills in order that individuals meet their financial commitments and thus avoid problems such as reduced access to affordable credit or fines for nonpayment. The questionnaire therefore asks each respondent whether they usually pay their bills on time (Table 22 Annex 2). Most respondents reported that they did putting themselves at 4 or 5 on the scale. However, a sizeable proportion of respondents in South Africa (15%) and Norway (15%) indicated that they never or rarely did so (less than 3 on the scale)20. A further 1 in 5 respondents in South Africa (19%) as well as 19% of respondents in Malaysia put themselves at the midpoint, suggesting that they too were not paying many bills on time. These negative responses may be due to a variety of reasons including lack of money, lack of access to electronic payment facilities or a tendency to be disorganised or unwilling to meet responsibilities on time, but in all cases they suggest that a sizeable proportion of consumers could be either encouraged or supported to improve this behaviour. A third behaviour statement asks respondents how often they keep a close personal watch on their financial affairs (Table 23, Annex 2). Keeping an eye on financial affairs is important for a variety of reasons. For those who use financial products, it is essential to be aware of anticipated withdrawals from
20

It is possible that people chose the Never category because they had no bills to pay. In Norway such people could have simply skipped the question. 35

an account and to check statements in order to address mistakes or fraudulent activity, such as duplicate amounts being withdrawn through computer error or unauthorised use of credit cards. Even those who do not use financial products need to oversee their financial affairs in order to keep their savings safe, smooth their expenditure and pay bills on time. Very few people claimed that they never keep an eye on their own finances ranging from 1% in Norway, Ireland, Germany and Peru to 8% in Hungary. However, in almost all the participating countries more than one in ten respondents put themselves in category 3 suggesting that were aware they could do more. In South Africa around a third of all respondents (32%) put themselves on the lower part of the scale (1, 2, or 3) suggesting a considerable need to help people see the value of watching over their own finances. In Peru, this question produced a very clear clustering effect; people tended to report that they either always kept a close personal watch (68%), or that they hardly ever did (point 2 on the scale: 19%). The final statement in this set relates to acting on longer term plans (Table 24, Annex 2). It asks whether respondents set long term financial goals and strive to achieve them; it does not specify how far away the goal should be, or how easy it might be to achieve. Long term financial goals may be related to accruing money for specific expenses, such as education fees or a wedding. Alternatively they could relate to investment strategies, saving for retirement, business ideas or career progression. The second phrase in this statement indicates that the respondent should be attempting to reach their goal, rather than simply thinking about it. Despite the fact that everyone can benefit from considering their longer term financial needs this particular behaviour does not appear to be widespread. As many as one in five people in the United Kingdom (22%) said that they never set a long term financial goal and worked to achieve it. It appears that Peruvians (55%), British Virgin Islanders (45%) and Armenians (43%) are the most likely to set long term goals. More than 1 in 10 Estonians responded that they didn't know whether this statement applied to them, perhaps indicating disengagement with long term planning. A further placed themselves at 1 or 2 on the scale suggesting that setting goals was not something that they did. A sizeable proportion of respondents in each country (ranging from 12% to 26%) put themselves at the midpoint on this scale. This could be classified as sometimes. Interpreted in this way, it indicates that people do not consistently work towards long term goals. Further behaviour questions provide us with information about the extent to which an individual takes responsibility for household finances and budgeting21. Responses to two questions have been combined to assess how many people report that they a) have either personal or joint responsibility for day to day money management decisions in their household and b) live in a household with a budget (Figure 4). This has been done to ensure that someone is not considered to be actively using a budget if they do not take on any responsibility for household finances22. The combination of the two sets of responses shows a very wide variation across countries, with fewer than 1/4 of respondents in Germany

21 22

The term budget was explained to the respondent to ensure that they did not misinterpret it. This measure will not capture those people who budget their own money but do not take responsibility for the household as both the questions relate to household money management. 36

and Estonia being personally or jointly financially responsible and budgeting through to almost 3/4 of those in Malaysia (74%)23.
Figure 4. Responsible and has a household budget

74% 59% 51% 37% 28% 22% 31% 54% 25% 49% 54% 43% 43% 43%

United Kingdom

Albania

Hungary

Estonia

Ireland

Poland

Germany

Malaysia

Norway

Peru

Base: all respondents. % of respondents have some responsibility for financial decisions in a household with a budget.

Saving behaviour is considered to be an important component of financial literacy building financial security and reducing the reliance on credit. As the actual amount that a person can save, and the length of time they can keep money to one side varies immensely, the financial literacy measure focuses exclusively on whether or not respondents save money. Respondents were asked In the past 12 months have you been saving money in any of the following ways?. The questionnaire then lists a variety of ways in which people typically save, in order to prompt recollection of any type of saving. This was tailored to the country context but typically included saving money at home, using informal savings clubs, putting money into savings accounts and buying investments. For the purpose of the international comparison a variable has been created that counts all kinds of saving as active saving, except for the passive approach of building up a balance in a current account24. This is an appropriate indicator of behaviour, since it indicates that saving was intentional rather than a default position due to income exceeding outgoings. As can be seen from Table 5, discussing savings is a sensitive issue in some countries. Almost one in 5 respondents refused to answer this question in the Czech Republic (19%), as did 14% in South Africa. In Poland a very large proportion claimed that they didn't know, which almost certainly indicates an unwillingness to divulge such information. Malaysia clearly has a culture of saving, 97% of respondents reported that they had been saving in the past 12 months and nobody appeared to feel that this was sensitive information. Conversely, in

23 24

Data from Armenia indicates that everyone takes some responsibility for household money management. The measure of savings used by the Czech Republic includes a product known as pension insurance. Whilst this has some of the characteristics of a pension, it is widely considered to be a standard savings vehicle by Czech consumers. 37

Czech Republic

South Africa

Armenia

BVI

Hungary, people were very unlikely to have been saving (just 27% responded positively), although again, nobody refused to answer.
Table 5. Actively saving or buying investments in the past 12 months Refused 2% 19% 2% Don't know 4% 3% 4% 2% 1% 1% 21% 9% 2% No 52% 64% 6% 58% 13% 71% 46% 2% 29% 36% 16% 24% 29% 17% Yes 42% 36% 72% 36% 86% 27% 53% 97% 71% 62% 51% 53% 68% 83%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

1%

12% 14% 1%

Base: all respondents. Row percentages by country.

The way people behave when choosing financial products is also an important aspect of their overall financial literacy. If people attempt to make an informed decision by shopping around or using independent advice they are more likely to choose appropriate products that meet their needs in a cost effective way, less likely to buy something inappropriate, and less likely to be subject to mis-selling or fraud. People do not typically choose financial products on a weekly, or even monthly, basis and so respondents are asked about a product chosen in the last 2 years (excluding simple renewals)25. It is important to note that this measure is specific to choosing products, and does not capture information about people who checked that their existing products were still suitable, unless they went on to shop for something new. Neither does it capture intention to behave such as how they think they might choose a product in the future. The possible approaches to choosing financial products may vary by country (and countries were able to add their own options to the questionnaire), but shopping around and gathering information are the behaviours that are most relevant. In the derived variable used in the final score, respondents are considered to have made some attempt to make an informed decision if they tried to compare across providers (even if they found out that there were no other providers), or if they sought information from someone. Figure 5 shows that consumers in Germany, Ireland and the United Kingdom were most likely to have made active financial product choices by shopping around and using independent information or advice.
25

The score should be seen as a conservative estimate since some of those who refused may actually have shopped around- i.e. their score could actually be higher. Some countries did not supply information to differentiate between those who did not choose a product, and those who refused to answer this question. 38

Figure 5. Shopping around for financial products 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI
Base: all respondents. Respondents who fall into the no score category include those who chose a product wit hout attempting to make an informed decision and those who had not chosen a product.

10%

20%

30%

40%

50%

60%

70%

80%

90% 100%

No score

1 point: Some attempt to make informed decision

2 points: Shopped around and used independent info or advice

Financially literate people will have strategies to smooth income flows and a tendency to avoid using credit for essentials such as food and utilities. The extent to which these strategies are successful will depend on the predictability of their income and expenditure as well as the extent to which they have the necessary skills. It is not always possible to prevent shortfalls in income, but a reliance on credit for basic living can become very dangerous and impossible to escape. A variable has therefore been created to identify people who reported that sometimes their income didn't meet their needs, and that the last time this happened they had to borrow to make ends meet. Figure 6 shows that on the whole, respondents were unlikely to have resorted to credit use to make ends meet. However, almost a half of Armenians had done so in the last 12 months (47%), indicating a worrying vulnerability to income fluctuations and the risk of facing spiralling debt problems.

39

Figure 6. Borrowing to make ends meet 47% 31% 22% 11% 4% Malaysia Czech Republic Albania Armenia Hungary Estonia Ireland Germany 14% 14% 7% South Africa 21% 27% 21%

26% 9% 13%

Peru

Norway

Poland

Base: all respondents. This variable combines the responses to two questions: Sometimes people find that their income does not quite cover their living costs. In the last 12 months, has this happened to you? & If so what did you do to make ends meet the last time this happened? Note that the question does not record all the strategies used to make ends meet in the course of 1 year, but only on the last occasion this occurred. The measure captures borrowing informally and formally, including (but not limited to) from family, pawn brokers, savings and loans club as well as relying on an overdraft, taking a bank loan and taking money from a flexible mortgage, It does not include late payment of bills.

A score for financial behaviours Information from the results described above has been combined into an overall score. This looks at the number of behaviours that are in evidence for each person, with the caveat that respondents who refused to answer particular questions (or opted for dont know in some cases) may appear to be lower scoring than is actually be the case26. The questionnaire captures a wide range of financial behaviours at an international level. Incorporating all of these measures into an overall score ensures a nuanced indicator that provides a good indication as to the extent to which individuals are behaving in a financially literate way.

26

There are various other ways in which missing data could be handled, including deleting the observations or replacing missing information with imputed variables. The approach taken will tend to underestimate overall scores, a conservative approach which is appropriate for this study. 40

United Kingdom

BVI

The score is created with 9 points for showing evidence of certain positive financial behaviours, as shown in Table 6. For reporting purposes this has been rescaled from 0 to 100.
Table 6. Creating a behaviour score
Behaviour Considered purchase Discussion This is a scaled response. Value towards final score 1 point for respondents who put themselves at 4 or 5 on the scale. 0 in all other cases. 1 point for respondents who put themselves at 4 or 5 on the scale. 0 in all other cases. 1 point for respondents who put themselves at 4 or 5 on the scale. 0 in all other cases. 1 point for respondents who put themselves at 4 or 5 on the scale. 0 in all other cases. 1 point if personally or jointly responsible for money management and has a budget. 0 in all other cases. 1 point for any type of active saving (excluding letting money build up in a current account as this is not active). 0 in all other cases 1 point for people who had tried to shop around or gather any information. 2 points for those who had shopped around and gathered independent information. 0 in all other cases.

Timely bill payment

This is a scaled response.

Keeping watch financial affairs Long term goal setting

of

This is a scaled response.

financial

This is a scaled response.

Responsible and has a household budget

This is a derived variable, created from the responses to two questions.

Active saving

This question identifies a range of different ways in which the respondent may save. People who refused to answer score 0. This is a derived variable drawing information from 2 questions. It is only possible to score points on this measure if the respondent had chosen a product: those with no score on this measure have either refused to answer, not chosen a product, or not made any attempt to make an informed decision. This is a derived variable that combines a question about running short of money and one that identifies a range of different ways in which the respondent made ends meet the last time they ran short of money. The derived variable indicates people who are making ends meet without borrowing (refusals will score 1).

Choosing products

Borrowing ends meet

to

make

0 if the respondent used credit to make ends meet. 1 in all other cases.

41

As with the knowledge scores, a statistical approach has been used to group countries according to their average financial behaviour scores (Figure 7). This provides an indicator of the extent to which individuals are exhibiting behaviour relating to the various aspects of financial literacy discussed above. Some countries have relatively similar financial behaviour scores; the analysis indicates there are 5 groups of countries that had similar average scores. Estonia exhibited significantly lower levels of behaviour than all other countries except Albania. The highest numbers of positive financial behaviours were apparent in Germany, Malaysia and BVI.
Figure 7. Country groupings by average financial behaviour scores
Lower behaviour scores Higher behaviour scores

Estonia Albania Hungary South Africa Poland Armenia Czech Republic UK Norway Ireland Peru Germany Malaysia BVI
Shaded boxes indicate homogenous groups of countries, computed in IBM SPSS19 using Scheffe Homogenous Subsets. Average scores range from 50.0 to 67.8.

42

It is also noteworthy that Germany and Malaysia had nobody who scored zero (identifying those with no positive financial behaviours), and just two countries had nobody who achieved a maximum score (Armenia and BVI) (Table 25, Annex 2). Figure 8 shows that large proportions of respondents in Germany, Malaysia and BVI scored relatively highly (characterised by a lighter colour on the bar charts). However, the distributions all tend towards a steep bell shape, indicating that there was a large proportion of the population of most countries who only exhibited around half of the positive behaviours identified. South Africa has the flattest distribution, showing that there was a wider range of behaviour in evidence across the country.
Figure 8. Distribution of financial behaviour scores 30% 20% 10% 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland

0.3 0.2 0.1 0 Malaysia Norway Peru Poland South Africa United Kingdom BVI

Scores from 0 (far left column for each country) 100 (far right column for each country).

43

Based on the distributions presented above and the fact that financial products are not chosen on a regular basis, those exhibiting 2/3 of the identifiable behaviours are considered to have a high score on this component of financial literacy. Analysis shows that only half of the countries had a majority of people gaining such a high score (Figure 9).
Figure 9. Financial behaviours: Percentage scoring 6 or more

67% 57% 48% 39% 41% 27% 38%

67% 59% 60% 51% 43% 43%

71%

Germany

Base: all respondents. Lighter shaded columns indicate countries where fewer than 50% achieved a high score.

44

South Africa

United Kingdom

Czech Republic

Malaysia

Albania

Peru

Armenia

Hungary

Estonia

Ireland

Norway

Poland

BVI

FINANCIAL ATTITUDES

Attitudes and preferences are considered to be an important element of financial literacy. If people have a rather negative attitude towards saving for their future, for example, it is argued that they will be less inclined to undertake such behaviour. Similarly, if they prefer to prioritise short terms wants then they are unlikely to provide themselves with emergency savings or to make longer term financial plans. The section that follows therefore reports the responses to three statements focusing on attitudes towards money, and particularly towards planning for the future. Attitudes and preferences: short term or longer term? The Core Questionnaire includes three scaled attitudinal questions: I find it more satisfying to spend money than to save it for the long term, I tend to live for today and let tomorrow take care of itself, and Money is there to be spent. Detailed results from each of the attitude statements are reported in Tables 26, 27 and 28 in Annex 2. Around a third of respondents in Germany, Hungary, Poland and the UK put themselves at the midpoint of the first statement I find it more satisfying to spend money than to save it for the long term suggesting that they found equal satisfaction in spending and saving. People in South Africa were the least likely to give a neutral response: just 14% put themselves at 3 on the scale, compared with 35% of UK respondents. Armenians were the most likely to report that they completely agreed that spending was more satisfying (56%), whilst Peruvian respondents were most likely to completely disagree with the statement (50%), followed by BVI (41%). The responses to the second attitude statement show that in most countries, respondents tended not to live for the day (Table 27). In Armenia, Czech Republic, Hungary and Peru, over half of the respondents completely disagreed with the statement. Despite this, almost 1 in 5 respondents in Armenia and Poland completely agreed with this statement showing considerable polarisation in Armenia. The third attitude statement relates specifically to peoples attitude towards money (Table 28). Here a large proportion of people were ambivalent (42% of Hungarians put themselves at 3 on the scale). Peruvians were the most conservative, with 31% completely disagreeing that money is there to be spent. In contrast, just 1% of Armenians and 4% of Polish respondents disagreed with the statement; indeed in Armenia 74% completely agreed.

45

Combining the various attitudes Exploratory factor analysis indicates that the three attitude questions capture an underlying attitude, indicating whether the respondent tends towards short-term gratification, or long term security. An average attitude score has therefore been created by adding together the responses to each of the three questions, and then dividing by 327. The average score varies significantly across countries (Table 29, Annex 2). The highest average attitudinal score across the 3 attitudes is 3.7 (Albania and Peru) suggesting that respondents in these countries are more likely than other countries to have a positive attitude towards planning for the future. The distribution of (rounded) scores is shown below (Figure 10). Estonia has a rather flat distribution compared with the other countries (indicating a wide range of attitudes in the country). Armenia has a very pointed distribution, showing that many people have the same attitudes. Poland exhibits large clusters of people who typically put themselves at 1 or 2 on the scale, whilst Albania, Hungary, and Peru there is a much more positive attitude towards the long term.
Figure 10. Distribution of financial attitude scores 60% 50% 40% 30% 20% 10% 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland

60% 50% 40% 30% 20% 10% 0% Malaysia Norway Peru Poland South Africa United Kingdom BVI

Scores from 1 (far left column for each country) 5 (far right column for each country).

27

Only the first of the attitudinal questions used in Norway have been used in this analysis in order to compare similar attitudes in our international comparison; some caution is therefore necessary when comparing scores from the other countries with the Norwegian score. Norway included a second attitude statement in their questionnaire: credit can be used for food and overhead costs. The vast proportion disagreed with this; just 15% agreed to some extent. This has not been included in the score as it is capturing something different. 46

As with the previous scores, it is also possible to identify those with higher scores (Figure 11). In this case, the higher scores identify those individuals with attitudes towards planning for the future that are considered to be positively related to financial wellbeing (which are described below as positive attitudes). In this instance, a high score is attributed to people whose average response across the attitude statements is greater than 3, even if only by a small percentage28. It shows that in most of the participating countries, people generally had positive attitudes towards financial matters. However, in Armenia and Poland this was not the case; in Armenia just over 1 in 10 respondents had positive attitudes and in Poland just over a quarter (27%) of respondents had generally positive attitudes.
Figure 11. Percentage of respondents with average score over 3 69% 69% 49% 53% 57% 27% 11% Albania Estonia Hungary Ireland Poland Norway 71% 54% 49% 67%

62% 46%

63%

Base: all respondents. Lighter shaded columns indicate countries where fewer than 50% with an average score over 3.

28

As 3 is the midpoint, these people are neutral. In order to have scored above 3 on average, they must have put themselves at 4 on the scale on at least one of the three questions: in the case of Norway, they must have put themselves at 4 or 5 on the single question used in the analysis. 47

United Kingdom

Germany

Peru

Czech Republic

South Africa

Malaysia

Armenia

BVI

RELATIONSHIP BETWEEN BEHAVIOUR AND THE OTHER SCORES

The literature often uses financial knowledge as a predictor of positive outcomes, which rely on behaviours that are consistent with financial wellbeing exactly the types of behaviours captured in our measure of financial literacy. This section therefore reports some very preliminary analysis to see whether behaviour scores increase with knowledge scores in each country. Figure 12 shows that across the countries surveyed there was a positive relationship between knowledge and behaviour higher knowledge scores are associated with higher behaviour scores29. However, the relationship does vary by country. For example, in Germany and Peru average behaviour scores only increase very slightly across the range of knowledge scores, and there is almost no increase in average behaviour scores in the higher knowledge categories in Estonia. Conversely, in Malaysia behaviour increases sharply with increased levels of knowledge, so that average behaviour scores amongst those with a high level of knowledge were almost double those of people with low levels of knowledge. These graphs also show that even people with very low levels of knowledge exhibited some positive financial behaviour. Whilst it is reassuring that they were financially active, it is very likely that they could benefit from additional knowledge in order to improve the quality of their decision making on matters such as choosing appropriate financial products or saving for the future.

29

The apparent dip in behaviour as knowledge moves from 0 to positive numbers should not be a concern: it is the result of analysing a very small number of people who had a score of 0. 48

Figure 12. Relationship between financial knowledge and behaviour

Average behaviour score by financial knowledge score

80 70 Average behaviour score 60 50 40 30 20 10 0 Financial knowledge score 80 70 Averagel behaviour score

Albania Armenia Czech Republic Estonia Germany Hungary Ireland

Malaysia Norway

60 50 40 30 20 10 0 Financial knowledge score Peru Poland South Africa United Kingdom BVI

It is also argued that those with positive attitudes towards the long term are more likely to behave in ways that are consistent with achieving long term goals, so further analysis looks at whether behaviour scores increase with attitude scores.

49

Figure 13 indicates a generally positive association between attitudes and behaviour: people with attitudes that tend towards short term gratification have lower behaviour scores than those with an attitude score over 3. There appears to be little relationship between attitudes and behaviour in BVI, and relatively little in Armenia. At the other extreme, behaviour increases notably with attitudes in Czech Republic and Ireland: for example average behaviour scores for Czechs with who prefer the short term are half those of people with positive attitudes towards the longer term.
Figure 13. Relationship between attitudes and financial behaviour
Average behaviour score by attitude score

80 70 60 50 40 30 20 10 0 1 2 3 4 5

Albania Armenia Czech Republic Estonia Germany Hungary Ireland

80 70 60 50 40 30 20 10 0 1 2 3 4 5

Malaysia Norway Peru Poland South Africa United Kingdom BVI

50

COMBINED MEASURES OF FINANCIAL LITERACY

Financial literacy is a combination of knowledge, attitude and behaviour, and so it makes sense to explore these three components in combination. This is done in two ways in this chapter- firstly by looking at whether people achieve high scores on several components, and secondly by adding the scores together. As these measures combine information from the scores developed in the previous chapters, the reader should apply the same caution when comparing countries that have made changes to the Core Questionnaire with other countries. Segmenting the population Table 7 summarises the proportion of respondents in each country achieving a high score on each of the three financial literacy components. There is some variation in relative strengths and weaknesses of respondents within countries: for example more than 4 times as many Armenians had a high knowledge score (46%) than a high attitude score (11%).
Table 7. High score on each of the financial literacy components High knowledge score
Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI
45% 46% 57% 61% 58% 69% 60% 51% 50% 41% 49% 33% 53% 57%

High behaviour score

39% 41% 48% 27% 67% 38% 57% 67% 59% 60% 43% 43% 51% 71%

High attitude score

69% 11% 62% 46% 63% 69% 49% 53% 57% 71% 27% 54% 49% 67%

Base: all respondents. Proportion scoring highly on each component (Cell percentages by country). Each of the columns reports % of respondents gaining a high score.

In addition to this aggregate data, it is possible to look at the relative strengths and weaknesses of individual respondents within countries, and use this to segment the population. The data has therefore been segmented according to whether respondents gained a high score in 0, 1, 2 or 3 of the three components described above. Figure 14 shows quite clearly that there are two typical patterns. Whilst all countries have people in each of the clusters, some tend to have the highest proportion in the cluster with one high score, whilst for others the highest proportion has two high scores. Germany and BVI stand out for having large portions of the population scoring highly on each component, indicating a high

51

overall level of financial literacy, whilst Armenia, Poland and South Africa have relatively large proportions of the population with low levels of financial literacy across the three measures. It is of some concern that more than 10% of the population in 10 of the countries had no high scores the exceptions being Germany, Hungary, Peru and BVI.
Figure 14. Financial Literacy Segments
Number of high scores (far left 0 high scores; far right: 3 high scores)

45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland

None 1 2 3

45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Malaysia Norway Peru Poland South Africa United Kingdom BVI

None 1 2 3

Developing an overall measure of financial literacy In order to assess overall levels of financial literacy, the three raw scores described above for knowledge, behaviour and attitudes have been summed to provide a single indicator. This takes into account the various aspects of financial literacy, including financial planning for the future, choosing financial products and managing money on a day-to-day basis. The final score, in its raw form can take a minimum value of 1, and a maximum value of 22. This has been rescaled to create a score that can take values between 1 100. As the three raw scores have different maximum values, the combined score is implicitly weighted. The most heavily weighted factor is behaviour. This is appropriate: behavioural questions make up a large

52

part of the questionnaire because financial behaviour is seen as a key component in financial literacy30. Financial knowledge also makes up a large percentage of the final score. Financial knowledge and behaviour are the two aspects of financial literacy most typically targeted by financial education initiatives. The score also contains a small component of attitudes towards money, and particularly towards planning for the future- those with more negative attitudes towards longer term planning therefore cannot achieve very high overall scores. Figure 15 shows that the countries can be grouped into 5 subsets with largely similar overall scores. Armenia, South Africa and Poland showed the greatest room for improvement, whilst Malaysia, Hungary, Germany and BVI had statistically similar, high scores. Taking an average of the country scores (not shown)31, countries typically scored around 63.2. Scores in the Czech Republic, Germany, Hungary, Ireland, Norway, Malaysia, Peru, the UK and BVI were above this average.
Figure 15. Country groupings by average overall financial literacy scores
Lower average overall score
Armenia South Africa Poland Estonia Albania Czech Republic Peru UK Norway Ireland Malaysia Hungary Germany BVI

Higher average overall score

Shaded boxes indicate homogenous groups of countries, computed in IBM SPSS19 using Scheffe Homogenous Subsets. Average scores range from 56.4 to 68.6.

30

Exploratory factor analysis also suggests that knowledge and behaviour should be weighted more heavily than attitudes. This average figure gives each country equal weight, regardless of population or sample size. 53

31

VARIATIONS BY SOCIO-DEMOGRAPHICS

Several approaches have been used to look at the association between socio demographic factors and the financial literacy of consumers. Firstly, tables are used to show the proportion of people within a particular demographic who fall into each segment described above. The average overall score for each demographic is also calculated; and finally multivariate linear regression analysis has been employed to explore the associations across several demographic categories simultaneously. This analysis focuses on key socio-demographic information: gender, age, education level, work status and income. It also takes into consideration attitude towards risk and income stability, as initial discussion with participating countries suggested that these may be important explanatory factors in understanding variations in financial literacy. Gender This section briefly highlights some of the key results of analysis of financial literacy by gender. Chapter 3 discusses gender differences in financial literacy and their implication for policy makers in more detail. Figure 16 shows that a larger proportion of male respondents gain high scores in knowledge than women in 13 of the countries studied. This is particularly marked in Norway, Poland and the UK with more than a 20 percentage point difference. There is no difference in the proportion of men and women gaining high scores in Hungary this is an important finding that cannot be explained by the different wording of one question.

54

In 8 of the countries surveyed, fewer than half of women have gained a high knowledge score; in comparison there are just 2 countries where less than 50% of men achieved such a score. It appears that in almost all countries where the average level of knowledge is relatively high, women are being left behind their male counterparts.
Figure 16. High knowledge score by gender
67% 69% 69% 50% 68% 54% 48%

63% 51% 35% 51% 52% 41%

66% 57%

Female
52%

Male
59% 45% 39%

67%

63% 53%

54%

29%

35%

36% 31%

40%

Peru

Czech Republic

Germany

Base: all respondents.

55

South Africa

United Kingdom

Malaysia

Armenia

Albania

Hungary

Estonia

Ireland

Norway

Poland

BVI

There are no differences in the proportions of men and women who gained high behaviour scores in Germany and Hungary (Figure 17) and only 2 countries show a difference of more than 10 percentage points (Norway and Ireland). In some countries, more men than women gained high marks (Albania, Armenia, Malaysia, South Africa and the UK), whilst in others (Czech Republic, Estonia, Ireland and Norway) women were more likely than men to achieve a high score. It is intriguing that this preliminary analysis suggests that the gender differences that are so apparent in the knowledge component are not entirely reflected in financial behaviours. This may be because many financial behaviours are undertaken at a household level such as saving or choosing products - whilst knowledge is entirely at the individual level. Even when the survey is designed to collect information about individuals it is inevitable that their behaviours will reflect household decisions.
Figure 17. High behaviour score by gender Female
67% 67% 43% 42% 38% 35% 51% 44% 30% 23% 37% 38%

Male
78% 66% 52% 46% 49% 43% 43% 41%

68% 65% 62% 62% 61% 55% 59% 51%

Czech Republic

Malaysia

Albania

Armenia

Peru

Hungary

Estonia

Ireland

Germany

Norway

Poland

South Africa

Base: all respondents.

56

United Kingdom

BVI

Analysis of the average attitude scores (Figure 18) shows that in most countries women were more likely than men to have high attitude scores showing that they typically had a more positive attitude towards the longer term. Only in Albania and Poland were men more likely than women to have attitudes that tend towards longer term preferences. In Armenia and South Africa there was no difference in attitudes by gender.
Figure 18. High attitude score by gender Female
72% 64% 66% 59% 47% 45% 65% 60% 73% 66%

Male
61% 54% 75% 69% 54% 54% 52% 46% 28% 26% 69% 66%

57% 55% 49% 43%

11% 11%

United Kingdom

Hungary

Albania

Estonia

Ireland

Poland

Germany

Norway

Peru

Base: all respondents. In Albania, Estonia, Germany, Poland and the UK women were more likely than men to have no high scores. This is not the case in other countries. In Hungary, men were over-represented in the lowest segment (showing that they were more likely than women to have no high scores), and the gender spread of the other categories is similar to the population as a whole. In South Africa, women were slightly more likely to have just one strength, and less likely to fall into the highest category.

Czech Republic

57

South Africa

Malaysia

Armenia

BVI

Figure 19 below shows that in some, but not all countries there was a small variation in overall score by gender. In no country did women score significantly more than men; almost certainly because of the large differences in levels of knowledge in most countries. The Czech Republic, Estonia, Hungary, Ireland and Malaysia exhibited no significant gender difference in overall score. Conversely, women scored significantly less in Albania, Armenia, Germany, Norway, Poland, South Africa, the UK and BVI.
Figure 19. Mean overall score by gender Female
57.8 62.3 55.4 57.2

Male

70.1 67.1 66.4 67.6 70.0 65.4 65.8 65.5 66.6 61.4 64.7 63.2 65.2 65.0 64.5 60.5 60.7 66.6 62.4 66.6 56.8 59.8 57.2 58.9

Estonia

Norway

Albania

Hungary

Poland

Germany

Ireland

Peru

South Africa

Malaysia

Base: all respondents.

Age It might be expected for financial literacy to increase with age, as people become more knowledgeable, and their attitudes and behaviours change accordingly. However, there are mediating factors that may reduce the financial literacy of the eldest respondents, such as: There is likely to be a cohort effect that impacts on older consumers. Older people, with experience of a very different financial marketplace may find it difficult to keep up with the fast pace of change in the financial market place, including the introduction of new technologies, modified rules and regulations and newly developed financial products. Cognitive deterioration may reduce the extent to which the oldest consumers can retain and apply financial knowledge.

Figure 20 shows that respondents with 3 high scores were most likely to be aged around 30-60. Conversely, it was common for the proportion of people with no high scores to be higher amongst the youngest and oldest respondents, and dip in middle age32. However, this pattern was less apparent in Armenia, Poland and South Africa. In Armenia, the number of people with no high scores rose steeply with age, whilst in Poland and South Africa there was no clear relationship.

32

Note that if there was no difference by age one would expect the same sized bands of colour for each row. Statistical tests confirm that in every country except BVI, average overall scores (not shown) vary significantly by age. 58

United Kingdom

Czech Republic

Armenia

BVI

Figure 20. Financial literacy segments by age

Proportion in each segment, by age (Equivalent to row percentages by country)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

18-19 40-49 70-79 18-19 40-49 70-79 18-19 40-49 70-79 18-19 40-49 70-79 20-29 50-59 80+ 20-29 50-59 80+ 20-29 50-59 80+

Czech German Republi Ireland Hungary Estonia Armenia Albania y c

None

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

18-19 40-49 70-79 20-29 50-59 80+ 20-29 50-59 80+ 20-29 50-59 80+ 20-29 50-59 80+ 20-29 50-59 80+ 20-29 50-59 80+

BVI

UK

South Africa

Poland

Peru

Norway

Malaysi a

None

Base: all respondents. Each bar reports the proportion of people within that age group who fall into each of the financial literacy segments.

59

Income A high level of financial literacy is possible at all income levels. Income itself does not impact on the ability of someone to gain knowledge, to form attitudes conducive to their own financial wellbeing or to exhibit positive behaviours. However, low income is often seen as an explanation for certain behaviours such as borrowing to make ends meet, and used as a reason not to undertake actions such as saving or making long term plans. Furthermore, low income may also be associated with other socio-demographic factors that have been shown to be associated with financial literacy, such as age33. The Core Questionnaire includes a single question to give a general impression of the household income of each respondent. As it is often difficult to get people to divulge income, the question was left quite general in order to maximise responses. This question was fine-tuned by participating countries to ask whether income fell into one of three bands, reflecting income at, below or above national medians. In the results below data is only analysed for the section of the population who responded to this question34. The segment with no high scores includes a high proportion of people who had low incomes (Table 31). Conversely, high income respondents were typically over-represented in the segment with 3 high scores. Overall the pattern indicates that respondents from the higher income households in each country were more than likely to be in the highest scoring segments. This pattern was less marked in Estonia and Norway but particularly noticeable in the Czech Republic, Hungary, Ireland, Malaysia and South Africa. Further analysis by income confirms that in general higher income is associated with higher average scores (Figure 21)35. However, in Armenia and Ireland, middle income consumers were, on average, the most financially literate and in Norway there was very little difference between the middle, and high income consumers. It is quite possible that these differences are driven in whole, or in part, by mediating variables such as age or education level; it is known that income varies by both age and education. However, this does not reduce the importance of the findings. Improving financial literacy amongst the poorest would clearly be welfare enhancing, helping them to make better use of their money, reducing the likelihood that they will make inappropriate decisions or ill-informed choices and potentially helping them to identify ways of increasing their income. It is also worth noting that the poorest consumers have less flexibility to learn by doing, as they cannot afford to make mistakes.

33

It is for this reason that the findings from initial regression analysis are also reported to explore whether income is associated with levels of financial literacy even after controlling for other factors. In the regression analysis that follows, a dummy variable is included to identify those who refused to state their income, to check whether they are systematically different. Additional analysis indicates that the differences in scores by income are significant in every country. 60

34

35

Figure 21. Average overall score by income Low 80 70 60 50 40 30 20 10 0 Albania Hungary Estonia Ireland Poland Czech Republic South Africa Germany United Kingdom Average High income

Base: all respondents with valid data on income. Caution necessary: Just 31 Armenian respondents had high income and 61 respondents in Ireland.

Analysis by income stability (not shown) indicates that there were significant differences in average overall scores by income stability in every participating country except Hungary and Ireland. In all countries the average score was higher amongst those that report that their income is regular and predictable, than those who said their income varied or that sometimes they did not receive it on time, if at all. Education level The Core Questionnaire captures detailed information about each respondents highest level of education. For the purpose of an international comparison, several categories have been combined in order to provide insight into how financial literacy varies according to whether an individual has or has not completed secondary school, or has continued formal education beyond secondary school level. There was a relationship between increased levels of education and high financial literacy scores in every country analysed (Figure 22). In Germany the relationship was particularly strong, and Malaysia and Poland also showed a clear pattern with higher educated individuals more likely to have high financial literacy scores. This initial analysis suggests that general levels of education impact on more than just knowledge. It is particularly striking in Armenia, Poland and Germany that many individuals with low levels of education had no high scores (Figure 22). However, it should also be noted that some people have achieved high scores despite low levels of education, indicating that high levels of financial literacy levels are possible even amongst those who have not completed formal education.

61

Malaysia

Armenia

Norway

Peru

BVI

Figure 22. Financial literacy segments by education

Proportion in each segment, by education (Equivalent to row percentages by country)
0% Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Beyond secondary school Complete secondary school Less than complete secondary school Czech Norway Malaysia Ireland Hungary Germany Estonia Republic Armenia Albania 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

None 1 2 3

Base: all respondents with valid data on education. Caution necessary on small bases for less than complete: Armenia (65) Germany (11) BVI (86).

BVI

UK

South Africa

Poland

Peru

62

Attitude to risk During the initial stages of the questionnaire design, attitude to risk was identified as an important explanatory variable for measures of financial literacy. Figure 23 therefore reports the average scores of those who disagreed (i.e. responded 4, or 5 on the scale) to the risk statement I am prepared to risk some of my own money when saving or making an investment. Only the Czech Republic exhibited a noteworthy difference in scores by this attitude risk averse respondents scored more than those who were risk tolerant. However, statistical tests indicate that the relatively small differences in scores in Armenia and Poland are also statistically significant. Interestingly, it was the risk tolerant who scored slightly higher, on average, in Poland.
Figure 23. Average overall scores by risk aversion Risk loving or neutral Risk averse

68.668.1 65.867.0 65.465.7 66.565.0 66.7 65.163.7 68.569.0 62.963.2 64.364.8 59.3 61.6 60.360.7 58.9 60.161.1 57.958.2 57.0 53.4

United Kingdom

Czech Republic

Malaysia

Armenia

Albania

Hungary

Estonia

Peru

Ireland

Germany

Norway

Poland

Base: all respondents

Multivariate analysis Various characteristics are associated with financial literacy; these characteristics are not observed in isolation and so it is useful to be able to undertake multivariate analysis to gain further insights into the associations. The findings discussed above show important variations by country, indicating that it is appropriate to run a series of regression analyses for each country. The first set of analyses undertaken controlled for characteristics that cannot be chosen or influenced by the participant (age and gender, table not shown). This preliminary check indicated that once the impact of age has been taken into account, men scored significantly higher than women in 9 of the participating countries; there was no country where women scored significantly higher than men. Financial literacy also appears to typically increase with age up to the age of 69 although in Malaysia respondents over the age of 60 had significantly lower scores than those aged 20-29 (after controlling for gender differences). In Armenia, young adults aged 18 -39 had significantly higher levels of financial literacy than other age groups, once gender differences are controlled for.

63

South Africa

BVI

The second set of analyses focused on a wider range of socio-demographic and personal characteristics that may be associated with financial literacy (gender, age, income, education, attitude to risk: Table 8). Gender remained significant in 8 of the survey countries even after controlling for these other factors. Only in South Africa did the other factors appear to outweigh any apparent differences by gender. It is interesting that once other factors such as education and income are controlled for, older participants appear to have been more financially literate in several countries - this indicates that it was not simply the impact of age, but factors associated with it (such as reduced income or lower levels of education) that resulted in lower scores. For example, in Ireland, individuals on a low income had significantly lower scores, as did those with incomplete schooling, and once these were taken into account, scores were significantly higher amongst all adults over 30 than amongst the comparison group. In every participating country, education was significantly related to overall scores, even after controlling for income, age and gender. In every country except the Czech Republic, scores were significantly higher amongst respondents that had continued their education compared with those who stopped when they completed secondary school. Comparing those people who disagreed that they were prepared to risk some of their own money when saving or making an investment (putting themselves at 4 or 5 on a scale from 1 to 5) with others also highlighted differences by country. In Armenia, risk averse people were less likely to be financially literate than their risk tolerant counterparts; whilst in the Czech Republic, Hungary, Norway and South Africa, a risk averse attitude was associated with higher levels of financial literacy.

64

Table 8. Regression results indicating significant variables and direction of association

Gender Albania Male (+) Age u20 (-) 30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) 60-69 (-) 70-79 (-) 80+(-) u20 (-) 30-39 (+) 50-59 (+) 60-69 (+) 30-39 (+) Income High (+) Income stability Stable (+) Education Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Risk averse (-) Risk aversion

Armenia

Male (+)

Low (-)

Stable (+)

Czech Republic

Low (-) High (+)

Risk averse (+)

Estonia

High (+)

Stable (+)

Germany

Male (+)

Hungary

Ireland

Male (+)

30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) 30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) 80+(-) u20 (-) 30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) 70-79 (+) 80+(+) u20 (-) 30-39 (+) 40-49 (+) 50-59 (+) 70-79 (-) 80+(-) u20 (-) 40-49 (+) 50-59 (+) 60-69 (+) 70-79 (+) u20 (-) 30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+)

Low (-) High (+)

Low (-) High (+)

Incomplete schooling (-) Education beyond secondary school (+) Stable (+) Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Education beyond secondary school (+)

Risk averse (+)

Low (-)

Malaysia

Incomplete schooling (-) Education beyond secondary school (+)

Norway

Male (+)

High (+)

Stable (+)

Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Education beyond secondary school (+)

Risk averse (+)

Peru

Low (-) High (+)

Stable (+)

65

Gender Poland Male (+)

Age

Income Low (-)

Income stability

Education Incomplete schooling (-) Education beyond secondary school (+) Incomplete schooling (-) Education beyond secondary school (+)

Risk aversion

South Africa

30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) Male (+) u20 (-) 30-39 (+) 40-49 (+) 50-59 (+) 60-69 (+) 70-79 (+) 80+(+)

Low (-) High (+)

Risk averse (+)

United Kingdom

Low (-) High (+)

Stable (+)

Incomplete schooling (-) Education beyond secondary school (+)

BVI

Male (+)

Education beyond secondary school (+)

The dependent variable is the overall financial literacy score at country level. Explanatory variables considered significant with p values of 0.05 or less. (+) & (-) signs in parentheses signify a positive or negative association. Age comparison is 20-29.

66

CONCLUSION

This chapter shows that it is possible to apply the same set of questions to very different populations around the world and create simple, meaningful indicators of financial literacy. Combining each of the indicators provides a way of capturing overall levels of financial literacy. The findings highlight reasons for concern. It appears that most people had some very basic financial knowledge, but understanding of other, everyday financial concepts such as compound interest and diversification was lacking amongst sizeable proportions of the population in every country. There is also some indication that certain respondents were over-confident, in that they gave incorrect responses rather than admitting that they did not know the answer. Furthermore, some countries need to work particularly hard to ensure that women are not left behind: women had lower levels of financial knowledge than men in almost every country studied. At least three in ten adults in each of the countries surveyed failed to gain a high score on the behaviour measure36. This suggests that certain people could benefit from initiatives designed to change their behaviour. In some countries such an effort would be consistent with consumer attitudes that are largely positive towards longer term plans, whilst in others policymakers would need to take into account the short term preferences of the majority of the population. Analysis of the relationship between behaviour and knowledge suggests a positive association in every country: when knowledge increases so does behaviour. This does not prove causation, and much more research is needed to understand the relationship between these variables. It may be that improved knowledge leads to more active participation in financial markets and more positive behaviours, but it is also possible that people who need to do something seek information, thus increasing their knowledge. Alternatively, knowledge and behaviour may be mutually reinforcing, so that people do something, learn a little, do more and continue to improve their level of knowledge in a continuous cycle. There is also a positive association between attitudes and behaviour. People with a positive attitude towards the longer term were more likely to be exhibiting financial behaviours than those with a preference for the short term. This relationship also warrants further investigation. Analysis by socio-demographics suggests that inequality in opportunities may be preventing individuals from being more financially literate. In particular, low levels of education and income are associated with lower levels of financial literacy, suggesting that certain groups may currently be excluded from activities and learning opportunities that would improve their financial wellbeing. The evidence also gives further emphasis to the need to support women in several of the countries studied37.

36 37

With the marginal exception of BVI (29%). See Chapter 3 for further analysis of financial literacy by gender, and OECD (2013c forthcoming) for further discussion about ways to address gender differences in financial literacy. 67

It is clear that the data collected for this measurement project provides the first ever rich and detailed insight into the financial literacy of diverse populations. Further analysis will help to address questions raised by this initial analysis, such as the relationship between risk aversion and financial literacy, and to further explore areas such as: which aspects of knowledge and behaviour people struggle with most; the issue of confidence in relation to financial knowledge (by analysing the dont know responses); the relationship between knowledge and specific behaviours the relationship between attitudes and specific behaviours the relationship between financial literacy and individuals socio-demographic characteristics, with a particular focus on gender, age and education; the extent to which people are financially included and active consumers; the extent to which the measures of financial literacy can be used to predict outcomes such as having sufficient savings to cover unexpected income shocks.

The results of this analysis provide evidence from which the participating countries can identify needs and gaps and develop appropriate national policies and strategies. The data provides a sound empirical evidence base from which to inform the current revision and elaboration of OECD recommendations on financial education, including the OECD/INFE High-level Principles on National Strategies for Financial Education (2012). It will also inform the development of targeted financial education programmes. It is intended that this first measurement project will lead to a regular programme of data collection, analysis and reporting. The countries participating in the first measurement project will be encouraged to repeat the survey in 3 to 5 years time and for other countries to use the questionnaire to conduct their own survey. By analysing the data in a consistent way, a large dataset of indicators can be developed across countries and over time.

68

Chapter 2

FINANCIAL INCLUSION AND FINANCIAL LITERACY. ANALYSIS OF DATA FROM THE OECD FINANCIAL LITERACY MEASUREMENT EXERCISE IN 14 COUNTRIES

INTRODUCTION

Financial Education, Financial Inclusion and the Need for Data Financial inclusion is recognised as an important policy issue in developed and developing countries, and is part of the G20 agenda on development. The OECD started working on the demand side of financial inclusion in 2003, devoting a chapter of its 2005 publication to the importance of financial education for bringing the unbanked and under-banked into the financial system (OECD, 2005a). In October 2010, upon the recommendation of its Advisory Board, the OECD International Network on Financial Education (INFE) created an Expert Subgroup on the Role of Financial Education in Financial Inclusion. The subgroup agreed on the following working definition: Financial inclusion refers to the process of promoting affordable, timely and adequate access to a range of regulated financial products and services and broadening their use by all segments of society through the implementation of tailored existing and innovative approaches including financial awareness and education with a view to promote financial wellbeing as well as economic and social inclusion. International efforts at measuring financial inclusion so far have mostly focused on supply-side data. Looking at the demand side of financial inclusion is important for several reasons. In particular, it helps to put supply side data into perspective. For example, supply side information about current accounts may tell us that there are 1.9 active accounts for every adult in the country, but the distribution of such accounts across the population can only be explored through consumer data. Furthermore, supply side data can only capture the provision of services; it cannot describe the awareness of such services or the public attitude towards them. There is clearly also benefit in combining information about both sides of the market in order to gain a comprehensive overview of the state of financial inclusion within a country. This work is in its infancy, although it has gained momentum with the creation of the Global Partnership for Financial Inclusions (GPFI) expert subgroup on financial inclusion data and target setting. This chapter reports research undertaken by the INFE to explore the relationship between financial literacy and demand side issues, including awareness, holding and use of financial products. The INFE agreed that the first and foremost consideration when analysing the financial literacy data to inform work on financial inclusion should be to explore patterns at a national level. Whilst tables and figures in this chapter present findings from a wide range of countries, readers are therefore encouraged to consider each of the results in relation to specific countries and their circumstances. The emphasis of the chapter is on the relationship between financial literacy and various indicators of financial inclusion within each country, rather than creating an overall indicator of financial inclusion that can be compared across countries.
71

This chapter provides country level information about important aspects of financial inclusion, including product awareness, holding and recent choices. It proceeds by exploring the extent to which the various indicators of financial inclusion are associated with financial literacy. This is followed by results of analysis of the financial inclusion variables and various key socio-demographic factors including gender, age and income. Conclusions about the extent to which the data is capturing something useful and the lessons that can be drawn from the findings are discussed in the final section, along with suggestions for next steps.

72

PRODUCT AWARENESS

Financial inclusion can only occur if products are available, and if people are aware of the products that are offered. The Core Questionnaire therefore begins the product section by checking whether respondents have heard of the main product types available in their country38. Annex 3 contains country level charts of financial product awareness, holding and recent choice. The products in these charts are sorted by product awareness; the product at the top of the chart is therefore the most widely known product in that country39. Country level charts show that across the participating countries, awareness was typically highest in relation to various bank accounts, such as savings or current accounts. However, in Armenia, more people were aware of retail credit than any other product. Overall, product awareness appears to have been relatively low in Albania, with fewer than 4 in 5 people having heard of a savings account, and slightly over half aware of a current account. In South Africa, too, some types of product awareness appear to have been rather low: whilst 9 in 10 had heard of a bank account, the next most widely recognised product was a credit card, and only 65% of respondents claimed to have heard of this product. The more products people are aware of and the more they can be considered to be engaged with financial services. In this regard, awareness of financial products can be seen as an important indicator of the demand side factors influencing financial inclusion. A simple indicator may just look at awareness of one product used widely around the world, or provide a count of the number of products that people are aware of. Reflecting the belief that people need to be aware of the financial market place more generally in order to identify suitable products, the second approach has been adopted, developing an indicator based on a count of products. However, the number of products differs across countries, and the number of products listed in the questionnaire also varies across countries. For these reasons it would not be appropriate to simply count how many products an individual has heard of; the indicator therefore identifies respondents who know of at least 5. A very small minority of the products listed by individual countries are informal products, such as Stokvels in South Africa. These have been included in the score for awareness to show a general connection with financial services of all kinds. As no country has listed 5 informal products, respondents must have heard of some formal products in order to be considered financially included on this indicator.

38

These products may not be available in all areas of the country. Country specific analysis will be necessary to explore whether lack of awareness is primarily caused by poor coverage. Whilst the questions were asked of everyone, some people did not provide a valid response. The percentages reported here are based on all respondents and will therefore provide a conservative estimate of the actual proportions being aware, holding or making recent purchases of the products listed. 73

39

Figure 24 indicates that in Albania around a third of those responding had heard of at least 5 financial products out of 9 listed. In Germany all respondents (100%) had heard of at least 5 financial products out of the 13 listed in their questionnaire.
Figure 24. Proportion of respondents aware of at least 5 products
88% 94% 84% 100% 87% 99% 95% 96% 98% 81% 62% 34% 97% 90%

Czech Republic

Malaysia

Armenia

Peru

Albania

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents: counting all positive responses. The proportion unaware therefore includes non -response, refusal, don't know and not aware.

Average financial literacy by levels of awareness Figure 25 looks at the average combined financial literacy scores at a country level, across 2 groups. The first group is aware of fewer than 5 of the financial products available in their country (as listed on the questionnaire), whilst the second group is made up of those that are aware of at least 5 such products. Figure 25 shows that individuals who were aware of at least 5 financial products had higher levels of financial literacy than the rest of the population. In some countries the difference in average scores is high: in the Czech Republic for example, there is difference in averages of more than 4 points.

74

South Africa

BVI

UK

Figure 25. Average financial literacy by product awareness Aware of fewer than 5 products
57.9 65.6 45.1 57.9 65.9 46.9 55.3 61.6 58.1 68.1 40.5 67.3 51.5

Aware of 5 or more products

65.5 60.3 64.5 49.8 60.2 52.3 61.6 53.8 64.7 66.6 68.9

Czech Republic

Albania

Malaysia*

Hungary

Estonia

Poland

UK*

Norway

Peru

Base: all respondents. Germany and Ireland had too few respondents who were not aware of 5 financial products to make meaningful comparisons. The UK and Malaysia had fewer than 50 respondents who were not aware of 5 financial products and so comparisons should be made with caution.

The relationship between awareness and financial knowledge Logically, one might assume that people who are financially knowledgeable will also be aware of financial products. Additional analysis of financial product awareness by knowledge has therefore been undertaken to explore whether this is the case. This explores the potential association by looking at the proportion of respondents with low and high levels of financial knowledge40 that were aware of at least 5 products. As seen from Figure 26 below, in some countries higher levels of financial knowledge were associated with product awareness. In Albania and South Africa the relationship was particularly marked: for example, 43% of Albanians with a high level of financial knowledge were aware of at least 5 financial products, compared to just 26% of those with a lower financial knowledge score. However, awareness levels in Germany, Ireland, Peru and BVI appear to have been equally high irrespective of knowledge. It is possible that the apparent lack of a relationship between awareness and knowledge in some countries is driven by country specific policies. For example, in countries where it is impossible to receive an income without some form of bank account, people will be aware of such products even if they know very little about financial matters.

40

As explained in Chapter 1, respondents who answered 3/4 of the knowledge questions correctly are considered to have a high level of financial knowledge. 75

South Africa

Armenia

BVI

Figure 26. Product awareness by financial knowledge Low

93% 88%99% 83% 43% 26%

High financial knowledge score

95% 99% 89% 91% 79% 53%

100% 100% 99% 93% 98%98% 98% 91%98% 91% 90% 90% 100% 75% 74% 72%

Poland

Malaysia

Peru

Czech Republic

Albania

Armenia

Estonia

Hungary

Ireland

Germany

Norway

Base: all respondents

76

South Africa

BVI

UK

PRODUCT HOLDING

Arguably, a measure of product holding is the most important demand side indicator of financial inclusion; and the most relevant for a combined measure of demand and supply side issues. A thorough exploration of the pattern of product holding can provide clear evidence of widespread financial inclusion or, alternatively high levels of exclusion. This can help policy makers to identify the size and nature of the problem of financial exclusion, if there is one. However, a measure of financial product holding cannot be considered to be a fully comprehensive indicator of the success of financial inclusion initiatives. If the primary goal of financial inclusion strategies is to improve access to appropriate products, current use may only provide a weak indicator of the success of strategies and should be used in conjunction with other information. For instance, the indicator may show that current product holding levels are high whilst other evidence such as consumer complaints data indicates that products are falling short of meeting the needs of consumers. Alternatively, current product holding may be found to be very low despite the availability of welldesigned products; in such cases it is worth checking if other barriers exist such as low levels of financial literacy or lack of geographical proximity. Furthermore, consumers may have been incentivised to hold products but make no use of them: for example in countries where income is received through a current account it is important to know whether that account is being used effectively, or whether the money is being withdrawn immediately in order for the account holder to manage a cash budget. The relationship between awareness and holding The country level analyses in Annex 3 show that more people were aware of a product than actually held it, and that in some countries, and with some products, the gap was large. The gap between awareness and holding is particularly noticeable in Armenia (where none of the individual types of products listed was held by more than 16% of the population see Annex 3) and on some products in Peru (where fewer than half had a general purpose account for salary or savings despite almost everybody being aware of such a product). Elsewhere, certain products were held by almost all of those people who were aware of them: such as bank/current accounts in Estonia, Germany and Norway or savings account in Malaysia and BVI. Indicators of product holding Indicators of product holding have been created by classifying products into 4 mutually exclusive categories, either: payment, savings/investments (excluding pension products), insurance or credit in order to give an overview of this aspect of financial inclusion in each of the countries surveyed. The percentage of the population in each country that holds a product from each of these categories has then be calculated (Figure 4)41. These indicators are not necessarily comparable across countries, as each
41

None of the product types listed is counted twice, although we acknowledge that, for example, a current account may also provide credit facilities, and certain savings products may provide various payment options. 77

country chose their own list of products to include in the questionnaire, and each financial market has product variants with different, and often overlapping, features. Figure 27 and the analysis reported below shows that within countries there is often a wide variation in the proportion holding each of the four categories of products42. To the extent to which it is possible to make any country comparison, it appears that across countries there is variation in the proportions of people holding products from within the same category.
Figure 27. Product holding Payment product 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Saving and/or investment Insurance product Credit

Estonia

Albania

Hungary

Ireland

Poland

UK

Norway

Peru

Base: all respondents.

The following sections discuss each of the product categories in more detail. Payment products An indicator has been created that identifies respondents who held products such as current accounts that are primarily designed to receive and make payments. This measure does not count savings products even if they have payment facilities, nor credit products with payment facilities, such as credit cards, as there is incomplete information about this in the data. It should also be noted that none of the countries listed mobile banking facilities in their questionnaire and so this indicator should be seen as a measure of the use of traditional payment products. Access to accounts with payment facilities is widely seen as an important aspect of financial inclusion. Such accounts allow people to automate bill payments whilst also lowering the cost of transactions and reducing the risk created by carrying cash. They almost certainly help with money management by allowing consumers to automate bill payments and may therefore reduce the likelihood of falling into arrears through accidental missed payments. Payment products can also be a prerequisite for other services (financial and other) including access to mobile phones and consumer credit.
42

Some people did not disclose which products they held, and so the percentages should be taken as an underestimate. 78

Czech Republic

South Africa

Germany

Malaysia

Armenia

BVI

The proportion of respondents that held a payment product ranges from over 9 in 10 in Germany, through to less than 1 in 10 in Armenia (see Figure 27 and Annex 3). In 7 of the countries (Estonia, Germany, Hungary, Norway, Poland, South Africa and the UK), more people held a payment product than savings, credit or insurance. Savings and investment products The indicator of savings products incorporates financial products such as savings accounts, credit union accounts, unit trusts and stocks and shares. Using this broad approach to categorisation, savings products were the most common of the four types of financial product analysed in 4 of the countries surveyed (Albania, Czech Republic, Malaysia and BVI) and held by a majority of those populations, with the exception of Albania (31%). They were also held by a majority of the population in Germany, Ireland, Norway and the UK. Insurance In some countries, certain insurance products can be compulsory across particular groups of the population: for example, car owners may have a legal obligation to hold car insurance whilst mortgage holders may be required to hold life insurance or income-protection. Elsewhere, insurance may be seen as a wise investment to protect against potential risks, or it may be treated as an alternative to a savings product (indeed in some countries term life-insurance is often bought as a savings vehicle). This analysis draws together reporting on insurance products as listed by participating countries. Only in Ireland were insurance products the most common products held. However, the majority of respondents had an insurance policy of some kind in Czech Republic, Germany, Hungary, Ireland, Norway, Peru, the UK and BVI. Credit Access to credit is widely considered to be an important part of financial inclusion; allowing individuals to smooth income, deal with unexpected events, invest in future income streams and develop entrepreneurial ideas. Nevertheless, this does not mean that credit is an essential product for all individuals; indeed amongst the countries surveyed, only in Armenia were credit products the most widely held product type. Variations in financial literacy by product holding Average financial literacy by payment product holding As noted above a payment product can help individuals to better manage their money. As financially literate people will exhibit effective money management and timely bill payment, it is logical to assume that payment product holding is likely to be associated with higher levels of financial literacy. The direction of any association is unclear: on the one hand high levels of financial literacy may increase demand for useful products, whilst on the other, access to such products may lead to more effective money management and thus higher financial literacy.

79

As in Figure 25, Figure 28 focuses on the differences in average financial literacy scores across two groups this time those without a payment product, and those with. This shows that payment product holders had an average financial literacy score that was higher than those without such a product.
Figure 28. Average financial literacy by payment product Doesn't hold a payment product
58.5

Holds a payment product

68.6
59.7

66.1

56.1

63.7

67.4
58.1

57.260.9 53.3

68.9

61.3

69.2

68.0
57.6

65.7

57.1

60.9

67.9

52.6

60.9

53.1

71.0 61.1 58.665.6 65.3

South Africa

Malaysia

Albania

Hungary

Estonia

Poland

UK

Ireland

Base: all respondents. 30 respondents in Germany lacked a payment account and so the comparison relies on very small numbers and should be treated with caution; conversely 41 respondents in Armenia had a payment account.

Further exploration of the relationship between timely bill payments and payment products empirically is possible by looking at responses to the statement I pay my bills on time in relation to payment product holding (Figure 29). There was a positive relationship between paying bills on time and having a payment account: in all countries a larger percentage of those with a payment product paid their bills on time than those without such a product. This is particularly noticeable in Albania, Ireland, Malaysia and South Africa. However, even amongst those without such facilities, the majority stated that they also made timely payments showing that there are other factors that influence the propensity to pay bills on time over and above financial inclusion.

Czech Republic

Germany

Armenia

80

Norway

Peru

BVI

Figure 29. Paying bills on time with and without a payment account Has no payment product
90% 94% 95% 73% 79% 88% 79% 83% 92% 96% 80% 83% 73% 88% 55%

Has a payment product

80% 81% 73% 77% 90% 73% 80% 50% 88% 89% 68% 81% 85%

Czech Republic

Albania

Hungary

Estonia

Ireland

Poland

UK
UK

Armenia*

Germany*

Malaysia

Norway

Peru

Base: all respondents: Percentage of respondents responding to the behaviour statement I pay my bills on time by putti ng themselves at 4 or 5 on the scale from Never=1 to Always=5. *30 respondents in Germany lacked a payment account and so the comparison relies on very small numbers and should be treated with caution; conversely 41 respondents in Armenia had a payment account.

Average financial literacy by savings product holding The financial literacy measure considers a wide range of financial behaviours. Amongst these is active saving. Although the active saving measure is not related specifically to account holding, it is likely that there is some relationship between financial literacy and savings account holding. Once again the focus is on the differences in average financial literacy scores across two groups this time those without a saving product, and those with. Respondents with a savings or investment product were typically more financially literate than those without (Figure 30). However, the size of the difference is too large to be explained by the active savings measure alone. This indicates that savings account holding is associated with higher levels of financial literacy.
Figure 30. Average financial literacy by saving and investment Doesn't hold a saving or investment product
67.3 61.7 57.4 56.2 65.8 47.3 75.4 69.0 65.0 67.4 64.9 59.4 59.0 59.0

Holds a saving or investment product

67.1 66.9 64.168.3 66.9 66.968.9 65.5 56.761.0 57.3 56.4 56.0 54.3

South Africa

Czech Republic

Malaysia

Armenia

Albania

Peru

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents. Note that in Armenia 16 respondents had a savings or investment product.

81

South Africa

BVI

BVI

In order to remove the potential influence of the active savings measure, additional analysis considers whether there is evidence of a link between financial knowledge and savings. Figure 31 below shows whether people who understood the impact of compound interest on savings were more likely to hold money in savings or investments than those who didnt. There was a positive association between this aspect of knowledge and savings in most countries, although in South Africa it appears that it was not knowledge but other factors such as income that mostly impacted on the likelihood of using a saving or investment product. The chart also highlights the fact that even people without a clear understanding of the benefit of compounding were using saving or investment products. Such people may not have chosen the most appropriate product for their needs.
Figure 31. Holds savings or investment account by knowledge of compound interest Incorrect
98% 92% 97% 90% 83% 76%

Correct
86% 76% 95% 83% 87% 69% 34% 31% 31% 20%

97% 90%

51% 29% 1% 1% 25% 20% 20% 14%

8%10%

Malaysia

Germany

Czech Republic

Peru

Albania

Armenia

Estonia

Hungary

Ireland

Norway

Poland

Base: all respondents. Left hand column per country is the percentage of all respondents who gave an incorrect response to the compound interest question that hold a savings product. Right hand column shows the percentage of all respondents who gave a correct response that hold a savings product. Note that in Armenia 16 respondents had a savings or investment product.

Average financial literacy by credit product holding The difference in financial literacy amongst those with and without a credit product is typically smaller than for those products discussed above (Figure 32). However, there are some significant differences (see for example Ireland, Malaysia and South Africa), and the association is consistently positive: those with credit products were more financially literature than those without. Of course, given the potential pitfalls from holding credit it is reassuring that credit users were not less financially literate than those without credit.

82

South Africa

BVI

UK

Figure 32. Average financial literacy by credit product Doesn't hold a credit product Holds a credit product

72.0 69.8 66.068.2 67.9 66.8 67.069.3 66.4 66.2 61.668.8 65.6 63.467.8 58.662.2 64.2 62.2 59.9 59.7 59.1 59.0 58.6 58.0 56.1 55.7 54.7

Czech Republic

Malaysia

Peru

Albania

Armenia

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents.

Average financial literacy by insurance product holding In some countries some insurance products (such as car insurance or household content insurance) will be mandatory for those individuals concerned (i.e. car owners or those living in rented accommodation). With a strong external influence on the propensity to hold insurance there may be little or no difference in the level of financial literacy amongst holders and non-holders. However, in fact the analysis shows that average levels of financial literacy are considerably higher amongst those holding insurance in several of the countries surveyed (Czech Republic, Germany, Ireland, Malaysia, South Africa, UK; Figure 33).
Figure 33. Average financial literacy by insurance product Doesn't hold an insurance product
60.4 61.2 70.4 58.2 58.1 63.4 60.3 69.6 63.2 69.8 55.1 67.5 72.7 61.4 59.7

Holds an insurance product

66.7 62.0 66.5 55.5 63.0 65.4 55.5 57.6 66.7 67.7 69.0

South Africa

Czech Republic

Malaysia

Albania

Peru

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents. The proportion with insurance in Armenia is too small to make a meaningful comparison

83

South Africa

BVI

UK

BVI

UK

ACTIVE PRODUCT CHOICE

Financial inclusion requires active participation in the financial market place. In most countries people could benefit from keeping an eye on the market, switching providers and choosing new products that meet their evolving needs. The Core Questionnaire therefore asks respondents whether, in the last 2 years, they have actively chosen any of the products listed even if they no longer hold these products. They are asked to exclude simple (passive) renewals. Figure 34 reports the proportion of respondents in each country that said they had made such a product choice. This is a useful measure of inclusion, but as with the previous measures, it will underestimate the true proportion of active consumers, since some people did not answer the question and some products may not have been listed43. Over half the population of Albania, Germany, Hungary, Ireland, Peru, South Africa, the UK and the British Virgin Islands had chosen some kind of financial product in the last 2 years.
Figure 34. Proportion making an active product choice in the last 2 years

62% 39% 39% 48%

70%

64%

60% 46%

67% 49% 56%

57%

60%

Albania

Hungary

Estonia

Ireland

Poland

UK

Norway*

Germany

Peru

Base: all respondents. *Norway did not ask this question.

The country level figures in Annex 3 show that, in Albania and Armenia, almost as many people had made a recent purchase of the products listed as reported that they held these products (although it should be remembered that overall percentages are relatively low). In all countries surveyed, average financial literacy was higher amongst those who had made a recent product choice than those who hadnt (Figure 35). In some cases the difference in averages is large.

43

Countries in the first measurement project were not required to separate out the refusals from the no responses, and so we cannot consistently identify non-response on this question. It should also be remembered that the list of products for this question was decided at the country level. 84

Czech Republic

South Africa

Malaysia

Armenia

BVI

Figure 35. Average financial literacy by recent product choice Hasn't made a recent product choice
71.4 71.0 70.2 68.7 67.7 64.7 63.2 62.4 62.3 61.9 61.2 61.1 59.0 58.7 54.7 53.7

Has made a recent choice

67.8 67.6 67.069.7 61.6 60.5 60.1 57.7 55.1 54.9

Poland

Norway*

Malaysia

Peru

Czech Republic

Armenia

Albania

Estonia

Hungary

Germany

Ireland

Base: all respondents. *Norway did not ask this question.

85

South Africa

BVI

UK

RELYING ON FRIENDS AND FAMILY

The previous sections focused on product awareness, choice and usage. The indicators capture the level of inclusion within a country. However, the data can potentially also be used to develop an indicator of exclusion. This section does this by identifying people who turned to friends and family for a loan when they are unable to make ends meet or ask them to hold money on their behalf when they want to save. This may provide a useful indication of people who did not have access to formal products that they felt met their needs, or alternatively of people who had less need for formal products because extended family networks provided valuable support the most likely explanation will depend on the country context. The full meaning of this indicator at a national level must be considered with care. In some countries, and amongst some groups, family and friends may be the first option for people to access financial support, and this behaviour would not be taken as a sign of lack of access to financial services. It would, however, still have policy implications: over reliance on family and friends can put financial pressure on households and informal networks and is of limited economic benefit. Figure 36 shows that a relatively large proportion of adults in Armenia (37%) and Malaysia (35%) had turned to family or friends to keep their savings or help them to make ends meet. In other countries the behaviour was less common.
Figure 36. Relying on friends or family for borrowing and/or saving

27%

37% 9% 17% 12% 10% 10%

35% 5%

27%

15%

27% 12%

18%

Czech Republic

Malaysia

Armenia

Peru

Albania

Estonia

Hungary

Ireland

Poland

Base: all respondents. Counting all those who reported that they borrowed from friends and family to make ends meet plus all those who gave money to friends and family to save on their behalf.

There is also a difference in average financial literacy scores by whether or not the respondent has turned to family and friends to borrow and/or save in most of the countries surveyed44. Financial literacy is consistently lower amongst individuals who have done so.

44

Some of the difference can be explained by considering the development of the score, but not all of it. Those who had not borrowed to make ends meet would exhibit a positive behaviour, as would those who had given money to friends to look after as a savings strategy. 86

Germany

Norway

South Africa

BVI

UK

Figure 37. Average financial literacy by reliance on family and friends Has not turned to family or friends Borrowed or saved via family or friends

69.066.9 68.9 66.665.0 65.4 66.2 65.162.5 65.0 64.3 67.361.4 66.8 62.9 61.9 59.9 59.5 59.4 59.3 56.5 54.5 54.4 54.1 54.1 51.3 51.1 51.0

Albania

Malaysia

Peru

Czech Republic

Armenia

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents.

87

South Africa

BVI

UK

FINANCIAL INCLUSION AND SOCIO-DEMOGRAPHICS

The analysis presented so far shows clearly that financial exclusion is associated with lower levels of financial literacy. This indicates a potential gain from developing financial education programmes designed to raise the financial literacy of the financially excluded. In order to identify who would benefit most from such financial education, this section explores the relationship between financial inclusion and various socio-demographic factors. There is no additional analysis of the exclusion indicator by sociodemographics as the small proportions make it difficult to split the categories further. Variations in awareness by socio-demographics Additional analysis shows that similar proportions of men and women were aware of at least 5 of the products listed in most of the countries analysed (Table 9). In Albania (29% vs 37%), Armenia (86% vs 90%), Malaysia (93% vs 98%), and South Africa (60% vs 64%), women had slightly lower levels of awareness than their male counterparts. Product awareness varied somewhat by age, typically being highest in young to middle age and falling off in old age across the countries surveyed. This contrasts somewhat with financial literacy, which is typically highest in middle age. With the exception of Germany and Ireland, levels of awareness increased notably with income. In South Africa, this finding is particularly stark, with just 50% of respondents from low income households aware of at least 5 products, compared with 83% of those from the higher income group. Awareness also varied by education in every country but Germany (awareness appears to be practically universal amongst German adults). Adults who had continued their education beyond secondary school were the most likely to be aware of 5 or more products.

88

Table 9. Awareness by socio-demographics

Gender Age 50 and over Income Education level Less than complete secondary*** Complete secondary Beyond secondary 58% 93% 97% 88% 100% 93% 99% 99% 98% 98% 86% 82% 98% 94%

18 to 29*

Average

30 to 49

Female

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa UK BVI

29% 86% 93% 82% 100% 87% 99% 93% 97% 98% 80% 59% 97% 92%

37% 90% 95% 86% 100% 86% 99% 98% 94% 98% 83% 64% 97% 88%

40% 91% 95% 85% 100% 91% 100% 96% 91% 97% 87% 63% 98% 88%

37% 92% 97% 90% 100% 90% 99% 98% 98% 99% 85% 65% 98% 90%

29% 79% 92% 76% 100% 81% 99% 88% 96% 97% 74% 55% 96% 91%

19% 81% 92% 79% 99% 78% 99% 93% 92% 97% 74% 50% 95% 86%

36% 93% 96% 86% 100% 91% 99% 96% 98% 99% 88% 76% 98% 94%

47% 100% 99% 89% 100% 96% 98% 99% 99% 99% 92% 83% 99% 95%

High**

Male

Low

20% 54% 92% 72% 100% 77% 98% 88% 94% 97% 64% 52% 87% 68%

38% 83% 96% 84% 100% 91% 100% 97% 96% 99% 84% 71% 98% 87%

Table suppresses data for refusals on age, income and education level respectively. *66 respondents from BVI fall into the youngest category. ** For Armenia the number of respondents with a high income is just 31 and for Ireland 61 respondents reported a high income. ***Few respondents had less than complete secondary school education in Armenia (65) Germany (11) and BVI (86). All analysis based on small numbers of respondents should be read with caution.

Variations in payment product holding by socio-demographics Analysis of payment product holding by socio-demographics indicates that women were noticeably less likely than men to hold a payment product in Albania (22% vs 29%), Malaysia (74% vs 82%), Peru (45% vs 53%) and BVI (54% vs 65%) (Table 10). In Norway, women were slightly more likely than men to have an account (90% vs 86%). Elsewhere the difference in payment product holding by gender was very small. Payment products were less likely to be held by young adults and very old respondents than those in middle age across the countries surveyed. Those with lower levels of education were also less likely than the more educated to hold payment products indeed a gap of 20 percentage points is not uncommon, as seen in Table 10.

89

Table 10. Payment products by gender, age and education level

Gender 18 to 29 30 to 49 Female Male Age 50 and over Less than complete secondary Education level Complete secondary Beyond secondary 52% 26% 86% 93% 99% 89% 91% 96% 92% 74% 84% 75% 91% 71%

Albania

22%

28%

25%

27%

25%

9%

21%

Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa UK BVI

2% 70% 90% 96% 68% 76% 74% 90% 45% 67% 59% 83% 54%

3% 73% 90% 96% 71% 78% 81% 86% 53% 69% 65% 84% 65%

3% 70% 89% 91% 67% 70% 89% 88% 38% 74% 59% 83% 39%

3% 84% 93% 97% 86% 84% 82% 93% 55% 79% 68% 86% 57%

2% 61% 88% 97% 57% 74% 53% 84% 60% 56% 58% 81% 65%

1% 61% 88% 93% 58% 72% 64% 87% 36% 58% 53% 79% 44%

3% 80% 90% 96% 71% 94% 84% 92% 57% 76% 74% 86% 68%

Table suppresses data for refusals on age and education level. *Very few people in Germany are lacking a payment account and so the comparison relies on very small numbers and should be treated with caution; conversely very few people in Armenia have a payment account.

Arguably the most noteworthy relationship between product holding and the socio-demographics studied is with income (Figure 38). Figure 38 shows the percentage of low, average and high income respondents that held a payment product in each of the participating countries. There was a clear association between level of income and payment product holding in every country; the relationship was particularly marked in Albania, Armenia, Hungary, Malaysia and Peru.

90

Figure 38. Payment product holding by income Low 100% 80% 60% 40% 20% 0% Albania Hungary Estonia Ireland Poland UK Norway Peru Czech Republic South Africa Germany Malaysia Armenia BVI Average High Income

Base: all respondents who provided detail of their income. Low, middle and high income refer to income relative to national median values. *Very few people in Germany lacked a payment account and so the comparison relies on very small numbers and should be treated with caution; conversely very few people in Armenia have a payment account.

Variations in savings product holding by socio-demographics There was no strong, consistent pattern between savings products and gender across the countries studied (Table 11). In Estonia (24% vs 19%) and Norway (85% vs 78%) women were slightly more likely than men to have a savings product. Elsewhere gender differences were very small, with the exception of Poland, where just 18% of women and 30% of men had a savings or investment product, and to a lesser extent the UK (71% vs 77%). Savings product holding appears to have been highest amongst individuals aged 30 to 49, and was strongly associated with income (Table 11). The relationship between saving holding and income was also positive. In two countries, however (South Africa and BVI), more middle income individuals held saving and investment products than either those on high or low incomes.

91

Table 11. Savings holding by gender, age and income

Gender 18 to 29 30 to 49 Female Male Age 50 and over Low Income Average High 53% 11% 97% 33% 98% 33% 89% 98% 86% 18% 41% 30% 87% 90%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway* Peru Poland South Africa UK BVI

26% 1% 94% 24% 92% 17% 79% 90% 85% 8% 18% 31% 71% 88%

35% 1% 94% 19% 94% 16% 78% 93% 78% 9% 30% 31% 77% 91%

32% 1% 94% 16% 87% 6% 69% 95% 83% 5% 23% 28% 69% 91%

28% 1% 97% 25% 95% 18% 82% 92% 81% 10% 29% 34% 74% 88%

34% 1% 92% 24% 94% 21% 80% 86% 80% 12% 20% 32% 76% 89%

21% 0% 92% 13% 87% 8% 76% 87% 77% 4% 15% 28% 62% 90%

23% 2% 96% 24% 95% 16% 85% 95% 84% 10% 28% 36% 76% 91%

Table suppresses data for refusals on age and income respectively.

There was a positive association between savings and investment products and education in most countries, which was particularly marked in Albania, Germany, Hungary, Poland and UK (Figure 39). However, in the Czech Republic there was relatively little variation in saving and investment product holding by education and in Peru and BVI there was no clear relationship.

92

Figure 39. Holds savings or investment account by education

Less than complete 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Complete secondary school

Beyond secondary school

Peru

Malaysia

Czech Republic

Armenia

Albania

Hungary

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents that provided information about their level of education. Graph shows proportions holding savings products by education. Note that in Armenia 16 respondents had a savings or investment product.

Variations in credit and insurance product holding by socio-demographics No clear relationship could be found between the use of credit products and gender (Figure 40). In some countries, approximately the same proportions of men and women reported holding some kind of credit product, whilst in others women were rather less likely to hold them than their male counterparts. In Estonia slightly more women than men held a credit product.
Figure 40. Credit products by gender Female
72% 76% 48% 49% 28% 16% 32% 29% 55% 52% 35% 33% 71% 72% 59% 43% 36% 41% 31% 39% 19% 24%

Male
81% 84% 69% 73% 72% 73%

Armenia

South Africa

Czech Republic

Malaysia

Hungary

Peru

Albania

Estonia

Ireland

Germany

Norway

Poland

Base: all respondents. Graph shows proportion of men and women holding credit products.

Women were slightly less likely than men to have insurance, but only Malaysia shows a marked difference with 34% of women reporting that they held an insurance product compared with 48% of men.

93

South Africa

BVI

UK

BVI

UK

As found with the other products, the use of both credit and insurance products varied by age and was highest amongst 30 to 49 year olds (not shown). There was a strong positive association between both credit use and income and insurance and income credit holding and the use of insurance products were highest amongst those with the highest incomes. The relationship with education was also positive. Variations in active product choice by socio-demographics In most of the countries surveyed there was very little difference between the proportions of men and women who had made a recent product choice (Figure 41). In Albania, however, 53% of women had made a recent choice out of the products listed compared with 68% of men, and in Peru 61% of women but 71% of men had made such a choice. Nowhere were women more likely to be active financial consumers than men.
Figure 41. Product choice by gender Female
68% 53% 47%49% 38%41% 37%41% 67%72% 62%66% 59%60%

Male
71% 46%51%

61% 47%44%

55%57% 54%60% 60%61%

Peru

Czech Republic

Norway*

Malaysia

Armenia

Albania

Hungary

Estonia

Germany

Ireland

Poland

Base: all respondents. *Norway did not ask this question. Graph shows proportion of men and women that made a recent product choice.

There was a similar relationship between product choice and age as with product holding those in the 30 to 49 year old age bracket were the most likely to be active (Table 12). A larger proportion of high income respondents had made a recent choice than those in the lower income brackets, and the higher educated were more likely to have made a choice than those who had not completed secondary school.

94

South Africa

BVI

UK

Table 12. Active product choice by age, income and education level
Age 50 and over 18 to 29 30 to 49 Income Less than complete secondary Average Education level Beyond Secondary 90% 42% 42% 49% 76% 72% 65% 65% 69% 60% 64% 62% 60% Complete secondary 67% 35% 45% 48% 66% 66% 60% 47% 60% 50% 61% 53% 58%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway* Peru Poland South Africa UK BVI

55% 46% 56% 52% 71% 66% 62% 56% 55% 66% 55% 64% 77%

59% 44% 46% 49% 78% 70% 67% 50% 75% 56% 60% 61% 53%

66% 27% 26% 41% 64% 58% 49% 20% 69% 33% 52% 50% 64%

45% 32% 32% 48% 51% 63% 56% 33% 63% 42% 51% 46% 61%

58% 44% 45% 45% 75% 64% 66% 50% 71% 52% 64% 60% 60%

87% 71% 55% 52% 81% 76% 67% 62% 77% 66% 60% 66% 58%

High

Low

46% 26% 35% 41% 31% 56% 49% 23% 57% 28% 52% 51% 59%

Analysis suppresses missing data. *Norway did not ask this question

95

CONCLUSION

The OECD/INFE financial literacy measurement survey has created a unique opportunity to explore some of the demand side factors of financial inclusion, including product awareness and holding. This analysis has provided insights into various aspects of financial inclusion in particular countries from a demand-side perspective. While this may not offer a full picture of financial inclusion, it provides an important complement to existing international financial inclusion measurements, which focus mostly on supply-side data from financial institutions. The approach also has the advantage of being applicable internationally, and already provides data from a range of developed and developing countries. The analysis undertaken in this chapter provides indications that the population of some countries have low levels of financial inclusion on a number of indicators. The positive association between financial inclusion and financial literacy suggests an important role for financial education in the drive to increase levels of inclusion in some countries. Moreover, the various indicators of financial inclusion are significantly associated with various socio-demographic factors such as gender, age, education, and household income; suggesting that targeted financial education may reap the highest rewards. However, financial education does not work in isolation. Financial services must be accessible and financial products must be well designed to provide users with safe, effective solutions to their financial needs. Consumer protection must be in place to protect consumers from unfair practices and regulation must be fit for purpose. With this structure in place, education can focus on informing consumers about their choices, rights and responsibilities, and providing them with the skills necessary to undertake those choices and manage their products appropriately. Product awareness The analysis presented here shows that basic awareness of the existence of products is high in many countries; even if product-use lags behind. Countries with high levels of awareness may wish to know more about the extent to which consumers are aware of the features of products, their benefits and costs, and how they can be accessed. There is a clear association between product awareness and financial literacy. Countries with low levels of product awareness may therefore wish to develop awareness campaigns and financial education strategies to improve both financial inclusion and financial education. Awareness was lower amongst Albanians and South Africans with low levels of financial knowledge than those with higher levels of knowledge. This indicates that there may be considerable benefit from providing financial education that increases both financial product awareness and financial knowledge in these countries. The fact that the least aware adults were typically older, less well educated and poorer, suggests that campaigns targeted on this segment of the population may be the most effective ways of increasing overall awareness. Effectiveness of any awareness campaign will depend on presenting information in a way that is engaging and easily understood by adults with little or no formal schooling. Repeated
96

messages are often seen as the most appropriate way of achieving awareness. Awareness campaigns also need to be immediately relevant to the target audience, suggesting that targeting different groups with different messages may be necessary. Countries wishing to develop this analysis further may consider certain refinements. For example, it would be possible to undertake country level analysis of awareness of specific categories of products (such as regulated/unregulated; products available online, or in branch, products available to under 18s). Alternatively, countries may experiment with the point at which consumers are considered to be sufficiently aware; this analysis focuses on 5 product types, but there is no precedent for this and there may be additional benefit from exploring other relationships such as how financial literacy changes with the number of product types the respondent is aware of. It may also be beneficial to look at lack of awareness, and focus on the types of products least likely to be in the public eye. Equally, countries may wish to undertake more in-depth analysis of the relationship between product awareness and financial knowledge. It may be that the cut-off used to identify a high level of financial literacy (a score of at least 6 out of 8), is too high to identify the point at which people become aware of financial products in some countries. In countries such as Germany and Ireland, there is little benefit from exploring the relationship between basic product awareness and other factors, since awareness appears to be almost universal. In these countries research may be more fruitful if it focuses on the other aspects of awareness such as whether individuals know the purpose of the products. There are many possible reasons for the gap between awareness and product holding that need exploring through more in-depth data collection. It may be that there are barriers to accessing the products that people are aware of, including rigid identification requirements, cost, or poor product design; it may even be that people do not believe that they can benefit from the products currently on the market. Additional analysis at the country level could also take into account factors such as the maturity of the financial market, the efficiency of regulation and consumer protection and the extent to which consumers can access impartial information and advice. Products held Of the four product types analysed, payment products were the most commonly held in 7 of the countries. This may reflect pressure or preference to make and receive payments electronically, but may also indicate that a current account is seen as a gateway to other products in some countries. The data also suggests that the majority of adults lack a basic payment account in Albania, Armenia, and Peru. National campaigns to explain the benefit of such accounts may be necessary to increase their use. Incentive schemes and/or redesigned products that encourage account opening may also have some initial impact, but experience elsewhere suggests that consumers are likely to continue to largely operate in cash without increased understanding of the benefits of electronic transactions and confidence with the necessary technology. The analysis shows that use of a payment account is associated with timely bill payments. To the extent that this reflects the ease with which bill payments can be automated, it suggests that education aimed to reduce the number of unbanked households should stress this benefit of payment products. However, countries should explore the association further, as it may be that people with poor payment histories have been refused an account, or cannot find one that meets their needs.
97

The analysis of payment product holding by socio-demographics suggests that in Albania, Malaysia, Peru and BVI financial education aimed at increasing the use of such products may be more effective if it includes provision tailored to the needs of women. The data suggests that information about payment accounts is an important component of financial education targeted on adults in low income households. This information should stress the overall benefit of using an account, and help individuals to develop their skills to manage the account effectively and thus minimise the cost of transactions and reduce the likelihood of additional fees. The use of savings and investment products is relatively low in Albania, Armenia, Estonia, Hungary, Peru, Poland and South Africa. With the exception of South Africa, there is some association between the use of savings or investment products and knowledge of the impact of compound interest; suggesting that account holding could be further encouraged by financial education that explains the longer term benefits of formal saving. In South Africa it appears that it is not knowledge but other factors such as income that mostly impact on the likelihood of using a saving or investment product. Whilst there is an association between holding saving and investment products and understanding the impact of compound interest, the analysis also shows that many people held savings despite their lack of understanding the benefit of compound interest. Of course, compound interest is not the only reason to save, but as an indicator of their level of financial knowledge in relation to savings it suggests that something other than knowledge led to their desire to put money aside. Additional analysis would be useful to explore the types of products that more knowledgeable people choose, and to consider how to encourage the least knowledgeable to open savings accounts. It is also important to provide additional knowledge to existing savers to make sure that they are making the most of the saving and investment opportunities available to them. Qualitative data may be necessary to understand the relationship between knowledge and the use of savings products in more detail. The analysis of the various types of product holding by socio-demographics shows that there is some association with financial inclusion and gender although the pattern is not marked on some product indicators- such as credit and insurance. This may reflect the propensity of households to hold such products as a couple: for example, a home loan or insurance on a car may be held in joint names. The extent to which credit and insurance products are being underutilised is difficult to assess through the OECD/INFE Financial Literacy Survey. However, it is clear that once again, the segment least likely to use these products is older, less well educated and poor suggesting that both types of product could be discussed in financial education campaigns aimed at this group. Across the products analysed there is also an association with education level. This may be related to specific knowledge gained through education, but is perhaps more likely to reflect the association between education and income, and in turn the association with income and the various products. The relationship between income and payment product holding needs to be analysed at the country level. It may be that financial institutions are excluding low income individuals in a variety of ways (such as keeping fees excessively high, marketing products in a way that suggests that are for a different type of clientele, opening branches exclusively in high income neighbourhoods or setting minimum income requirements on certain products). However, it is also possible that consumers have the impression that there is nothing to gain from managing their money through a payment account: an attitude that could potentially be countered through financial education.

98

Product choice At least 1/3 of the population in each of the countries studied had made an active product choice. However there is potential to improve engagement still further, with a large group of individuals remaining passive or excluded. It is particularly noteworthy that the proportion of people who had made a recent purchase of general insurance was much lower than the proportion holding insurance products. As many insurance policies can be renewed on an annual basis, this lack of recent product choice suggests that there were a lot of passive consumers in the countries surveyed who let their providers renew their product automatically. Reliance on friends and family It is difficult to be certain that there is unmet demand for financial products. One way in which this chapter has attempted to identify this is through an indicator that captures the extent to which people turn to friends and family. If people are relying on friends to keep their money safe or to lend them money in times of need, this may suggest that they do not have access to appropriate financial products such as safe savings accounts or payment products with overdrafts. However, authorities should decide for themselves whether this is an acceptable indicator or a crude measure of unmet demand amongst their own population, particularly taking into account cultural factors. In most of the countries studied, financial literacy was lower amongst those who had turned to family and friends to borrow money or to safeguard savings than those who hadnt. As those who are aware of financial products are more financially literate than those with lower levels of awareness it might tentatively be concluded from the various findings that people turn to friends and family because their low level of financial literacy makes them less aware of the alternatives. This could be further explored in additional research.

99

Chapter 3

GENDER DIFFERENCES IN FINANCIAL LITERACY45

45

See also OECD (2013c forthcoming). 101

INTRODUCTION

The policy interest in the financial literacy of women and girls Both women and men need to be sufficiently financially literate to effectively participate in economic activities and to take appropriate financial decisions for themselves and their families. However, as women tend to live longer and earn less than men, therefore being more likely to face financial hardship in old age, it is particularly important that they have the financial skills necessary to ensure their financial wellbeing. Unfortunately, existing evidence indicates that their level of financial knowledge tends to be lower than mens, suggesting that they may not be well prepared for the challenges they face. The need to address the financial literacy of women and girls as a way to improve their financial empowerment is gaining global relevance and is reflected in various initiatives at a national and international level. With the aims of improving policies and promoting gender equality in the economy in both member and non-member countries, in 2010 the OECD launched a Gender Initiative to examine existing barriers to gender equality in education, employment and entrepreneurship. The OECD Horizontal Project in Gender Equality presented its Final Report on Gender Equality in Education, Employment and Entrepreneurship to the Ministerial Council Meeting in May 2012 to inform, share policy experiences and good practices, and help governments promote gender equality (OECD, 2012). The report also addresses gender differences in financial literacy and how financial education can contribute to womens financial empowerment. Moreover, in June 2012 the G20 Leaders meeting in Mexico recognised the need for women to gain access to financial services and financial education. G20 Leaders also asked the OECD/INFE together with the Global Partnership for Financial Inclusion (GPFI) and the World Bank to identify the barriers they may face, and called for a progress report to be delivered by the next Summit (G20, 2012). This chapter supports the work of the OECD/INFE in helping policy makers and relevant stakeholders by identify key gender differences in levels of financial literacy. It draws on the OECD/INFE financial literacy survey and other relevant data to explore differences in knowledge, behaviour, attitudes and also looks at issues around product choice and holding46.

46

See OECD/INFE (2013c) and INFE (2013). 103

FINANCIAL KNOWLEDGE OF WOMEN

Analysis of financial knowledge by gender shows that women have lower average financial knowledge scores than men in 13 of the 14 countries that participated in the OECD Financial Literacy Measurement survey the exception being Hungary (Figure 42). This result is consistent with the overall international evidence on gender differences in financial knowledge and understanding (Hung et al., 2012). Gender differences in financial knowledge are found also by Lusardi and Mitchell (2011), who report cross-country comparable evidence from studies on eight countries Germany, Italy, Japan, the Netherlands, New Zealand, the Russian Federation, Sweden and the United States. Using the same datasets exploited in Lusardi and Mitchell (2011), Bucher-Koenen et al. (2012) further explore gender differences in financial knowledge in a subset of countries (the United States, the Netherlands, and Germany). The overall evidence from these studies shows that in most countries women have lower levels of financial knowledge than men based on short tests of financial knowledge. However, in Russia and East Germany financial knowledge is very low for both men and women, and gender differences are not significant (Bucher-Koenen and Lusardi, 2011, and Klapper and Panos, 2011). Other national financial literacy surveys have documented gender differences. In Azerbaijan and Bulgaria, women achieved a lower share of correct answers than men to six numeracy and financial knowledge questions (Alpha Research, 2010; Azerbaijan Micro-finance Association, 2009). The ANZ Survey of Adult Financial Literacy in Australia provides evidence that females had lower scores than males on a financial knowledge and numeracy index (ANZ, 2011). Similarly, the 2012 Survey on the Financial Literacy of the Portuguese population shows that men obtained higher scores on average than women in the knowledge of basic financial concepts (Banco de Portugal, 2011). In New Zealand, women made up a higher proportion of the low knowledge group, and a lower proportion of the high knowledge group, even though they improved over time (ANZRetirement Commission, 2009). Also women in the United Kingdom scored lower on a financial knowledge test compared to men (Atkinson et al., 2006).

104

Figure 42. Average financial knowledge score by gender

61.8 54.9

65.9 61.9

74.7 72.3 68.9 67.5

77.076.6 74.3 75.9 66.8 66.6 68.464.1 66.0

Male

Female
75.5 62.6 71.4 67.0

70.7 60.5 58.2 60.3 56.0 54.4 52.7

Armenia *

Albania *

Ireland *

Norway *

Estonia *

Czech Republic *

Peru *

UK *

Germany *

Malaysia *

Note: In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Gender differences in financial knowledge at young ages Gender differences in financial knowledge may to some extent be the result of historical gender inequalities. If this is the case, it is possible that gender differences in financial knowledge are smaller among younger generations exposed to a more egalitarian environment. However, analysis of financial knowledge amongst respondents aged 18 to 29 identifies gender differences among young generations, even though differences are significant for fewer countries (Armenia, Ireland, Norway, Peru, Poland and the UK Figure 43). Bucher-Koenen et al. (2012) show that in Germany, the Netherlands and the US gender differences in financial knowledge exist within all age brackets, including among respondents younger than 35. A survey of financial understanding of 8 to 18-year-olds in the Netherlands also finds that more boys than girls can answer knowledge questions correctly (CentiQ, 2008). In a study among college students in the US, Chen and Volpe (2002) find that young women have less knowledge about personal finance topics than men. Also Lusardi et al. (2010) find gender differences in financial knowledge among US young adults. The evidence suggests that more research is necessary to understand whether there are factors inducing gender differences in financial knowledge from a very young age, and to what extent these gender differences are transmitted across generations.

105

South Africa *

Hungary

Poland *

BVI *

Figure 43. Average financial knowledge score by gender (young people aged 18-29)

63.063.2

70.9 68.1 66.5 62.9

74.572.3 60.759.2

77.878.8

Male
67.968.2 67.7 59.7 58.2 45.1

Female
75.0 66.0 58.3 52.1 69.0 59.459.9 57.7

Armenia *

Ireland *

Norway *

Peru *

Czech Republic

Germany

Malaysia

Hungary

Poland *

Albania

Estonia

Subsample of respondents aged 18-29. The data for the BVI have been excluded due to a very low number of observations in the relevant age range. Note: In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Less well-educated and low-income women have the lowest financial knowledge In addition to looking at the differences in financial knowledge between women and men, it is also interesting to focus on women only by investigating which subgroups of women exhibit the lowest financial knowledge. This is relevant as it can help policy makers to identify the subgroups of women who may be most in need of improving their financial literacy. The results show that there are marked differences across women according to their education, their occupational status and their household income, while there is hardly any difference by marital status or age. Women who are married have similar levels of knowledge to those who are single, cohabiting and divorced women. This suggests that married/partnered women are not learning from their partners and single women are not learning by doing. In spite of the limited difference across marital statuses, in some countries (including Albania, the Czech Republic, Malaysia, and South Africa) widows show lower financial knowledge than married women, even after controlling for age and income. Even though the age pattern is not significant in many countries, in some of them (such as Ireland, Norway, Peru and the United Kingdom) young women in the age group 18-29 have lower knowledge than women in the 30-59 age bracket. Some evidence of a hump-shaped pattern of financial knowledge by age comes also from other studies. In the US, women under the age of 35 are found to have lower financial knowledge than women in the age group 36-65 (Bucher-Koenen et al., 2012). In Germany, women over the age of 65 gave fewer correct answers than younger women, and widows gave fewer correct answers than married/single women (Bucher-Koenen et al., 2012). Overall, these results suggest that young and elderly women/widows have lower financial knowledge than middle aged ones (as is the case for the population in general).

106

South Africa

UK *

Finally, there appears to be a marked association of financial knowledge with womens occupational status, their education and their household income. In Albania, Ireland, Poland, South Africa and the UK women who are not working (including the unemployed, the retired, students, as well as homemakers) have lower financial knowledge than women who are employed. Similarly, women who live in households with below median income, and who have not completed secondary school have the lowest financial knowledge. Gender differences are smaller but still significant after controlling for socio-demographic factors Gender difference may, to some extent, be related to the different opportunities that women and men have to experience with financial issues along their life, and may therefore be related to observable demographic, social and economic factors. Indeed, empirical evidence shows that gender differences in financial literacy are strongly correlated with differences in socio-economic conditions of men and women. Bucher-Koenen et al. (2012) show that taking into account age, marital status, income and education reduces the apparent gender difference in the Netherlands, although it does not disappear completely. Analogously, a study on the US population shows that accounting for education, income and current and past marital status reduce the observed gap in financial knowledge by about 25% (Fonseca et al., 2010). Also the analysis of the OECD/INFE survey data supports the evidence that gender differences in financial knowledge remain even after taking into account the different socio-economic conditions of men and women47. Figure 44 shows that in Albania, Armenia, Czech Republic, Germany, Ireland, Norway, Peru, Poland and the UK the difference between the financial knowledge score of men and women is lower but still significant after controlling for demographic, social and economic factors with respect to the case when no explanatory variables are considered (with the exceptions of Hungary, where there is no gender difference, and Estonia and the BVI, where the difference does not become smaller). Malaysia and South Africa stand out as exception, as gender differences disappear once socio-demographic factors are accounted for.

47

Based on a multivariate analysis, including a series of separate ordinary-least-squares regression analyses for each country on the whole set of female and male respondents. The outcome variable is the overall financial knowledge score and the explanatory variables are gender, age, marital status, education, household income and occupational status. 107

Figure 44. Gender differences in financial knowledge controlling for socio-demographic factors

Hungary

Albania **

Armenia **

Norway **

Malaysia *

Ireland **

Czech Republic**

Gender difference (M-F) with no controls Gender difference (M-F) controlling for socio-demographics
Notes: the graph is based on the coefficients from an ordinary least squares estimation for each country separately. The results on the left column are based on a model including only a female dummy as explanatory variable; the results on the right column are based on a model including as explanatory variables gender, age, marital status, income, education, and working status. The gender difference is the difference between the average financial knowledge score for men and women. In columns indicated with an asterisk * the gender difference is statically significant at 5% level.

108

South Africa *

Estonia **

Germany **

Peru **

Poland **

BVI **

UK **

WOMENS FINANCIAL ATTITUDES

Evidence from academic and policy research highlights different attitudes and preferences among men and women in relation to financial issues. Even though this evidence is often limited to specific countries or age groups, it helps to shed more light on the different approaches women and men take to financial matters, and it offers some insights about possible strategies to address womens vulnerabilities in the financial domain. Women appear to be aware of their lack of financial knowledge Evidence suggests that women tend to be aware of their lack of knowledge. For instance, BucherKoenen et al. (2012) show that womens self-reported levels of financial knowledge are lower than mens in Germany, the Netherlands and the US. Women also gave themselves lower ratings than men when assessing their personal financial knowledge in a US study about understanding debt concepts (Lusardi and Tufano, 2009). Similarly, when studying US college students, Chen and Volpe (2002) found that young womens selfrated knowledge was lower than that of young men. Similar findings have also been reported amongst high school seniors in the United States (Capital One, 2009). Women have lower confidence than men in their financial knowledge and skills Women are more likely than men to answer do not know to financial knowledge questions, suggesting that they are not only less knowledgeable about finance, but that they are also less confident than men about their financial knowledge. The OECD/INFE data show that in most countries women were significantly more likely than men to say they did not know the answer to a financial knowledge question rather than attempt to answer it (with the exception of Hungary) (Figure 45)48. Women have also been found to be more likely to answer do not know to financial knowledge tests in a variety of other countries, including Azerbaijan, Bulgaria, Germany, Italy, Japan, the Netherlands, New Zealand, the Russian Federation, Sweden and the United States (Alpha Research, 2010; Azerbaijan Micro-finance Association, 2009; Lusardi and Mitchell, 2011; and Bucher-Koenen et al., 2012). Evidence also indicates that women have lower confidence then men in their ability with certain financial issues decisions. The Women Understanding Money research campaign conducted in Australia in 2008 highlighted that women were less confident than men when it came to more complex issues like investing, understanding financial language and ensuring enough money for retirement (Australian Government and Financial Literacy Foundation, 2008). Overall, the evidence on self-assessed knowledge and confidence provides interesting insights. Womens lower confidence with respect to men is consistent with the awareness of their lower knowledge. If women can correctly recognise their limited knowledge, they may be more prudent in their
48

Men may be more likely than women to guess when they don't know the answer. The OECD/INFE questionnaire minimises the likelihood of respondents making correct guesses by using several open response questions. 109

financial behaviour. However, the combination of lower levels of knowledge and a lack of confidence also means that women are less likely to be willing to deal with financial issues, or to access financial products and services, and that they may not necessarily grasp potential opportunities for investment or for income generation.
Figure 45. Percentage "do not know" replies by gender Male
17.5 15.1 11.6 11.3 8.9 4.5 12.9 9.7 8.9 9.3 9.1 14.7 11.1 11.1 14.4 9.9 11.1 6.7

Female

21.5 16.8 13.7 14.2 14.7

19.8

Poland *

Armenia *

Norway *

Estonia *

Ireland *

Czech Republic *

Germany *

Peru *

Notes: The above score is computed as the percentage of do not know answers to seven questions. The figure does not report result for Albania and Malaysia because do not know was not available as an answer in three financial knowledge questions. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Gender differences in interest for financial matters Lower knowledge and confidence may, to some extent, be related to a lack of interest in financial matters. If women are less interested in finance than men it is natural that they may be less motivated to learn. However, it may also be that women show limited interest in money matters because they feel they have too little knowledge to engage in these issues. In a survey of US college students, young women expressed less interest in personal finance than young men. Although both girls and boys thought that personal finance literacy and planning would help them improve their quality of life, male students were more likely than young females to feel that personal finance was important (Chen and Volpe, 2002). In the Netherlands inquisitiveness about money and financial issues and therefore financial motivation is generally very limited among young people. Research found that, on average, no more than a quarter of youths aged from 12 to 18 expressed a desire to learn more about financially-related issues such as savings accounts, taxes, insurance and borrowing money. Boys expressed this interest a little more often than girls (CentiQ, 2008). In Australia, the Women Understanding Money study found that women were not very willing in learning more about everyday money management issues, where they felt more confident, but found that women thought it was important to learn about more complex money management issues, such as
110

South Africa *

Hungary

BVI *

UK *

planning for the financial future, understanding rights and responsibilities when dealing with money and ensuring enough money for retirement (Australian Government and Financial Literacy Foundation, 2008). However, despite of their acknowledgement of the importance to learn, significant numbers of Australian women held attitudes and beliefs that could limit improvements in money management knowledge and skills. Women were more likely than men to find money stressful, uncomfortable or boring and less likely to feel in control of their financial situation. Women are more risk-averse than men There is evidence from experiments and surveys showing that women are less likely to invest in risky assets than men, and that women are more risk-averse compared to men. These results are found in a variety of countries and settings, and are robust to controlling for womens lower income, wealth, and financial knowledge. Using US sample data, Jianakoplos and Bernasek (1998) examined household holdings of risky assets to determine whether there were gender differences in financial risk-taking. As wealth increased, the proportion of wealth held as risky assets was estimated to increase by a smaller amount for single women than for single men, leading to the conclusion that single women exhibited relatively more risk aversion in financial decision making than single men. In another study about the US, Sundn and Surette (1998) examined whether workers differ systematically by gender in the allocation of assets in DC plans. They concluded that gender and marital status significantly affected how individuals chose to allocate assets in defined-contribution plans, after controlling for a wide range of demographic, financial, and attitudinal characteristics. Similarly to previous studies, Halko et al. (2011) studied the relation between gender and stock holdings in Finland. They found that amongst individuals participating in the stock market, men were more likely than women to have a portfolio with a large equity share, even after controlling for other factors including financial knowledge and financial resources. Two studies review experimental evidence on risk aversion. Charness and Gneezy (2011) conducted a meta-analysis of 15 different studies on risk-taking in investment (conducted by different researchers in different countries, with different instructions, durations, payments, and subject pools). Each study used the same mechanism for eliciting risk preferences and gathered data by gender, but gender differences were not the focus of the experiments. They found a very consistent result that women invested less, and thus appeared to be more financially risk averse than men. Eckel and Grossman (2008) reviewed the results from experimental measures of risk aversion for evidence of systematic differences in the behaviour of men and women and again, in most studies, women were found to be more averse to risk than men.

111

GENDER DIFFERENCES IN FINANCIAL BEHAVIOUR AND STRATEGIES

In addition to gender differences in financial knowledge, to some extent men and women also display different financial behaviours. However gender patterns in financial behaviour are more complex than those for financial knowledge. Women are not always outperformed by men in all domains: for instance they are more likely than men to have a budget. Nevertheless, in several countries there are areas where women show vulnerabilities, including product choice, information seeking and saving behaviour. Women are more likely to have a budget Short-term money management is a crucial aspect of behaviour, and as such keeping a close watch on every-day financial expenses is a first step in building long-term financial security and avoiding unsustainable levels of debt. Analysis of the OECD/INFE Financial Literacy Survey allows exploration of the extent to which an individual takes responsibility for household finances and budgeting. This section first looks at who was responsible for day-to-day money management in the household, and then considers whether this person was using a budget. Even though there are no striking gender differences, in many countries women were mainly responsible for household money management in a large share of households. For instance, in the Czech Republic, Peru, Poland, the UK and the British Virgin Islands (BVI) almost one third of married/cohabiting respondents indicated that the wife/female partner had main day-to-day financial responsibility (Figure 46).

112

Figure 46. Responsibility for day-to-day money management decisions in the household 0% Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Peru Poland South Africa UK BVI Female partner Male partner Both partners Other 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Notes: Base: married and living with partner. Other includes responses indicating that the person responsible for money management is: the respondent and another family member; another family member; someone else; nobody. The question about marital status was not asked in Norway.

Combining the responses to two questions that assess how many people report that they a) had either personal or joint responsibility for day to day money management decisions in their household and b) lived in a household with a budget shows that in several countries women were more likely to be responsible for a household budget than men (Figure 47)49. This is in line with the fact that women are significantly more likely than men to report that they keep a close watch on their financial affairs in Estonia, Norway, Poland and the UK (Figure 48), even though gender differences are not very large.

49

This result holds also if we consider who has a budget without conditioning on having personal or joint responsibility for day to day money management decisions in the household. 113

Figure 47. Responsible for money management decision and has a household budget

Male
76% 72% 59%59% 52%50% 60% 49% 35% 32% 26% 24% 23% 20%

Female

43% 30%

60% 53% 47% 46% 43%43% 39%46% 44%43% 29% 22%

Poland *

Hungary *

Czech Republic *

Norway *

Estonia *

Albania

Ireland *

Peru *

UK * UK *

Germany

Malaysia

Notes: Base: all respondents. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Figure 48. I keep a close personal watch on my financial affairs Male

4.0 3.9

Female
4.3 4.4 4.4 4.3 3.8 3.8

4.4 4.5 4.2 4.3 4.3 4.5 4.5 4.4 4.3 4.4 4.3 4.4 4.1 4.3 4.2 4.4 4.5 4.5 4.0 4.1

Albania *

Norway *

Estonia *

Poland *

South Africa

Armenia

Czech Republic

Germany

Malaysia

Armenia

Hungary

Ireland

Peru

Notes: Base: all respondents. The graph indicates to what extent the respondents agree with the statement I keep a close personal watch on my financial affairs on a scale from 1 to 5. The question has been recoded to reverse the original scale so that 1 is never and 5 is always. * the gender difference is statically significant at 5% level.

This is consistent with evidence from other national studies showing that women appear to be better than men at short-term money management behaviour. For instance, in the United Kingdom, women outperformed men at keeping track of finances (Atkinson et al., 2006) and in Canada women marginally outperformed men on making ends meet and keeping track of finances (although these differences were not statistically significant) (McKay, 2011). In Portugal, women gained higher scores than men in a composite index measuring short-term money management capability and saving propensity (Banco de Portugal, 2011). Moreover, the results of the 2006 and 2009 Financial Knowledge Survey in New Zealand suggest that men were less likely than women to control their expenses: women
114

South Africa

BVI

BVI

were more likely than men to report that they kept a fairly close eye on their expenses, or that they used written/electronic records. However, gender differences in the 2006 survey were smaller than in the 2009 one (ANZRetirement Commission, 2009). In the Netherlands, research conducted by the non-forprofit organisation Nibud showed that women were more likely than men to know what their balance is, to give estimates of household expenses, and to plan their spending on costly items (Nibud, 2012). Making ends meet: women tend to cut down on spending while men try to earn extra money Further analysis suggests that men and women are not equally able to make ends meet and that they engage in different strategies when they find that their income does not cover their living costs. First analysis looks at the ability to make ends meet, and then at coping strategies. The OECD/INFE survey results presented in Figure 49 show that there were not only significant differences across countries, but also between men and women, in the ability of respondents to cover their living expenses. In some countries, including Armenia, Germany, South Africa and the BVI women were more likely than men to have experienced problems in covering living costs in the previous year. Since making ends meet can be closely related to the allocation of resources within the household, checks have been made to ensure that the results are very similar when the analysis is restricted to respondents who do not have a partner (i.e. are single, divorced/separated, or widowed). A more in-depth data analysis indicates that in the four countries where there was a gender difference, the difference can be accounted for by controlling for socio-demographic factors (i.e. age, marital status, income, education, work status).
Figure 49. Making ends meet

Male
59%60% 72% 65%

Female
56%57%

47%47% 36%37% 25%24% 22% 13%

46% 41%

35%36% 13%12%

49% 40% 43% 35%

30%33%

46% 37%

Armenia *

Germany *

Czech Republic

South Africa *

Estonia

Norway

Malaysia

Poland

Peru

Albania

Hungary

Notes: Base: all respondents. The graph indicates the share of respondents who answered affirmatively to the question Sometimes people find that their income does not quite cover their living costs. In the last 12 months, has this happened to you? In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

The Core Questionnaire asks respondents about their coping strategies if they report having had problems in making ends meet in the previous year. The flexibility of the questionnaire allows for several options, including drawing on existing resources, cutting expenses, borrowing formally or informally and falling behind with payments.
115

Ireland

BVI *

UK

To some extent men and women appear to have employed different strategies when they needed to make ends meet. In particular, in most countries women are more likely than men to answer that they spent less (Figure 50), while in most countries men are more likely than women to answer that they worked overtime, earned extra money, or took a second job (Figure 51). Respondents did not display differences across genders with respect to other strategies, such as running down existing savings, borrowing, or falling behind with payments. Making ends meet typically involves household resources and is not necessarily an individual behaviour in married or cohabiting couples, but, results are qualitatively the same after excluding respondents who are married or live with their partner from the analyses.
Figure 50. Strategies for making ends meet: cutting back on spending

70%72% 74%73% 55% 50% 60% 46% 31% 41%

80% 69% 70% 64%

Male

Female
64%64%

49% 45% 44% 46% 42% 41% 40% 40% 36% 31%28% 26%

Estonia *

Ireland *

Germany *

Malaysia *

Peru

Albania

Armenia

Hungary

Norway

Poland

Notes: Base: respondents who said that in the 12 months before interview experienced situations where their income did not cover living costs; all marital statuses. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

These results have a number of implications, some of which may differ across countries. The tendency to reduce expenses may be a wise choice for relatively high-income women, but for women who already had problems in covering their living costs a reduction in expenses may make them particularly vulnerable. Reducing current consumption may expose them, for instance, to malnutrition and/or sickness, and may reduce the well-being of their children whenever women are responsible for child-related expenses. In addition, cutting back on insurance and pension payments could undermine their financial security in the future. Moreover, the evidence that women are less likely than men to engage in working activities to increase their income points to their vulnerability in the labour market as well as various potential constraints, including a lack of policies seeking to reconcile work and family life (e.g. childcare, part-time employment opportunities, etc.). To the extent that womens difficulty in accessing economic opportunities is related to a lack of financial knowledge and confidence, many of them could benefit from financial education combined with entrepreneurial education. These could improve the extent to which they can increase their income in a flexible way rather than having to cut expenditure when they fail to make ends meet.
116

Czech Republic

South Africa

BVI

UK

Figure 51. Strategies for making ends meet: earning extra money

Male
20% 28%24% 28% 20% 18%18% 10%15%

Female
22% 13% 18%16% 35%29% 24% 21% 11% 7% 8% 4% 7% 4%

8% 2% 0%

10% 7%

Armenia *

Peru *

Malaysia *

South Africa *

Albania *

Poland *

Estonia

Hungary

Czech Republic

Germany

Ireland

Base: respondents who said that in the 12 months before interview experienced situations where their income did not cover living costs; all marital statuses. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Gender differences in product holding Whenever access to safe and regulated financial products is possible, their appropriate use and choice can be considered as indicators of savvy financial behaviour. It is interesting to look at gender differences in product holding before considering how they are used and how they are chosen in the following paragraphs. The analysis of the INFE survey data reveals different gender patterns in the holding and use of transactional, investment, and insurance products, in addition to sizeable differences between countries (Table 3 in the Appendix collects detailed information on gender differences in product holding). Holding of saving products and saving behaviour is discussed more extensively in the following sub-section. Men are significantly more likely to hold transaction products (i.e., current account, ATM card, etc.) especially in developing and emerging economies (e.g., Albania, Malaysia, Peru, South Africa and the BVI), while there are no significant gender differences in more developed countries (i.e. Czech Republic, Estonia, Germany, Hungary, Ireland, Poland and the UK Figure 52). In Norway women are even more likely than men to have payment products. This is in line with the aggregate evidence showing that in most regions of the world women have lower access than men to formal banking products and that the gender difference is larger in non-OECD countries.

117

Norway

BVI *

UK

Figure 52. Percentage of respondents who hold a payment product Male

71% 68% 90% 86% 78%76% 82%74% 53% 45% 28% 22% 3% 2%

96% 96% 90% 90% 73% 70%

Female
69% 67% 65% 59% 84% 83% 65% 54%

Malaysia *

Peru *

South Africa *

Albania *

Armenia

Hungary

Norway *

Estonia

Notes: In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Men are significantly more likely than women to hold investment products in most of the countries analysed. However, as the holding of investment products is quite low in general, differences tend to be very small in most countries (Figures 53). Gender differences are more pronounced in Germany and the UK. Also in the European Union overall men are more likely than women to hold shares, bonds, and investment funds (European Commission, 2012). Differences in risky financial holdings among men and women have been found also in another study about the UK. Women and men were equally likely to save into low-risk (and typically low-return) saving vehicles, but men were more likely to save into higher-risk/higher-return products, like shares, personal equity plans and unit trusts, compared to women (Westaway and McKay, 2007).

Czech Republic

Germany

Ireland

118

Poland

BVI *

UK

Figure 53. Percentage of respondents who hold an investment product Male Female

45% 34% 19% 2% 0% 0% 0% 7% 4% 7% 7% 5% 4% 13% 3% 2% 29% 31% 31% 27% 8% 4% 8% 5%

38%

27% 30% 28%

Ireland *

Germany *

Peru *

Czech Republic

South Africa *

Albania *

Malaysia

Norway *

Hungary

Poland *

Estonia

UK *

Note: In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Gender differences in saving behaviour Saving behaviour is also an important component of financial literacy, as it builds financial security and reduces the reliance on credit. Exploring gender differences in saving behaviour is relevant because economic, social and cultural factors may generate differences in the ability and propensity of men and women to save. Moreover, women and mens respective decision-making power within the household may influence household decision-making patterns, and therefore also household overall saving decisions (Floro and Seguino, 2002). In analysing gender differences in saving behaviour, it is important to remember that saving typically refers to the household and individual saving decisions are often part of the whole household decisionmaking. Bearing this caveat in mind, it is worth looking at gender differences in saving behaviour because it can help identify womens vulnerabilities in the financial domain, especially for non-married women. According to the OECD/INFE survey, gender differences in saving products holding are not uniform across countries. As shown in Figure 54 men are more likely to have savings products (including savings accounts, term deposit, etc.) in Albania, Malaysia, Poland and the UK, whereas women are more likely to hold formal savings products in Norway and South Africa. In most countries covered by the survey, gender differences in saving holdings are not statistically significant (e.g., Armenia, Czech Republic, Estonia, Germany, Hungary, Ireland, Peru, and the BVI). Also other studies found that women were about as likely as men to have savings products (see Westaway and McKay, 2007 for the UK; Whitaker et al., 2013 for the US).

Armenia

119

BVI

Figure 54. Percentage of respondents who hold a saving product

Male
89% 86% 79% 79% 93% 88%

Female
80% 73% 91% 86% 74% 69%

49%46% 34% 26% 1% 1% 20% 15% 15% 14% 25% 25%29% 15%

7% 7%

Albania *

Malaysia *

Poland *

South Africa *

Hungary

Norway *

Estonia

UK *

Ireland

Notes: In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Another way to gauge the extent of peoples financial reserves is to look at how long they would be able to cover their living expenses (without borrowing money or moving into a different house) in case they lost their income. Results from the OECD/INFE survey shows that in Hungary, Peru, Poland, South Africa, the UK and the BVI, non-married men would be able to cope for a longer period than non-married women without their main source of income (Figure 55)50. Gender differences in the ability to save and the amount of saving are likely to be driven by several factors, whose respective importance may differ across countries. The existing evidence suggests that women are typically able to accumulate significantly less in savings than men during their life cycle because of lower earnings, due to lower labour market participation, occupational segregation, and a higher likelihood to work part-time. This is related to the fact that in many households women are responsible for the majority of unpaid work, including caring for children and the elderly. At the same time, patterns of marriage and family formation have changed over time, with more women being single for some periods in their life, with couples being more likely to separate or divorce, and with more women heading sole parent households. All these factors may reduce womens ability to save, especially for retirement, and may result in women being more likely than men to face financial hardship later in life (Lewis and Messy, 2012; Heathrose, 2012). As women live longer, on average, than men, and typically retire at a younger age, different abilities to save for the long term between men and women whether through pensions or other saving vehicles may imply that women will be more likely to be less well off, to face poverty, or to depend on family members or welfare in old age.

50

This is self-reported information. As men tend to be more over-confident than women, they may also be overoptimistic about how long they would be able to cope in case of an income loss. 120

Czech Republic

Germany

Armenia

Peru

BVI

Another factor potentially explaining gender differences in wealth accumulation is womens higher risk aversion, leading to more conservative portfolio allocations and lower investment returns. More research would be needed to further explore the relation between womens lower savings and gender differences in preferences (especially risk and time preferences) and in financial knowledge.
Figure 55. Ability to cover living expenses in case of an income loss 80% 60% 40% 20% 0% < 1w 1w - 1m 1m - 3m 3m - 6m > 6m < 1w 1w - 1m 1m - 3m 3m - 6m > 6m < 1w 1w - 1m 1m - 3m 3m - 6m > 6m < 1w 1w - 1m 1m - 3m 3m - 6m > 6m < 1w 1w - 1m 1m - 3m 3m - 6m > 6m UK < 1w 1w - 1m 1m - 3m 3m - 6m > 6m BVI

Hungary

Peru

Poland Female

South Africa Male

Notes: Base: respondents who are not married or living with a partner. The graph reports the percentage of male and female non-married respondents who report that they would be able to cover expenses for: less than a week (< 1w); at least a week, but not one month (1w 1m); at least a month, but not three months (1m 3m); at least three months, but not six months (3m 6m); more than six months (> 6m). The graph only reports countries for which the distribution of the outcome variable is significantly different between men and women at 5% level based on a Pearson's chi-squared test. In the other countries the difference is not statistically significant.

Complementing the evidence about womens and mens ability to save, the OECD/INFE survey suggests the presence of some differences in relation to how men and women save. Questions explore whether respondents had been actively saving in the previous year and through which channels, including saving cash at home, paying money to savings account or investment products, participating to informal savings clubs, etc51. In some countries, including Peru, Poland, South Africa, and the UK among non-married respondents, men are more likely than women to be actively saving through formal financial products (such as paying money into a savings account; buying financial investment products other than pension funds; paying in for a term deposit, and/or paying into a building society contract - Figure 56). This is consistent with the results of Figure 53, showing that in some of these countries women are also less likely to use investment products. On the contrary, non-married women are more likely than non-married men to be saving cash at home or in their wallet, or to be saving in an informal savings club in Ireland, Malaysia, and Poland (Figure
51

Gender differences in saving by building up a balance on ones current account and by giving money to ones family to save are not reported here. Men tend to be more likely to be passively saving by accumulating money in their bank account in many countries. The option of saving by giving money to a family member was rarely chosen. Men were more likely than women to choose this option in Estonia, Germany and Malaysia, but less likely in Peru. 121

57)52. Such a propensity to save informally may be a result of several forces, including demand and supply factors. Women may be unable to access formal financial institutions due to supply-side barriers (e.g. geographical distance, high fees, minimum balance requirements, etc.), and/or they may lack sufficient trust and knowledge to approach formal financial institutions and use financial products.
Figure 56. Saving behaviour: Actively saving through formal products Male Female
80% 81% 66% 49% 32% 35% 17% 8% 14% 14% 0% 0% 5% 8% 11% 9% 20% 12% 26% 11% 30% 22% 41% 67% 64%

71%

Albania *

Peru *

UK *
55% 31% 28%

Czech Republic

Notes: Base: respondents who are not married or living with a partner. The question about marital status is not asked in Norway. Actively saving through formal products includes: including paying money into a savings account; buying financial investment products other than pension funds; paying in for a term deposit, and/or paying in to a building society contract. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Figure 57. Saving behaviour: Saving informally Male Female

85% 75% 43% 40% 37% 41% 17% 13% 9% 18% 25% 45% 23% 23%

38% 25%

33% 32%

40% 22% 23% 28%

35%

Ireland *

Malaysia *

Poland *

South Africa *

Germany

Malaysia

Armenia

Estonia

Hungary

Ireland

Poland *

Czech Republic

Albania

Armenia

Hungary

Estonia

Peru

52

The informal saving category in Malaysia also includes buying poultry. In Germany non-married men are more likely than women to save through formal products (savings accounts, investments and building societies), but also to save cash (at home/ in their wallet); saving in informal savings clubs was not among the options in the German survey. 122

South Africa

Germany

BVI

UK

BVI *

Notes: Base: respondents who are not married or living with a partner. The question about marital status is not asked in Norway. Informal saving includes: saving cash at home or in ones wallet, and/or saving in an informal savings club. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Not only do women appear to be saving through different vehicles from men but they also seem to have different motivations. A study about the UK found that there were gender differences in the saving habits of men and women (Westaway and McKay, 2007). Women and men were equally likely to save into low-risk (and typically low-return) saving vehicles, but men were more likely to save into higherrisk/higher-return products, like shares, personal equity plans and unit trusts, compared to women. Moreover, the purpose of saving was different across gender: women were more likely to save for shortterm use (such as special events, holidays, home improvements, and children), whereas men were more likely to save for long-term use (including for old-age, house purchase and cars). As women live longer, on average, than men, and typically retire at a younger age, different longterm saving patterns between men and women whether through pensions or other saving vehicles may imply that women will be more likely to be less well off, to face poverty, or to depend on family members or welfare in old age. Women are less likely to choose financial products appropriately Product choice is also an area of financial behaviour that is relevant for consumers financial literacy. If people attempt to make an informed decision by shopping around or using independent advice they are more likely to choose appropriate products that meet their needs in a cost effective way, less likely to buy something inappropriate, and less likely to be subject to mis-selling or fraud.

123

The OECD/INFE Financial Literacy Survey indicates whether respondents tried to compare financial products across providers in their recent products choices. The results show that, in most of the countries analysed, men were more likely than women to have shopped around for financial products (Figure 58)53.
Figure 58. Has tried to compare financial products across providers Male Female

42% 33% 22% 16% 12% 22% 20% 29% 25%

34% 31%

28%

37% 37% 21% 18%

34% 24%

33%

27% 24%

33% 20%

41% 27% 32%

49% 47%

Germany *

Albania *

Peru *

South Africa *

Armenia *

Norway *

Estonia

Hungary

Ireland

Poland

UK *

Notes: Base: all respondents (Not comparing across providers also includes no recent product choice). In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

The OECD/INFE Financial Literacy Survey also allows us to examine which sources of information respondents were most influenced from when deciding which products to take out. While the Core Questionnaire lists several possible answers, some categories have been combined and the focus is on whether individuals made some attempt at taking an informed decision using non-independent sources (i.e., whether they gathered information from bank staff, friends, trusted individuals or from general media coverage), and whether they consulted independent professional info/advice (including best buy and specialist magazines). Gender differences in the use of information sources were significant only in a few countries. In such cases, women were usually less likely to choose financial products appropriately. Figure 59 shows that in Albania, Armenia, Peru and South Africa men were more likely than women to have made an attempt at taking an informed decision (i.e. by consulting either non-independent or independent sources of information). On the contrary, women in Norway were more likely than men to have tried to take an informed decision. Of the respondents who tried to take an informed decision, men were more likely to have consulted independent professional sources of information and advice in Germany, Norway, South Africa, and the

53

We know that women in some countries are less likely to hold various types of financial products. Since respondents who did not compare across providers also include those who had not made any recent product choice in the past two years, checks have been made to ensure that gender differences in shopping around are qualitatively unchanged if respondents who did not make recent product choices are excluded. This avoids that gender differences in shopping around are driven by gender differences in product holding and product choice. 124

Czech Republic

Malaysia

BVI

UK. On the contrary, women in Ireland were more likely than men to have consulted an independent source (Figure 60). Other financial literacy surveys also highlight womens difficulties with choosing products. The UK financial capability survey showed that women had lower scores in the choosing products domain, which captured various aspects including information collection, the main source of information, product choice, and reading products terms and conditions (Atkinson et al., 2006).
Figure 59. Information sources: made an attempt at taking an informed decision from independent or nonindependent sources Male Female

48% 36% 31% 38% 36% 25% 14% 16%

44% 37% 40% 39% 34% 36% 41% 38% 38%

49% 48%

52% 40% 24% 19%

44%

51% 37% 33%

45%

Armenia *

Albania *

Norway *

Peru *

South Africa *

Malaysia

Czech Republic

Hungary

Estonia

Notes: Base: all respondents. This is derived from a categorical variable taking three values: 0 = No attempt at taking an informed decision (used only unsolicited info received by post, TV ads, other ads) / no recent product choice; 1 = Used only nonindependent sources (gathered info from bank staff, friends, trusted individuals or from general media coverage); 2 = Consulted independent professional information/advice (Independent professional info/advice, best buy and specialist magazines). This graph represents the share of female and male respondents who fall into categories 1 or 2. In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

Other national financial literacy surveys highlight womens difficulty with choosing products more broadly. The UK financial capability survey showed that women had lower scores in the choosing products domain, which captured various aspects including information collection, the main source of information, product choice, and reading products terms and conditions (Atkinson et al., 2006). Analogously, the financial capability survey conducted in Ireland shows that women scored lower than men at staying informed (measured by a composite index based on respondents answers to four variables about the importance of keeping up to date with financial matters, the number of financial topics monitored, the frequency of monitoring financial topics, and the knowledge of whether specified savings and investments are affected by the stock market) (Irish Financial Regulator, 2009). A study by the UK Financial Services Authority (FSA) on women and personal finance highlighted that men were much more likely than women to make decisions for savings accounts and mortgages based on product characteristics and price, while having an existing relationship with a financial institution was the most important factor in womens decision-making. Moreover, the analysis indicated that women were less likely than men to say that they read about financial products and services in newspapers (Graham and Warren, 2001).
125

Germany

Ireland

Poland

BVI

UK

Using a survey of US households, Loibl and Hira (2011) showed that female investors were generally more likely to practice a lower-information search strategy (based on the number of sources used and on the frequency of use), using fewer online and mass media sources compared with male investors. This result is confirmed in a multivariate analysis controlling for demographic and attitudinal factors (including risk tolerance and self-confidence). Womens reluctance to use formal sources of advice may not only be a sign of low financial literacy, but may also be related to the quality of the advice they receive. In an audit study of the market for financial advice in the Boston area in the US, Mullainathan et al. (2012) show that some advisers refused to offer any specific advice as long as the potential client had not transferred his or her account to the company of the adviser. The results show that this behaviour was more pronounced towards females, meaning that female investors were more frequently expected than male investors to engage with advisors before being able to assess their quality. As the evidence on the supply side of financial advice, especially in relation to the quality of advice, is quite limited, more research is needed to explore the existence of any differential treatment to male and female investors, and why.
Figure 60. Information sources: consulted independent professional info/advice Male
82% 54% 35% 20% 7% 7% 3% 3% 10% 25% 15% 12% 16% 12% 34% 41% 28% 21% 19% 24% 20% 13% 9% 10% 8%

Female

73% 44% 50%

Ireland *

Norway *

Germany *

South Africa *

Estonia

Poland

UK *

Czech Republic

Base: respondents who made some attempt at taking an informed decision. This is derived from a categorical variable taking three values: 0 = No attempt at taking an informed decision (used only unsolicited info received by post, TV ads, other ads) / no recent product choice; 1 = Used only non-independent sources (gathered info from bank staff, friends, trusted individuals or from general media coverage); 2 = Consulted independent professional information/advice (Independent professional info/advice, best buy and specialist magazines). This graph represents the share of female and male respondents who fall into category 2 given that they made some attempt at taking an informed decision (categories 1 and 2). In countries indicated with an asterisk * the gender difference is statically significant at 5% level.

126

Malaysia

Albania

Armenia

Hungary

Peru

BVI

FACTORS AFFECTING GENDER DIFFERENCES IN FINANCIAL LITERACY

As shown in the previous section, women display on average lower financial knowledge and confidence, and have more difficulties in some aspects of financial behaviour with respect to men. This evidence calls for a better identification and understanding of what barriers affect gender differences in financial literacy. A number of elements can be considered as potentially driving such gender differences. In particular, various factors may reduce womens opportunities to learn about financial matters and to acquire financial skills, as suggested also by previous research (Hung et al., 2012) and evidence from INFE member surveys. Limited access to education, employment and formal financial markets not only reduce womens financial well-being per se, but also limit the extent to which women can improve their knowledge, confidence and skills about economic and financial issues. These elements will be discussed in the following paragraphs, together with examples of issues and policies from a selected number of countries. The literature on the potential causes of these gender differences in financial literacy is still in its infancy and it is difficult to establish causal links. In practice, causality probably goes both ways: factors limiting womens economic opportunities may reduce their ability to gain more financial literacy, and at the same time gender differences in financial literacy may reinforce disparities in other domains. This highlights the need for policies addressing gender inequalities both in economic opportunities and in financial literacy as a means to improve womens financial well-being. Gender roles in household financial decision making have limited impact A potential explanation for womens lower financial literacy may have to do with specialisation of tasks within the household. The idea that men acquire greater financial knowledge and skills than women because they specialise in financial decision making within the household (for various reasons, including cultural and social norms) is consistent with the fact that men tend to display higher financial knowledge than women across a wide range of countries. However, the evidence regarding a link between gender differences in financial literacy and specialisation within the household is at best mixed. Hsu (2011) finds that the financial literacy of older married women in the US increases as they approach widowhood, thus supporting the idea of household division of labour. Anticipating possible widowhood, women have incentives to begin learning. However, Fonseca et al. (2010) and Bucher-Koenen et al. (2012), studying the US and Dutch population respectively, do not find support for the specialisation hypothesis. Womens lower financial inclusion and access to finance Women have been shown to face greater difficulties than men in accessing finance for business and personal use. In most regions of the world, fewer women than men have an account at a formal financial institution. Moreover, in some regions of the world women are more likely than men to use informal

127

products: in Sub-Saharan Africa more women savers report using an informal savings clubs and not a formal financial institution to save as compared to men (Demirguc-Kunt and Klapper, 2012). The evidence reviewed in the section highlights the existence of substantial gender disparities that can affect gender differences in financial literacy. Even though financial education alone cannot address all these challenges, nevertheless targeted financial education initiatives can address womens needs and help them to improve their awareness, knowledge, confidence and skills in dealing with financial issues.

128

CONCLUSION

The need to address the financial literacy of women and girls as a way to improve their financial empowerment is gaining global relevance and is reflected in the G20 Leaders Declaration in June 2012 recognising the need for women to gain access to financial services and financial education. This chapter identifies variations in the levels of financial literacy by gender, and indicates a pressing need for financial education targeted at women to address the gender gap. This is particularly important in terms of increasing levels of financial knowledge amongst women and girls, and improving ability to make ends meet and choose appropriate financial products. Women are often facing unequal economic and financial opportunities with respect to men in developed and developing countries. This report identifies various womens financial education barriers and needs, and discusses financial education initiatives addressing these issues. It draws on previous OECD/INFE work, including the financial literacy measurement survey and stock-taking exercises among INFE members, as well as on extensive academic research and national evidence on financial education programmes for women and girls. Overall, the analysis of the evidence on gender differences in financial literacy highlights a number of issues: Women display lower financial knowledge and confidence than men in many countries. In particular, young women, elderly/widows, less well-educated and low-income women lack financial knowledge the most; Women appear to be more vulnerable than men in some aspects of financial behaviour, namely in relation to making ends meet; saving strategies; and in choosing products.

A number of factors and barriers appear to affect womens financial opportunities. Gender differences in financial literacy are correlated with differences in socio-economic conditions of men and women, suggesting that limited access to education, employment and formal financial markets not only reduce womens financial well-being per se, but also limit the extent to which women can improve their knowledge, confidence and skills about economic and financial issues. In light of these challenges, various countries have acknowledged the need to improve the financial literacy of women and girls by implementing targeted national financial education policies and initiatives. The most important policy goals of financial education initiatives for women include: Addressing the needs of specific subgroups, such as young and old women, low-income women, and small and micro female entrepreneurs; and

129

Improving womens financial inclusion and financial strategies, including fostering the use of formal saving accounts, preventing over-indebtedness, and supporting women in planning for retirement.

130

REFERENCES

Alpha Research (2010), Financial Literacy Survey, Report on the Key Findings of the Survey Prepared for the World Bank, Alpha Research, June 2010, Sofia, Bulgaria http://siteresources.worldbank.org/FINANCIALSECTOR/Resources/World_Bank_Bulgaria_FinLit_Report.pdf (Accessed on July 12, 2012) ANZ (Australia and New Zealand Banking Group) (2011), Adult Financial Literacy in Australia. Full Report of the results from the 2011 ANZ Survey. December 2011. http://www.anz.com/resources/f/9/f9fc9800493e8ac695c3d7fc8cff90cd/2011-Adult-Financial-LiteracyFull.pdf.pdf?CACHEID=f9fc9800493e8ac695c3d7fc8cff90cd (Accessed on July 12, 2012) ANZRetirement Commission (2009), 2009 ANZRetirement Commission Financial Knowledge Survey http://www.financialliteracy.org.nz/research-/results-anz-retirement-commission-financial-knowledgesurvey/2009 (Accessed on July 12, 2012) Atkinson, A, S. McKay, E. Kempson, and S. Collard (2006), "Levels of Financial Capability in the UK: Results of a Baseline Survey", Consumer Research 47, Prepared for the Financial Services Authority by Personal Finance Research Centre University of Bristol, Financial Services Authority. http://www.fsa.gov.uk/pubs/consumerresearch/crpr47.pdf (Accessed 13 June 2012). Atkinson, A. and Messy, F-A. (2012), Measuring Financial Literacy: Results of the OECD / International Network on Financial Education (INFE) Pilot Study, OECD Working Papers on Finance, Insurance and Private Pensions , No. 15, OECD Publishing. http://dx.doi.org/10.1787/5k9csfs90fr4-en (Accessed 13 June 2012). Australian Government and Financial Literacy Foundation (2008), Financial Literacy Women Understanding Money, Australian Government, Financial Literacy Foundation. http://www.financialliteracy.gov.au/media/209296/women-understanding-money.pdf (Accessed 13 June 2012). Azerbaijan Micro-finance Association (2009), Final Report - Results of the Financial Literacy Survey, Baku, Azerbaijan Micro-finance Association, Azerbaijan: December 30, 2009. http://siteresources.worldbank.org/INTECAREGTOPPRVSECDEV/Resources/Financial_Literacy_Survey_Repor t_EN.pdf (Accessed online on June 20, 2012) Banco de Portugal (2011), Survey on the financial literacy of the Portuguese population http://clientebancario.bportugal.pt/ptPT/Publicacoes/InqueritoLiteraciaFinanceira/Documents/Survey%20on%20the%20Financial%20Literacy%20o f%20the%20Portuguese%20Population%20(2010)%20-%20Final%20Report.pdf (Accessed on July 12, 2012) Bucher-Koenen, Tabea, Annamaria Lusardi, Rob Alessie, and Maarten van Rooij (2012), How financially literate are women? Some new perspectives on the gender gap. Netspar Panel Paper 31, http://arno.uvt.nl/show.cgi?fid=126935 Capital One (2009), Capital One Survey of High School Seniors Reveals Gender Gaps in Financial Literacy, http://phx.corporate-ir.net/phoenix.zhtml?c=70667&p=irol-newsArticle&ID=1301899&highlight (Accessed 6 February 2012).

131

CentiQ (2008), Summary. Financial understanding and behaviour of 8- to 18-year-olds in the Netherlands. CentiQ. http://www.wijzeringeldzaken.nl/media/13194/summary_financial_understanding_%208to18_netherlands. pdf Charness G. and U. Gneezy (2011), Strong Evidence for Gender Differences in Risk Taking. Journal of Economic Behavior and Organization, 83 (1), pp. 50-58. http://www.sciencedirect.com/science/article/pii/S0167268111001521 Chen, H., and R.P. Volpe (2002), "Gender Differences in Personal Financial Literacy Among College Students," Financial Services Review, Vol. 11, No. 3, 2002, pp. 289-307. http://www2.stetson.edu/fsr/abstracts2/Vol11_A18.pdf Demirguc-Kunt, Asli, and Leora Klapper (2012), "Measuring Financial Inclusion: The Global Findex", World Bank Policy Research WP 6025. http://econ.worldbank.org/external/default/main?pagePK=64165259&theSitePK=469372&piPK=64165421& menuPK=64166093&entityID=000158349_20120419083611 (Accessed on July 12, 2012) Eckel, C. C. and P. J. Grossman (2008), Chapter 113 Men, Women and Risk Aversion: Experimental Evidence, In: Charles R. Plott and Vernon L. Smith, Editor(s), Handbook of Experimental Economics Results, 2008, Volume 1, Pages 1061-1073. http://www.sciencedirect.com/science/article/pii/S1574072207001138 (Accessed 13 June 2012). Floro, Maria, and Stephanie Seguino (2002), Gender Effects on Aggregate Saving, Policy Research Report on Gender and Development, Working Paper Series No. 23, the World Bank http://siteresources.worldbank.org/INTGENDER/Resources/wp23.pdf (Accessed on July 12, 2012. Fonseca, Raquel, Kathleen J. Mullen, Gema Zamarro, and Julie Zissimopoulos (2010), What explains the gender gap in financial literacy, RAND Working Paper No. WR-762, June 2010. http://www.rand.org/content/dam/rand/pubs/working_papers/2010/RAND_WR762.pdf G20 (2012), G20 Leaders Declaration, Los Cabos, Mexico. http://www.g20.org/images/stories/docs/g20/conclu/G20_Leaders_Declaration_2012.pdf Graham, M. and R. Warren (2011), Women and personal finance: the reality of the gender gap . Financial Services Authority, Consumer Research 7. http://www.fsa.gov.uk/pubs/consumer-research/crpr07.pdf Halko, M.-L., Kaustia, M., Alanko, E. (2011), The gender effect in risky asset holdings. Journal of Economic Behavior and Organization, 83 (1), pp. 61-81. http://www.sciencedirect.com/science/article/pii/S0167268111001569 Hsu, Joanne W. (2011), Aging and strategic learning: The impact of spousal incentives on financial literacy, Networks Financial Institute Working Paper, No. 2011-WP-06, Indiana State University. http://www.federalreserve.gov/pubs/feds/2011/201153/201153abs.html Hung, A., J. Yoong and E. Brown (2012), Empowering Women Through Financial Awareness and Education, OECD Working Papers on Finance, Insurance and Private Pensions , No. 14, OECD Publishing. http://dx.doi.org/10.1787/5k9d5v6kh56g-en (Accessed 13 June 2012). Jianakoplos, N. A. and A. Bernasek (1998), Are Women More Risk Averse? Economic Inquiry. 36 (4), pp. 620630, October 1998. http://onlinelibrary.wiley.com/doi/10.1111/j.1465-7295.1998.tb01740.x/abstract (Accessed 13 June 2012).

132

INFE (2013), INFE Policy Guidance on Addressing Womens and Girls Needs for Financial Awareness and Education, in OECD/INFE set of criteria, principles, guidelines and policy guidance to improve financial education. Part 2: addressing youths and womens needs for financial education . OECD, Paris. Kelly, S and H. Gong (2010), Saving Tomorrow - The savings and spending patterns of Australians, AMP.NATSEM Income and Wealth Report, Issue 25, April 2010, AMP, Sydney. http://www.melbourneinstitute.com/downloads/hilda/Bibliography/Other_Publications/AMP.NATSEM_Savi ng_Tomorrow_Report.pdf Klapper, L. and G. A. Panos (2011), Financial literacy and retirement planning: the Russian case, Journal of Pension Economics and Finance, 10 (4): 599618, October, 2011. http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8403939 Lewis, S. and F-A. Messy (2012), Financial Education, Savings and Investments: An Overview, OECD Working Papers on Finance, Insurance and Private Pensions, No. 22, OECD Publishing. http://dx.doi.org/10.1787/5k94gxrw760v-en Loibl, C. and T. K. Hira (2011), Know your subject: A gendered perspective on investor information search, Journal of Behavioral Finance, 12, pp.117-130. http://www.tandfonline.com/doi/abs/10.1080/15427560.2011.600841#preview (Accessed on July 12, 2012) Lusardi, A. and O. S. Mitchell (2011), "Financial literacy around the world: an overview", Journal of Pension Economics and Finance, October 2011, 10 (4), pp. 497-508 . http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8403921 (Accessed 13 June 2012). Lusardi, A., O. S. Mitchell and V. Curto (2010), Financial literacy among the young, Journal of Consumer Affairs, 44(2), pp. 358-380 http://onlinelibrary.wiley.com/doi/10.1111/j.1745-6606.2010.01173.x/abstract Lusardi, A. and P. Tufano (2009), Debt literacy, financial experiences, and overindebtedness, NBER Working Paper, no. 14808. http://www.nber.org/papers/w14808 McKay, S. (2011), "Understanding Financial Capability in Canada Analysis of the Canadian Financial Capability Survey", Research paper prepared for the Task Force on Financial Literacy http://publications.gc.ca/collections/collection_2011/fin/F2-213-2011-eng.pdf (Accessed on 19 January 2012). Mullainathan, S., M. Noeth and A. Schoar (2012), The Market for Financial Advice: An Audit Study, NBER Working Paper 17929, http://www.nber.org/papers/w17929 Nibud (2012), Mannen, vrouwen en geldzaken, Nibud Factsheet June 2012. http://www.nibud.nl/fileadmin/user_upload/Documenten/PDF/onderzoeken/2012/Factsheet_man_vrouwe n_geldzaken.pdf OECD (2005a), Improving Financial Literacy: Analysis of Issues and Policies, OECD, Paris. OECD (2005b), Recommendation of the Council on Principles and Good Practices on Financial Education and Awareness. OECD (2012), Gender Equality in Education, Employment and Entrepreneurship: Final Report to the MCM 2012 [C/MIN(2012)5], OECD, Paris. http://www.oecd.org/els/familiesandchildren/50423364.pdf OECD/INFE (2012), High-Level Principles on National Strategies for Financial Education

133

OECD/INFE (2013a), OECD/INFE toolkit to measure financial literacy and inclusion: Guidance, Core questionnaire and Supplementary Questions, OECD, Paris. OECD/INFE (2013b), Current Status of National Strategies for Financial Education: OECD/INFE Comparative Analysis and Relevant Practices OECD/INFE (2013c), Gender differences in financial literacy: identifying barriers and addressing womens and girls needs through financial education in OECD, Women and Financial Literacy: OECD/INFE Evidence, Survey and Policy Responses, OECD, Paris. Schmidt, Lucie and Purvi Sevak (2006), Gender, Marriage, and Asset Accumulation in the United States, Feminist Economics, 12 (1-2), pp. 139-166 http://www.tandfonline.com/doi/abs/10.1080/13545700500508445 Siermiska, Eva M., Joachim R. Frick, and Markus M. Grabka (2008), Examining the Gender Wealth Gap in Germany, DIW Discussion paper 806, Deutsches Institut fr Wirtschaftsforschung, Berlin. http://ssrn.com/abstract=1424899 Sundn, A. and B. Surette (1998), Gender differences in the allocation of assets in retirement savings plans. American Economic Review Papers and Proceedings 88, 207211. http://www.jstor.org/stable/10.2307/116920 (Accessed on July 12, 2012). Westaway, Jenny, and Stephen Mckay (2007), Womens financial assets and debt, Fawcett Society, London. http://www.fawcettsociety.org.uk/documents/Fawcett%20Assets%20Report.pdf (Accessed on July 12, 2012).

134

ANNEX 1: TESTING THE CORE QUESTIONNAIRE AND DEVELOPING A SCORE

Introduction This annex outlines the lessons learned whilst developing the OECD/INFE Financial Literacy Core Questionnaire and undertaking the first measurement project. It also provides detailed information about the method used in analysing data collected from the OECD/INFE Financial Literacy Survey. It includes information about the processes used to prepare the data for analysis, and the decisions made during the analysis in order to create a measure of financial literacy that will provide cross-country comparisons. It is based on the first 11 countries to submit data. Design of the Core Questionnaire The OECD/INFE Financial Literacy Survey uses a unique questionnaire, developed specifically to capture financial literacy at an international level. The questionnaire is described below as the OECD/INFE Financial Literacy Core Questionnaire, or Core Questionnaire for short. The OECD/INFE has defined financial literacy as follows: A combination of awareness, knowledge, skill, attitude and behaviour necessary to make sound financial decisions and ultimately achieve individual financial wellbeing. The Core Questionnaire provides a number of ways in which the country can tailor responses to national circumstances, in the interest of allowing flexibility across countries. It was designed to minimise any negative reaction to sensitive questions for example by carefully positioning them within the overall set of questions - and to ensure that there was no unnecessary repetition or wordiness (hence asking whether the respondent has a partner, before asking questions about the household). Probes and read out categories should maximise the chance of getting the type of response intended and open ended questions are used in some places to limit guessing. Data collection and preparation process Countries were invited to apply the questionnaire according to the recommended method and submit their data to the OECD/INFE Secretariat for comparative analysis. The collection process was therefore undertaken at a country level. Once each country had finalised data collection and entry, the OECD Secretariat gave each dataset consistent labels and checked the data for obvious errors or omissions in order to prepare them for international comparison. This checking process continued throughout the analysis stage, as some issues are only apparent once the data from one country is compared with other countries54.

54

Whilst it is very likely that there are differences across countries, it is also possible to identify potential problems when one country appears to be extremely different. Often this is because of an alternative approach to data coding (for example reversing the scale on an attitude statement). Where issues have been identified in the data submitted, these have been addressed with the relevant agencies and revised datasets have been submitted if necessary. 135

Data preparation also included deriving new variables from existing information. This included gathering information that was spread over two or more questions and separating out multiple responses that had been collected from a single question. It is very unusual for every participant in a survey to answer every question, although every effort should be made to keep refusals to a minimum. In some situations, an individuals record may be removed from the dataset because the responses are considered to be unreliable, or unusable. This was necessary in a very small number of cases during the preparation of the data. Other missing data was handled as appropriate during the analysis process (typically by taking the conservative approach of assuming the respondent was not financially literate on that measure). From survey questions to financial literacy scores The Core Questionnaire asks a range of questions about issues that are widely considered to be aspects of financial literacy. Each of these questions provides valuable information, but no single question can adequately indicate levels of financial literacy. As the definition above highlights, financial literacy incorporates a combination of factors. Together, the questions provide valuable information about the knowledge, behaviour or attitudes of the respondents. An overall financial literacy measure is expected to indicate something that cant be captured simply by asking the right question, it has to be created through combining information about a respondents knowledge, behaviour and attitudes. In order to incorporate information about a combination of factors it is necessary to develop a score across several questions or segment the population according to their responses to particular questions or groups of questions. Both of these approaches have been used successfully in a national context. As the definition of financial literacy stresses that it is a combination of three components, namely knowledge, behaviour and attitude, the approach taken with the OECD/INFE data incorporates measures of each of these three components as well as combining them to develop an overall measure of financial literacy. It is important that the data preparation and analysis is undertaken in such a way that another researcher will be to be able to achieve the same results using the same data. Similarly, it should be possible to apply the same approach to new data to create comparable measures of financial literacy. Two analysts at the OECD have worked together on the analysis, using IBM SPSS Statistics v19 to manage, check and analyse the data. Developing an internationally applicable method The questions have a certain amount of flexibility in order to ensure that they are contextually relevant across countries (this includes the possibility of changing the types of financial products being discussed). This flexibility does not alter the substance of the question, and it is this that should be captured in a measure of financial literacy. For example, a list of possible approaches to saving money may vary by country, but the process under investigation is the act of saving, and this substantive issue can be explored irrespective of the actual approaches taken. Consistency is vital when comparing data across countries, and can be challenging when questions have been contextualised. In order to consistently reflect the substance of the questions it is essential that the responses to the questions are mapped to a central analytical framework and that multiple
136

responses are combined in the same way. Such an approach will ensure comparability across countries and enable future comparisons. The Core Questionnaire includes questions relating to managing money, planning ahead, choosing financial products and staying informed. National surveys have often been analysed according to these domains. However, exploratory factor analysis indicated that whilst there are considerable similarities across countries, there is no single way of identifying which questions relate to regular money management and which are capturing longer-term planning ahead behaviours. It is therefore difficult to create comparable measures of the domains across countries. In contrast, knowledge, behaviour and attitudes are concepts that do not vary by country, making comparisons meaningful and achievable.

137

KNOWLEDGE SCORE

Knowledge of key financial concepts and the ability to apply basic numeracy in financial situations are seen as core requirements in order to consider a person to be financial literate. The Core Questionnaire therefore asks a range of questions in relation to concepts such as simple and compound interest, risk and return, and inflation. This is in contrast to some national surveys, which focus more on knowledge of particular financial services available or issues such as tax and retirement funding. Survey questions designed to test knowledge The Core Questionnaire includes eight questions designed to test knowledge. These vary in style in order to reduce potential biases that could be caused either by different ways of processing information across certain groups of people or by cultural norms. In particular, some are open ended whilst others provide alternative options from which the respondent must choose their response. The knowledge questions also vary in content and complexity although none of them requires technical (financial) knowledge. They have been chosen to reflect the kinds of knowledge that would help individuals to make informed decisions and monitor their own finances. During the design phase of the Core Questionnaire, it was argued that some of the questions were too complex or abstract to ask amongst low income, financially excluded respondents. Alternative wording was therefore provided to capture the same basic concept whilst minimising the risk of losing comparability. Lessons from the first measurement project Level of difficulty In order for the questions to provide meaningful information about the level of financial literacy of individuals and populations, the questions must be able to differentiate high and low achievers. This means that the question should be neither too hard nor too easy and should not cause the respondent to refuse to answer. Several country representatives expressed concern about the level of complexity of certain knowledge questions used in the Core Questionnaire. Both points of view were heard the questions were too easy, and the questions were too difficult. If the questions were indeed too easy or too difficult, then very few people would get them wrong, or right, respectively, and it would be impossible to create a meaningful indicator of the level of financial knowledge. Questions that are too easy may also potentially confuse respondents they may think they are missing some information, or that there is a trick in the question. The questionnaire includes a statement to help overcome any such problem (the questions are not designed to trick you, so if you think you have the right answer, you probably do).

138

An additional concern with difficult questions is that people will be dissuaded from continuing to answer the survey, which would negate the whole measurement process. To overcome this, and also to reduce the possibility of people guessing, the questionnaire reminds respondents if you don't know the answer, just say so. The analysis indicates that the level of complexity was appropriate. Some questions were sufficiently easy that most people were able to answer them (this is good, as it helps the respondent to feel at ease, and it identifies the real outliers who cant answer even the simplest question). Other questions were more difficult, identifying those with a higher level of knowledge. Just two countries used the alternative wording of the questions on Risk and Return, and Diversification. South Africa used only the alternative wording whilst Malaysia tested both versions. The Malaysian data shows a significant and high correlation between the two versions, providing reassurance that they are capturing knowledge of the same underlying concept. Very few people refused to answer the knowledge questions across all participating countries. Given the range of countries participating and the heterogeneity of their populations, this indicates that the questions are appropriate across a wide range of respondent types. It is recommended that countries that are interested in using the Core Questionnaire, but still have concerns about the level of difficulty of the knowledge questions draw on additional questions from other sources to complement the questionnaire, including from our catalogue of Supplementary Questions (available online). Drawing on additional questions will also provide countries with the opportunity to ask about the same concept in more than one way, so providing greater opportunities to check whether respondents really understand. Changes to the question wording Whilst the instructions for the measurement process made clear that questions should be translated for meaning but should not be changed, some countries nevertheless made adjustments to the questions before asking them. This created the need for additional tests to ensure that the changes had not jeopardised the international comparability of the data. However, as it will be difficult to run similar tests in future, it is strongly advised that countries using the questions keep to the original wording. Additional questions can be added to the questionnaire if necessary to address country level issues or to collect additional information about understanding of certain concepts. A slight adjustment has been made to the finalised Core Questionnaire in response to these unilateral changes. Specifically, the phrase no fee has been added into the question designed to test the calculation of interest plus principle, so that respondents in countries where savings accounts attract fees do not take the cost of fees into consideration when answering. The OECD/INFE also considered changing the wording of the Interest question, to reflect the fact that the word interest may be culturally sensitive. However, as the question is specifically designed to test whether people know how to identify interest on a payment, it was felt that a change was not appropriate. Individual countries will need to decide whether they can use the word interest, or should replace it with something less sensitive (such as return); they can be reassured that the results of the analysis suggest that the overall results are not particularly sensitive to such a change. Changes to the wording should be clearly identified in the data collection process, however, in order to make sure that the correct response is identified during data coding.
139

The Time Value of Money question (QK3) has been adjusted to reflect the fact that it was difficult to identify the correct response across respondents with very vivid memories of inflation. This also addresses concerns voiced by analysts in South Africa who felt that the question may not be fully capturing understanding of inflation. The question now includes the phrase and inflation stays at X 55 percent to make the purpose of the question clear and ensure that it is relevant within countries. Malaysia tested the alternative wording for the risk and reward and diversification questions by asking every respondent both questions, and analysis shows that responses to the two alternatives were significantly correlated. This is reassuring for countries wishing to rely on the alternative wording. Developing a score Analysis of the responses to each question by country indicates that the combination of knowledge questions adequately identified high and low achievers in all countries. It also shows that relatively few people refused to answer the questions56. The score is therefore based on derived variables that identify correct answers to each of the knowledge questions. The following approach was used to assign a value to the responses to each question.
Table 13. Creating a knowledge score
Question Discussion This is open response and a correct answer is therefore a good indicator of applied numeracy Value towards final score 1 for correct response. 0 in all other cases.

Division Time-value of money

This is multiple response and context specific

1 for responses c, d, e unless country situation suggests otherwise (in the final version of the questionnaire, the correct response will depend on X). 1 for correct response. 0 in all other cases.

Interest paid on a loan

This is open response and a correct answer is therefore a good indicator of understanding This is open response and a correct answer is therefore a good indicator of applied numeracy This is multiple response. It is assumed that if the respondent couldn't calculate 2% they also cannot calculate 5*2%. This is a yes/no question so guessing may occur This is a yes/no question so guessing may occur This is a yes/no question so guessing may occur

Calculation of plus principle

interest

1 for correct response. 0 in all other cases.

Compound interest

1 for a correct response IFF the previous response was also correct. 0 in all other cases. 1 for a correct response. 0 in all other cases. 1 for a correct response. 0 in all other cases. 1 for a correct response. 0 in all other cases.

Risk and return Definition of inflation Diversification

55 56

X will be replaced by a valid percentage in each participating country. The analysis also shows that some people responded don't know. In the case of the knowledge questions, don't know is considered to be a valid response indeed respondents were encouraged to say if they didn't know the answer - and so this will not cause a problem in analysis. 140

There are several approaches that could be used to develop a score from these individual items. After considering several factors influencing the development process it was agreed to develop a simple count the number of correct responses. In particular: Some questions are more difficult than others, and, arguably, some may be more important. Should all correct answers be given equal weight? Some experts have argued that the first question is numeracy, rather than financial literacy, and have suggested that this should not go into the score. Some countries have made adjustments to the questions or opted for the alternative wording. How much does this affect the relative scores?

Two types of test were used to identify the most appropriate approach. Factor analysis was used to look at the extent to which each question adds to an underlying financial knowledge factor. Several additive knowledge scores were also created, to test the impact of using or omitting certain questions on the overall score. An attempt at undertaking confirmatory factor analysis of the knowledge questions highlighted the difficulties in using such an approach to create a score for this survey. In particular, the analysis is sensitive to decisions about which observations (or countries) and questions are included, and how missing data is handled. The apparent weighting or importance of each item included varies accordingly. It should also be noted that factor analysis seeks to correlate a range of factors with an unobserved, underlying factor. If some of the questions appear to be important to experts, but do not correlate to this artificially derived factor, then they are effectively deleted from the measure despite being valid components of financial literacy. Despite the challenges, the factor analysis provides some insights57. An analysis of seven questions across 10 of the participating countries58 (excluding the compound interest question, which relies on the correct response to the simple interest calculation making collinearity an issue) indicates that there is an underlying construct related to knowledge of simple interest, risk diversification and the relationship between risk and return. Neither knowledge of inflation nor interest on loans is well captured in this construct. It would be possible to create a score for each respondent from the factor analysis, and this approach is widely considered to be good practice when scoring complex data59. However, there is also a strong argument for giving each component of financial knowledge equal weighting, as each has benefits for individuals, and each has been identified as important by international experts. There is also some sense in avoiding complex statistical approaches if these are likely to be applied or interpreted in
57

Following the recommendations of Rowe 2006 to the extent possible, we used confirmatory factor analysis with a one factor solution under Maximum Likelihood estimation using IBM SPSS 19. South Africa was excluded because we wanted to test a specific set of questions, rather than the alternatively worded questions. Albania and Peru was excluded because of late receipt of data. BVI was not part of the first measurement project and so data is not included in any of these analyses. This argument is based on the fact that a score that counts correct answers may be misinterpreted; people may assume it is equally difficult to gain one additional point from anywhere on the scale when in fact some questions are more difficult than others. 141

58

59

different ways in different countries or if changes/problems with data from one country are likely to influence the way the data from other countries is analysed. The final scores are therefore based on a simple count of correct answer, rescaled to 100, as may be used in a school test. The questions within the Core Questionnaire were all chosen because they were considered to capture essential aspects of financial knowledge and so have been given equal weighting in the score. The set of eight questions has been tested by excluding certain questions which may be having an unintended influence on scores and looking at the this changes the ranking of country scores to see if the change has a major impact on the apparent relative level of financial knowledge in any country (defined as changing the relative position of a country by more than 2 places - e.g. dropping from 2nd to 5th relative to a score using all the knowledge questions with equal weighting).
Table 14. Financial knowledge scores
Ranking of country according to scoring method used
Knowledge score across all 8 questions
Hungary Estonia Germany Ireland Czech Republic UK Malaysia Poland Armenia South Africa

Knowledge division
Hungary Germany Estonia UK Ireland

score

excluding

Knowledge score compounding

Hungary Estonia Ireland Czech Republic Germany UK Malaysia Poland Armenia South Africa

excluding

Knowledge score interest on loan

Hungary Estonia Germany Ireland Czech Republic UK Poland Malaysia Armenia South Africa

excluding

Czech Republic Malaysia Poland Armenia South Africa

Norway is excluded from this analysis because they have both substituted and omitted questions, making direct comparisons difficult. Albania and Peru are also excluded as the analysis predates receipt of their data.

Excluding the division question has no major impact on the ranking of countries (because the question was answered correctly by the majority of respondents). As the majority of countries and experts felt that basic applied numeracy is an important aspect of financial literacy, this is kept in the measure. Excluding the compounding question, which was made easier in Hungary, makes very little difference to the relative ranking of the countries, and Hungary does not move position. This question is only considered to be correct if the previous question, on simple interest, was also answered correctly. Removing the question indicates that the Hungarian population do have relatively high levels of financial knowledge and are not simply being mis-graded through the reworded question. Malaysia made a slight change to the question about interest on a loan, as the wording of the question was not felt to be appropriate for Muslim consumers. Removing this question from the final score also makes very little difference to the ranking of the countries.
142

The findings of this analysis suggest that it is appropriate to keep all eight knowledge questions. However, for the uncertain questions to add something, they must also have some explanatory power. One way of testing this, is to see whether people who give correct answers to these questions have higher levels of financial knowledge based on the other questions. This test confirmed that financial knowledge (tested using the remaining questions) was higher amongst those that got the three uncertain questions right than those who gave incorrect answers. Given these results all eight knowledge questions are used in the final measure.

143

BEHAVIOUR SCORE

Survey questions designed to capture behaviour The OECD/INFE Financial Literacy Core Questionnaire has a wide range of behavioural questions asked in different ways to capture the maximum amount of information. This gathers information on the way people manage their money, including whether they borrow to make ends meet, whether they typically pay bills on time, if they are personally (or jointly) responsible for a household budget, if they report that they keep a close watch over their finances and if they consider carefully whether they can afford something, whether they attempt to save and set long term goals, and how they choose financial products: including a question about shopping around and one about information and advice. Lessons from the first measurement project These questions are behavioural questions but they included quite sensitive financial questions such as whether the respondent is actively saving and whether they hold certain financial products. The number of refusals on some questions indicates that some interviewers may need to be provided with additional tools and guidance in order to maximise responses. In particular, the respondent must genuinely believe that the responses will not be stored along with any personal information and that the interviewer has no interest in their personal situation other than to record the information for research purposes. The interviewer must also believe that there is no embarrassment in asking personal questions of the respondent, and must neither show surprise nor make judgements or comments in relation to the responses. Given that some people refused to answer some of the savings questions, it was difficult to incorporate all of the information about saving in the final measure. In particular, many people refused to provide an answer to the question about how long they could continue to cover living expenses if they lost their main source of income; this question has not been included in the financial literacy scores, but could have been a useful indicator of the outcome of financial education. The question has been left in the final version of the questionnaire because it can capture valuable information once non-responses are minimised. Changes to the Core Questionnaire The first measurement project uncovered some issues with the instructions on the financial product questions. In particular, several countries recorded the way in which a recent product was chosen but did not have an indicator of which product was being discussed. Whilst this does not affect the way in which the data is used in the financial literacy score, it does reduce the usefulness of the data for national purposes. An instruction has therefore been added into the revised questionnaire to keep a record of the question being discussed. Country feedback also indicated that it was very difficult to implement rotation of products from the list of options and that this caused mistakes to occur in the field. This instruction has therefore been removed (this is a compromise; rotation is useful to avoid order effects in long lists).
144

There have been some changes to the behavioural statements in the final version of the questionnaire combining them with the attitude statements. The reasons for this are as follows: The combination allows interviewers to work their way through just one set of statements with one kind of answer code reducing the time taken to ask the questions, and reducing the complexity of the questions for respondents (previously the two sets of questions had differently labelled scales) Combining the questions reduces the chance of biases caused by repeated patterns. This is because there will be a mixture of questions that require a 4 or 5 on the scale to indicate financial literacy and those that require a 1 or 2.

The approach taken in particular countries also suggested that it was necessary to reconsider the use of a 5 point scale on these questions. South Africa provided a verbal description of each of the points (always, often, some of the time, seldom, never)60 whilst Norway used additional points (a 7 point scale). The verbal description is not usually recommended when survey questions are going to be translated because of the difficulty in translating qualitative nuances that are captured with words such as seldom. Instead, the use of a show-card during face-to-face interviews can be considered. These would have a scale from Always to Never, so that the respondent can place themselves on the scale in front of the interviewer. The financial literacy score does not require 7 points on the scale and so the final version of the questionnaire continues to use 5. The original version of the questionnaire included an optional definition of budget that the interviewer could read out if they felt that the respondent had not understood the word budget. It was not clear from the analysis if this worked successfully in the field, and so reading the definition is now obligatory in order to ensure consistency across countries.

60

Recent exploratory work undertaken by the World Bank under the Trust Fund suggests that people with low levels of education find it difficult to put themselves on a scale. They have developed an alternative approach using yes/no responses and follow-up questions. 145

Deriving variables and developing a score In order to create a financial behaviour score several variables needed deriving as shown in the table below:
Table 15. Financial Behaviour Variables
Derived variable for inclusion in financial behaviour score Personally or jointly responsible for a household budget Carefully considers affordability before making a purchase Pays bills on time Keeps a close watch Sets long term financial goals and strive to achieve them Has not borrowed to make ends meet Has been actively saving or buying investments in the past 12 months. Making informed financial product purchases Questions used to create derived variable (Paraphrased wording of questions from the draft questionnaire) QF1 Who is responsible for day to day money management decisions in your household? QF2 Do you have a household budget? QM1 Before I buy something I carefully consider whether I can afford it QM1 I pay my bills on time QM1 I keep a close personal watch on my financial affairs I set long term financial goals and strive to achieve them QM2 Sometimes people find that their income does not quite cover their living costs QM3 What did you do to make ends meet QP1 in the past 12 months have you been saving money in any of the following ways? QC2 which statement best describes how you last chose QC3 which sources of information influences your decision Approach taken

Score of 1 IFF respondent is responsible for day to day money management (personally or jointly) AND the household has a budget. Give a score of 1 if respondent puts themselves at 1 or 2 on the scale Score of 1 if respondent puts themselves at 1 or 2 on the scale Score of 1 if respondents puts themselves at 1 or 2 on the scale Score of 1 if respondent puts themselves at 1 or 2 on the scale Score 1 unless QM3 response includes borrowing (Note that for the purposes of this score following behind with bills is not considered to be borrowing). Score 1 if actively saving in any of the ways listed: i.e. any option other than building up a balance in current account which is not an active behaviour. This is a 3 stage process. a) Score 1 if considered several products from different companies, or looked around at QC2. b) Score 1 if took some advice at QC3- not independent; Score 2 if took independent advice. c) Create the final score for this aspect which takes the value of 2 if a+b=3, and the value of 1 if a+b=1 or a+b=2.

The approach taken above was developed through discussion with various stakeholders and following exploratory analysis. There was considerable discussion about the household budget measure. It should be noted that the final measure only identifies people who have some responsibility for the household budget. Other respondents may be budgeting their own money; this information could be captured with an additional question. The behavioural questions include four that use a scale, enabling people to provide a more detailed description of their behaviour. Only a small percentage of people refused to provide an answer to one or more of the questions using a behavioural scale. A refusal has therefore been recoded to be scale point 3 before undertaking analysis. As noted above, these questions have been asked in the same way in every country except Norway and South Africa. Where the scale is changed (and particularly if it is reversed) it is important that care is taken to create the derived variable appropriately). Financially literate people will have strategies to smooth income flows and a tendency to avoid using credit for essentials. The extent to which these strategies are successful will depend on predictability of both income and expenditure, as well as the skill involved. It is not always possible to prevent shortfalls in income, but a reliance on credit for basic living can become very dangerous, and may create a situation that becomes impossible to escape. A variable has therefore been created from QM2 and relevant responses to QM3 to identify people who reported that sometimes their income doesn't meet their needs, and that the last time this happened they had to borrow to make ends meet.

146

Respondents were asked In the past 12 months have you been saving money in any of the following ways? A list was then provided, which was tailored to the country context. For the purpose of the international comparison a variable has been created that counts all kinds of saving as active saving, except for building up a balance in a current account. This is an appropriate indicator of behaviour, since it indicates that saving was intentional rather than just a default situation due to high income. Discussing savings is a sensitive issue in some countries. Almost one in 5 respondents refused to answer this question in the Czech Republic (19%), as did 14% in South Africa. In Poland a very large proportion claimed that they didn't know, which almost certainly indicates an unwillingness to divulge such information. These individuals have been considered to be non-savers in the score, which provides a conservative approach to measuring the true level of financial literacy. The way people behave when choosing financial products is also an important aspect of their financial literacy. If people attempt to make an informed decision, by shopping around or using independent advice they are more likely to choose appropriate products that meet their needs in a cost effective way. People do not typically choose financial products on a weekly, or even monthly, basis. The questionnaire therefore asks about a product chosen in the last 2 years. Even so, a large proportion had not actively chosen a product in this timeframe61. This suggests that they could be encouraged or supported to become more active consumers; it is therefore appropriate for them to get a low score on this aspect of financial literacy. The possible approaches to choosing financial products may vary by country (and countries are able to add their own options to the questionnaire), but the general willingness to shop around to gather information is the behaviour that is being captured. In the derived variable used in the final score, respondents are considered to have made some attempt to make an informed decision if they tried to compare across providers (even if they found out that there were no other providers), or if they sought information from someone. Given the way that the derived variable is created, it is important that all authorities using the questionnaire think carefully about the options that should be included in the survey instrument, as this will directly influence the quality of the data. Some individuals may assess their financial product holding regularly, and decide to keep their current portfolio or renew rather than replace specific policies and products. The product choice score does not take into account product monitoring, but this is something that could be captured through an additional question and may partially explain low levels of product choice in some countries.

61

Some countries did not keep a record of respondents who refused to discuss product holding or product choice and so the only way we could deal with this systematically for the financial behaviour score was to assume that all non-responses were due to respondents that had not made a recent product choice. 147

Table 16. Testing the behaviour score

Ranking of country according to scoring method used

Behaviour score using 9 points Malaysia Germany Ireland UK Czech Republic Armenia Poland South Africa Hungary Estonia

Behaviour score without budgeting question Germany Malaysia UK Ireland Czech Republic Hungary Armenia South Africa Poland Estonia

Behaviour score without active saving Ireland Germany Malaysia UK Armenia Hungary Czech Republic Poland South Africa Estonia

Behaviour score without any points for product choice Malaysia Germany Ireland UK Czech Republic Armenia Poland South Africa Hungary Estonia

As reported above for knowledge, a simple test has been used to see whether the removal of certain behavioural indicators affects the ranking of countries. The budget question was tested because of concerns that it overlooks individuals who do not budget for their household. The active savings question has been tested due to the relatively high level of missing data, whilst the product choice question has been tested due to the low levels of product purchasing activity in some countries. This indicates that the inclusion of the budgeting question and the active saving measure reduce the relative score of Hungary. The active saving measure was not refused by many people in Hungary, and the reduced ranking reflects the small proportion of individuals who reported that they were saving suggesting that the ranking is appropriate. Other countries are not unduly affected by the behaviour measures tested. Additional tests show that average behaviour scores are higher amongst those who gained a score on the omitted variable than those who did not or refused- another important indicator that the omitted variable is a valid measure. As behaviour is considered to be a key component of financial literacy, and as each of these elements has been identified as a valuable and valid indicator they have all been kept in the final measure. Comparative data should be read with caution, given the impact of missing information on saving. The larger number of questions will help to provide a more nuanced indicator. It should be remembered, however, that it will be less likely that an individual gets a top score in this section, through a greater propensity to refuse, lack of responsibility for household finances or lack of recent product choice. As with the knowledge score, the behaviour score has been rescaled for reporting to take values from 0 to 100.

148

ATTITUDINAL SCORE

Survey questions designed to capture attitudes The Core Questionnaire includes three scaled attitudinal questions that are intended to be used to measure financial literacy (and one that has been included as an explanatory factor). Lessons from the first measurement project People were generally happy to try to answer the attitude statements very few refused, and a small percentage claimed that they did not know what response to give. Changes to the question wording As with the behavioural scale questions, these attitudinal questions were asked in the same way in every country except Norway and South Africa. In Norway the questions were asked on a 7 point, rather than 5 point scale62. In addition, in Norway, just 1 of the three core questions was asked. In South Africa, a 5 point scale was used, but each of the scale points was given a verbal description (completely agree, agree, neither agree nor disagree, disagree, completely disagree). In South Africans were far more likely to put themselves at 2 or 4, rather than 1 or 5 on the scale (something that was not found in other countries). This may reflect the fact that South Africa labelled each of the values. This is something that should be taken into account for future surveys. Changes to the Core Questionnaire In the revised version of the questionnaire, the behavioural statements have been combined with the attitude statements, as discussed in the previous section on behaviours. Developing a score With very low levels of missing data, refusals have been recoded as neither agreeing nor disagreeing (point 3 on the scale). Those who responded dont know have also been given a score at the midpoint as they appear to be ambivalent. The scale-points of the attitude questions have been used as scores (where necessary they have been reversed so that 1 is the least financially literate attitude). The scores from each of the three

62

The Norwegian scale has been recoded as follows for the purpose of comparisons: 1=1 (2,3=2) (4=3) (5, 6=4) (7=5). 149

questions have then be summed, before dividing by 3 to give an average attitude. Anyone achieving more than 3 across the three questions has been identified as having a generally positive attitude63.

63

If 3 is the midpoint, then these people are neutral. In order to have scored above 3 on average, they must have put themselves at 4 on the scale on at least one of the three questions. 150

AN OVERALL MEASURE OF FINANCIAL LITERACY

Segmenting the population Some countries are particularly interested in categorising their population according to their financial literacy. The approach applied by the OECD/INFE considers whether the respondent gained a high score in 0, 1, 2 or 3 of the three components. This has the advantage of creating mutually exclusive categories, and ones that are driven exclusively by responses to the financial literacy questions. The size of the categories is of course related to decisions made about what constitutes a high score, but as the purpose is to compare proportions in each category across groups, countries and times, the most important aspect is that a) they genuinely represent good enough levels of financial literacy and b) they are easily replicable with different data. Developing a single measure of financial literacy In order to assess overall levels of financial literacy the three raw scores discussed above for knowledge, behaviour and attitudes have also been summed64. This gives a simple measure that takes into account the various aspects of financial literacy, including financial planning for the future, choosing financial products and managing money on a day to day basis. The three raw scores have different maximum values, and so the combined score is implicitly weighted. The most heavily weighted factor is behaviour which can take a value between 0 and 9. This is appropriate: behavioural questions make up a large part of the questionnaire because financial behaviour is seen as a key component in financial literacy65. Financial knowledge also makes up a large percentage of the final score (scores range from 0-8). Financial knowledge and behaviour are the two aspects of financial literacy most typically targeted by financial education initiatives. The score also contains a small component of attitudes towards money, and particularly towards planning for the future (the scores for attitude are based on the respondents average position on the scale and range from 1-5). As the attitude score does not go to zero, the raw combined financial literacy score runs from a minimum of 1 to a maximum of 22. This has been rescaled to take values from 1 to 100.

64 65

We have also rounded final values to whole numbers, in order to facilitate reporting. Exploratory factor analysis also suggests that knowledge and behaviour should be weighted more heavily than attitudes. 151

COLLECTING DATA ON SOCIO-DEMOGRAPHICS

It is important to be able to analyse financial literacy data by key demographics, in order to identify target groups for financial education initiatives. Target groups typically reflect age, gender, income and location. Work status may also be used to identify recipients of particular initiatives. Across the participating countries, quite large proportions of the population refused to answer the income question, despite the very broad categories. This means that most analysis by income is only possible on a subset of respondents. However, it has been possible to include an indicator of missing income data in the regressions in case they were systematically different from other respondents. The rural/urban marker also caused difficulties in an international context. Changes have been made to the final questionnaire to try to capture this data. Following feedback about the use of 30 hours as an indication of full time work, the question on work status has been edited, and it is now followed by a question on the number of hours worked. For international comparisons it is recommended to use 30 hours, but this gives countries more detailed data with which to explore financial literacy according to national working hour norms. The questions have also been reordered to make sure that the demographic questions are asked in a logical order (from education through to income), and to try to reduce the number of refusals to the income question by making it shorter, and by noting that it is the last question.

152

ANNEX 2: FURTHER DETAIL OF COUNTRY LEVEL DATA

Table 17. Details of sample Country Albania Survey details Age 18+ Useable sample 1,000 Three step random sample: 1) Primary Select ion Unit 2) Families selected 3) Random selection of adult. Face-to-face interviews Main fieldwork in early 2011 Only using OECD/INFE core questionnaire Weights created based on socio-demographic data. Age 18+ Useable sample 1,545 Stratified sample: 1 stage in urban areas and 2 stages in rural areas. The sample size of each selected areas was defined proportional to population size Face-to-face interviews Main Fieldwork October to November 2010. Only using OECD/INFE core questionnaire Weights created based on socio-demographic data. Age 18+ Useable sample 1,005 (1,047 responded, but some incomplete scripts) Face-to-face interviews OECD/INFE core questionnaire combined with others: total 80 questions Weights created based on socio-demographic data. Piloted prior to full scale survey (main survey September 2010) Report available in Czech http://www.mfcr.cz/cps/rde/xchg/mfcr/xsl/ft_finvzd_vyzkum_gramot.html

Armenia

Czech Republic

Estonia

Age 18+ Useable sample 993 Multi-stage probability random sampling by regions and types of settlement (urban/rural) Face-to-face interviews Main Fieldwork November to December 2010. OECD/INFE core questionnaire combined with 10 others Weights created based on socio-demographic data Age 18+ Useable sample 1,005 Random sample based on the ADM-Basis for CATI samples, all private phone numbers whether or not registered in directories. Telephone interviews 1,441 calls were made Main fieldwork November 2010 until January 2011 Weights created based on socio-demographic data

Germany

153

Country Hungary

Survey details Age 18+ Useable sample 998 (gross sample 3131) Face-to-face interviews Only using OECD/INFE core questionnaire Weights created based on socio-demographic data, individuals in household, number of telephone lines. Age 18+ Useable sample 1,010 Stratified random sampling, sampling points selected proportion to population, random selection of adult in household Face-to-face interviews Piloted prior to full scale survey (main survey October to November 2010) Only using OECD/INFE core questionnaire Weights created based on socio-demographic data Age 18+ Useable sample 1,046 Multi Stage Stratified Random Sampling. Response rate 65% Face-to-face interviews Data collection period September to October 2010 OECD/INFE Questions combined with others on household indebtedness and levels of literacy Weights created based on socio-demographic data Age 18+ Useable sample 2122 Web panel (robustness checks are possible as certain questions have also been added to a telephone omnibus). Weights created based on socio-demographic data Age 18+ Useable sample 2254 Weights created based on socio-demographic data Age 18+ Useable sample 1008 Telephone omnibus survey Questions used as part of a longer survey and also incorporating additional knowledge questions from INFE supplementary questions Fieldwork September to December 2010 Weights created based on socio-demographic data Original sample of Age 16+ reweighted to be a sample of adults aged 18+. Complex stratified multi-stage sample Useable sample 3,112 Face-to-face interviews Fieldwork November to December 2010. Weights created based on socio-demographic data

Ireland

Malaysia

Norway

Peru

Poland

South Africa

154

Country UK

Survey details Age 18+. Useable sample 1579 Fieldwork September 2010. Only using OECD/INFE core questionnaire Weights created based on socio-demographic data Age 18+. Useable sample 535. Sample drawn from a random pool of anonymous telephone numbers: 1489 calls made. Telephone interviews. Only using OECD/INFE core questionnaire Data unweighted; tested for representativeness across key socio-demographics

BVI

Table 18. Financial knowledge: division, time-value of money, interest paid on a loan
Division Dont know Incorrect Time-value of money Dont know Incorrect Interest paid on a loan Dont know Incorrect

Refused

Refused

Refused

Correct

Correct

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway* Peru Poland South Africa United Kingdom BVI

1% 2% 2% 1%

10% 4% 3% 3% 5% 2% 4%

10% 2% 2% 10% 2% 3% 1% 18% 6% 4% 10% 17%

89% 86% 93% 93% 84% 96% 93% 93% 61% 90% 91% 79% 76%

4% 1% 1% 2% 2% 1% 2% 1% 4% -

25% 9% 6% 7% 4% 5% 7% 14% 3% 7% 8% 13% 6%

9% 6% 13% 5% 35% 16% 35% 23% 10% 28% 14% 34% 33%

61% 83% 80% 86% 61% 78% 58% 62% 87% 63% 77% 49% 61% 1% 1% 3% 12% 8% 11% 22% 2% 85% 65% 90% 3% 3% 1% 6% 1% 1% 7% 8% 9% 7% 3% 6% 6% 16% 5% 1% 4% 3% 1% 1% 1% 23% 87% 88% 84% 88% 95% 88% 93% 61%

1% 1% 4% 1% -

5% 20% 3% 1% 10% 8%

14%

2%

84%

1%

16%

9%

74%

1%

Correct 99%

Base: all respondents. Column percentages by country. A dash [ ] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded. *The results reported for Norway under the Division column actually refer to an alternative questions posed: What is the nominal interest rate. Norway also slightly reworded the time value of money question, as they had not asked the previous question. Under interest for Norway responses to: What is meant by the effective interest rate are reported.

155

Table 19. Financial knowledge: interest plus principle, compound interest

Responses to knowledge question (Column percentages by country, weighted data, all respondents) Calculation of interest plus principle
Refused Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI 2% 1% 6% 4% 2% 1% 4% 2% 4% Dont know 45% 14% 20% 20% 17% 18% 15% 29% 3% 20% 2% 28% 19% 27% Incorrect 13% 33% 14% 12% 18% 20% 9% 16% 22% 36% 36% 24% 19% 10% Correct 40% 53% 60% 64% 64% 61% 76% 54% 75% 40% 60% 44% 61% 63%

Compound interest AND Correct response to previous question *

No 90% 82% 68% 69% 53% 54% 71% 70% 46% 86% 73% 79% 63% 80% Yes 10% 18% 32% 31% 47% 46% 29% 30% 54% 14% 27% 21% 37% 20%

Base: all respondents. Column percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

156

Table 20. Financial knowledge: risk and return, inflation, diversification

Risk and return Definition of inflation Diversification

Dont know

Dont know

Dont know

Refused

Refused

Refused

False

False

False

True

True

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway** Peru Poland South Africa United Kingdom BVI**

2% 1% 3% 1% 1% 11% 1% 2% 2% 1% 0% 20% 16% 37% 7% 9% 6% 16% 10% 22% 4% 7% 10%

21% 16% 7% 4% 16% 6% 6% 8% 18% 13% 13% 19% 14% 11%

77% 67% 81% 72% 79% 86% 84% 82% 61% 69% 48% 73% 77% 83%

2% 0% 2% 1% 0% 14% 1% 1% 1% 1% 0% 16% 4% 12% 8% 3% 7% 10% 13% 11% 3% 3% 8%

17% 33% 14% 2% 10% 6% 3% 12% 15% 9% 6% 13% 4% 6%

81% 57% 70% 85% 87% 91% 88% 74% 68% 86% 80% 78% 94% 87%

3% 1% 4% 2% 1% 1% 27% 1% 3% 1% 1% 0% 37% 29% 31% 9% 22% 39% 12% 34% 35% 11% 25% 34%

34% 28% 7% 6% 28% 14% 18% 29% 51% 17% 12% 42% 23% 41%

Base: all respondents. Column percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.* If the respondent did not know how to calculate simple interest over 1 year, it is unlikely that they could calculate 5 times this amount, and therefore it is assumed that a correct answer to the follow-up question is a guess.**For diversification Norway and BVI asked Buying a single companys stock usually provides a safer return than a stock mutual fund: note that in this case False is the correct response.

157

True
63% 59% 54% 57% 60% 61% 47% 43% 11% 51% 55% 48% 55% 20%

Table 21. Behaviour: Before I buy something I carefully consider whether I can afford it
Refused 1% Dont know 1% 1% 4% 1% 4% Never 4% 1% 2% 4% 4% 2% 2% 1% 2% 1% 4% 3% 6% 2% 2 4% 2% 3% 4% 3% 2% 3% 1% 12% 2% 5% 3% 4% 2% 3 4% 6% 18% 21% 11% 10% 12% 6% 13% 6% 20% 10% 13% 5% 4 16% 9% 21% 20% 23% 18% 20% 14% 41% 11% 23% 21% 15% 15% Always 70% 81% 54% 47% 59% 68% 63% 78% 31% 80% 47% 62% 62% 72%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

1% 1%

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

Table 22. Behaviour: I pay my bills on time

Refused 2% 1% 1% 0% 3% 1% 1% 2% Dont know 3% 3% 1% 1% 3% 0% 4% Never 3% 1% 1% 1% 2% 1% 4% 5% 1% 2% 9% 3% 1% 2 6% 1% 2% 2% 2% 3% 5% 10% 1% 4% 6% 1% 2% 3 11% 5% 9% 10% 3% 12% 11% 19% 5% 12% 14% 19% 6% 10% 4 30% 13% 20% 19% 13% 16% 20% 30% 29% 18% 21% 26% 9% 19% Always 48% 81% 65% 64% 83% 66% 64% 39% 50% 68% 57% 35% 80% 64%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

158

Table 23. Behaviour: I keep a close personal watch on my financial affairs

Refused 1% 1% 2% 1% 1% 2% Dont know 1% 2% Never 3% 4% 3% 2% 1% 8% 1% 2% 1% 1% 2% 7% 4% 3% 2 8% 4% 5% 4% 2% 5% 3% 2% 3% 19% 4% 8% 3% 2% 3 17% 11% 15% 14% 9% 15% 10% 15% 6% 2% 12% 17% 12% 11% 4 35% 16% 25% 23% 23% 16% 21% 29% 34% 23% 28% 15% 20% Always 36% 65% 51% 55% 64% 54% 64% 50% 55% 68% 58% 37% 65% 59%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

1% 1% 1% 4%

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

Table 24. Behaviour: I set long term financial goals and strive to achieve them
Refused 4% 2% 1% 2% 1% 1% 1% 2% Dont know 6% 1% 11% 1% 1% 1% 1% 2% 2% 1% 6% Never 15% 15% 18% 12% 11% 13% 12% 7% 4% 9% 14% 12% 22% 5% 2 27% 9% 15% 13% 6% 10% 11% 7% 18% 5% 16% 9% 9% 3% 3 22% 18% 26% 21% 22% 24% 20% 20% 18% 12% 22% 19% 25% 17% 4 19% 15% 18% 19% 25% 21% 23% 34% 39% 16% 21% 24% 15% 23% Always 12% 43% 19% 22% 36% 31% 33% 30% 19% 55% 25% 32% 27% 45%

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

Base: all respondents. Row percentages by country. A dash [ ] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

159

Table 25. Average behaviour score by country

Mean 54.4 55.6 57.8 50.0 65.6 54.4 62.2 66.7 61.1 63.3 55.6 55.6 61.1 67.8 Min 0.0 0.0 0.0 0.0 11.1 0.0 0.0 11.1 0.0 0.0 0.0 0.0 0.0 0.0 Max 100.0 88.9 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 88.9

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI

Base: all respondents.

Table 26. Attitude: I find it more satisfying to spend money than to save it for the long term
Refused Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI 1% 2% 2% 1% 1% Dont know 2% 2% 6% 1% 1% 1% 3% 1% 1% 6% Completely agree 7% 56% 10% 19% 8% 5% 15% 19% 4% 11% 28% 11% 17% 6% 2 10% 15% 14% 14% 12% 8% 20% 16% 17% 8% 17% 26% 12% 8% 3 20% 21% 26% 21% 30% 30% 27% 18% 21% 15% 33% 14% 35% 20% 4 24% 4% 18% 9% 20% 19% 14% 11% 35% 15% 11% 28% 14% 19% Completely disagree 37% 4% 27% 30% 29% 37% 24% 35% 23% 50% 8% 20% 21% 41%

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

160

Table 27. Attitude: I tend to live for today and let tomorrow take care of itself
Refused Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI 1% 1% 1% 1% 1% Dont know 2% 1% 3% Completely agree 3% 19% 8% 16% 8% 6% 12% 10% 11% 19% 7% 15% 9% 2 9% 10% 9% 12% 7% 8% 16% 13% 5% 14% 19% 11% 6% 3 18% 11% 13% 18% 20% 18% 18% 20% 11% 20% 14% 24% 14% 4 25% 7% 18% 14% 19% 16% 19% 16% 15% 21% 35% 16% 21% Completely disagree 41% 53% 51% 35% 46% 52% 35% 41% 58% 24% 25% 34% 44%

1% 1% 1% 6%

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

Table 28. Attitude: Money is there to be spent

Refused
Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI 1% 2% 1% 1% 1% 1% -

Dont know
4% 2% 7% -

Completely agree
15% 74% 14% 28% 21% 12% 14% 21% 19% 38% 14% 21% 21%

2
9% 13% 19% 16% 14% 13% 23% 22% 11% 24% 26% 15% 13%

3
25% 11% 35% 24% 39% 42% 33% 29% 23% 25% 18% 34% 32%

4
21% 1% 13% 9% 12% 14% 12% 11% 14% 8% 22% 12% 14%

Completely disagree
24% 1% 16% 16% 14% 19% 18% 15% 31% 4% 17% 17% 17%

2% 1% 1% 4%

Base: all respondents. Row percentages by country. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

161

Table 29. Average combined attitude scores Average combined score

Albania Armenia Czech Republic Estonia Germany Hungary Ireland Malaysia Norway Peru Poland South Africa United Kingdom BVI 3.7 2.3 3.4 3.1 3.4 3.6 3.2 3.2 3.6 3.7 2.6 3.2 3.1 3.5

Base: all respondents. A dash [] refers to more than 0 but less than 0.5%. Empty cells have no relevant observations, including those where the response was not recorded.

162

Table 30. Financial literacy segments by gender

No Strengths One strength Two strengths Strong on all components All respondents (Unweighted count of non missing data) 410 590 1000 1042 503 1545 532 473 1005 580 413 993 548 457 1005 533 465 998 581 429 1010 423 623 1046 1126 991 2117 798 1456 2254 639 369 1008 1798 1219 3017 840 739 1579 305 230 535

Albania

Armenia

Czech Republic

Estonia

Germany

Hungary

Ireland

Malaysia

Norway

Peru

Poland

South Africa

UK

BVI

Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All Female Male All

19% 8% 13% 33% 29% 32% 16% 15% 15% 19% 15% 17% 10% 6% 8% 7% 11% 9% 14% 13% 13% 11% 10% 10% 12% 12% 12% 11% 10% 10% 30% 22% 26% 21% 23% 22% 19% 14% 17% 10% 5% 8%

38% 35% 36% 44% 39% 42% 26% 28% 27% 40% 44% 42% 27% 28% 27% 31% 26% 29% 26% 32% 29% 32% 31% 32% 30% 29% 30% 28% 28% 28% 39% 38% 38% 44% 33% 39% 36% 30% 33% 25% 18% 22%

33% 41% 38% 20% 28% 24% 32% 33% 33% 30% 31% 30% 34% 32% 33% 36% 41% 38% 37% 35% 36% 33% 37% 35% 41% 36% 38% 44% 41% 42% 26% 30% 28% 25% 28% 27% 30% 34% 32% 33% 40% 36%

11% 16% 14% 2% 4% 3% 26% 24% 25% 11% 10% 10% 29% 34% 32% 25% 22% 24% 23% 20% 22% 25% 22% 23% 17% 23% 20% 18% 22% 20% 6% 11% 8% 10% 15% 13% 15% 22% 19% 32% 36% 34%

Base: all respondents. Row percentages by country

163

Table 31. Financial Literacy Segments by Income

Number of high scores across three components No high scores 1 2 3 All respondents (Unweighted count of non missing data)
380 321 283 984 733 755 31 1519 451 332 139 922 363 375 151 889 249 347 384 980 319 367 141 827 558 176 61 795 431 410 205 1046 349 493 905 1747 822 793 341

Albania

Low Average High Total

19% 12% 5% 12% 46% 20% 23% 32% 21% 11% 7% 15% 19% 16% 11% 16% 10% 9% 4% 8% 14% 8% 6% 10% 15% 5% 6% 12% 16% 8% 4% 10% 11% 10% 9% 10% 13% 8% 7%

44% 39% 26% 37% 39% 45% 29% 42% 29% 26% 21% 26% 41% 42% 46% 42% 40% 26% 19% 28% 35% 25% 16% 28% 32% 22% 20% 28% 39% 31% 17% 32% 32% 28% 26% 28% 31% 25% 20%

33% 39% 40% 37% 15% 31% 46% 24% 32% 35% 34% 33% 31% 32% 26% 31% 29% 35% 34% 33% 36% 41% 37% 38% 37% 42% 36% 38% 32% 34% 44% 35% 43% 41% 36% 39% 40% 45% 42%

4% 10% 29% 14% 1% 4% 3% 3% 19% 28% 38% 25% 8% 10% 17% 11% 20% 30% 44% 32% 15% 26% 41% 24% 17% 31% 38% 22% 14% 27% 36% 23% 14% 21% 28% 24% 16% 22% 31%

Armenia

Low Average High Total

Czech Republic

Low Average High Total

Estonia

Low Average High Total

Germany

Low Average High Total

Hungary

Low Average High Total

Ireland

Low Average High Total

Malaysia

Low Average High Total

Norway

Low Average High Total

Peru

Low Average High

164

Total Poland Low Average High Total South Africa Low Average High Total UK Low Average High Total BVI Low Average High Total

10% 36% 19% 11% 26% 27% 16% 11% 22% 21% 14% 11% 16% 10% 6% 5% 7%

27% 37% 43% 35% 39% 46% 31% 25% 39% 39% 35% 25% 33% 25% 15% 21% 21%

42% 23% 27% 41% 27% 20% 35% 31% 27% 31% 29% 35% 32% 40% 36% 31% 36%

21% 4% 12% 12% 8% 7% 18% 33% 13% 9% 22% 29% 20% 24% 42% 42% 36%

1956 455 314 144 913 1347 1327 137 2811 478 453 444 1375 153 151 153 457

Base: all respondents. Row percentages by country. Caution should be taken when interpreting the results of the high income groups in Armenia and Ireland due to small bases.

165

ANNEX 3: PRODUCT HOLDING BY COUNTRY

Figure 61. Products: Albania Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Savings Account Current Account Insurance product Credit Card Secured Loan Treasury bill Unsecured bank loan Pension fund Microfinance Loan

Figure 62. Products: Armenia Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Retail credit Loan with gold coverage Mortgage Secured bank loan Rural loans Savings Account Credit Card Consumer loans Current Account Pension fund Insurance product Stocks and shares Bonds Overdraft loans

167

Figure 63. Products: Czech Republic Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Current Account Additional Pension Insurance Housing Focused Saving Life Insurance Proprietary Insurance Mortgage Credit Card Savings Account Shares Consumer Lending Debit Card Debentures Investment Fund Income Loss Insurance Charge over Property

Figure 64. Products: Estonia Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Bank Account Insurance Home Loan Pension Fund Credit Card Small Loan Term Deposit Secured Bank Loan Stocks and Shares Unsecured Bank Loan Shares and Investment Fund Bonds

168

Figure 65. Products: Germany Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Savings Account Life-Pension Insurance Current Account P&C Insurance Credit Card Building Society Contract Shares Mortgage Short Term Savings Investment Funds Installment Credit Fixed Interest Securities Savings Plan

Figure 66. Products: Hungary Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Bank account Insurance Credit for housing Private pension fund Personal consumer loan Term deposit Cash loan Credit card Voluntary pension fund Mortgage loan State securities Stocks Investment funds

169

Figure 67. Products: Ireland Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Current account Credit card Mortgage Car insurance Home insurance Credit union savings account Savings account (bank or building society) Private health insurance Post office account Credit union loan Pension (including PRSA-AVC) Life insurance (including mortgage protection) Stocks and shares Bank loan secured on property Hire purchase loan Unsecured bank loan Bonds Investment account Serious illness insurance Consumer hire-leasing Income protection insurance Loan insurance or payment protection insurance

170

Figure 68. Products: Malaysia Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Savings account ATM card EPF Insurance Home loan Car loan Credit card Current account Unit trust: permodalan nasional berhad Fixed deposits Pension funds: gov servants Debit card Unsecured bank loan Stocks, shares and warrants Unit trust: banks or private mutual Bank loan secured on property Overdraft Pensions funds: private Microfinance loan Bonds

Figure 69. Products: Norway

Awareness Holding

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Savings account Credit card Checking account Mortgage Insurance Consumer loan Pension fund Investment products Credit line/housing credit/flexible loan Senior loan

171

Figure 70. Products: Peru Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Savings, salary account, zero account Health insurance ESSALUD Pension fund AFP Credit, business card store Compulsory Insurance: Traffic Accidents Health insurance SIS (comprehensive) Life insurance Pension fund ONP Loan, mortgage Personal loan Health insurance EPS, insurance companies Fixed term deposit Current account, money order Working capital loan Personal accident insurance Vehicle insurance School insurance Cooperative savings account Commercial loan, SME Car loan Collective funds Supplementary insurance for hazardous work Cooperatives loan Loan, credit for motorcycle taxi, motorbike Investment account, mutual funds Investments in the stock exchange Insurance on debt Loan for fixed assets Collateral loan (not on housing) Leasing

172

Figure 71. Products: Poland

Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Current account Credit card Pension fund
Insurance

Mortgage Saving account Payment cards (not credit cards) Bonds Stocks and shares Bank loan secured on property Unsecured bank loan
Investment account, such as a unit trust

Microfinance loan Structured products

Figure 72. Products: South Africa

Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Bank account Credit card

Post office savings

Policies (sanlam, Old Mutual) Insurance (car and house, funeral) Stokvels Pension fund House bond Investment account Bank loan secured on a property Microfinance loan

Stocks and shares

Unsecured bank loan Retail bonds

173

Figure 73. Products: UK

Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Current account Credit card Life insurance Savings account, other than a cash ISA Mortgage General insurance Health or medical insurance Cash ISA Pension fund, company or personal Stocks and shares Bonds e.g. gilts and corporate bonds Loan secured on property Unsecured loan Investment account e.g. unit trust, Insurance to protect income or mortgage Microfinance loan

Figure 74. Products: British Virgin Islands

Awareness Holding Recent purchase

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Insurance Credit card Saving account Pension fund Current account Mortgage Stocks and shares Bonds Bank loan secured on property Unsecured bank loan Investment account Microfinance loan

174

ANNEX 4: ANALYSIS BY SUBGROUPS OF WOMEN

Table 32. Financial knowledge across subgroups of women

Germany Malaysia Czech Republic Armenia Hungary Norway Albania Estonia Ireland Poland

South Africa

Peru

Marital status Married (ref.) Living Coeff 15.04 with a . 4 partner Signif Divorced 4.501 Single 2.267 Widow -7.051 * Age 30-59 (ref.) 18-29 3.889 60+

0.000 *** 6.526 2.753 2.526

9.458 ** 1.600 4.252 6.533 **

3.486

-4.557

-4.128

-2.701

0.402 6.393

-2.430

0.816 0.929

2.844

0.154

-0.246

1.506

-3.162

-0.112

-1.783

2.885 2.724

-0.750

-1.647

-3.953

-4.405 *

-1.354 19.381 *** 9.281 *** 0.523 1.563

4.212 ** 1.418

1.218

2.168 4.338 *

-6.689 *

-4.109

-4.106

-5.699 *

-3.085

0.621

1.592

3.335

3.295

-5.719 *

2.110

-9.093 ***

-2.613

6.206 ***

2.042

3.384 ** 2.864

5.507 **

-4.119 2.941 7.355 -0.718 Working status (not working as reference category ) Working 6.218 2.619 2.914 -1.910 ** Household income Median (ref.) Below median 1.646 5.461 *** Above median 5.515 Education Secondary (ref.) Below secondar -5.845 y ** Above secondar y 7.415 7.779 9.151 **

-0.462

-1.932

1.882

-8.381

5.345

0.164

1.404

-0.308

2.976

1.658

5.803 ***

3.586 *

2.426

6.492 **

4.099 *** 8.793 ***

6.037 *** 4.852 **

-3.467

6.607 **

-2.162

-0.695

-1.434

-6.683 ***

1.699

7.409 *** 0.324 8.243 ***

0.810

-4.759 *

-4.206

4.048 7.922 ***

1.070 12.774 ***

12.065 13.747

1.192 10.425 ***

-0.445 12.791 ***

2.656 14.191 ***

1.981 7.919 **

6.063 11.562 ***

0.502 7.111 *** 0.371

0.635 8.938 ***

-7.637 **

5.906

3.260

3.748

9.588

5.190

2.481

9.285

5.577

6.592

4.845

5.715

Notes: ordinary least squares regressions. Coefficients: Empty cells: the explanatory variable is not available for this country (e.g. marital status not available for Norway)/ no observations in the cell (e.g. the category living with partner was not coded in the Armenian survey). Dummies for refusals/DK were included in the regressions, but not reported in the table. Omitted categories: marital status: married is the omitted category for marital status; age: 30-59 omitted; working status: not working was omitted; household income: median omitted; education: complete secondary school omitted. Significance levels: *** 1%; ** 5%; * 10%. The financial knowledge score is based on 8 questions and has been rescaled on a 0-100 scale.

175

BVI 14.85 2 ** 14.40 7 *** 3.835 12.39 6 *** 0.204 0.543 6.439

UK

The Russia Financial Literacy and Education Trust Fund was established in 2008 at the World Bank with funding provided by the Ministry of Finance of the Russian Federation. The work supported by the Trust Fund is jointly managed by the World Bank and the Organisation for Economic Co-operation and Development (OECD) and is directed toward improving public policies and programs to enhance nancial knowledge and capabilities in low- and middleincome countries. This effort has focused on the review of national strategies for nancial education, the development of methods for the measurement of nancial knowledge and capabilities, methods for evaluating the impact and outcome of programs, and research applying these methods to programs in developing countries. The products of this program of work can be found at the Trust Fund website at:

www.nlitedu.org

MINISTRY OF FINANCE OF THE RUSSIAN FEDERATION