Journal of Macroeconomics 34 (2012) 1007–1019

Contents lists available at SciVerse ScienceDirect

Journal of Macroeconomics
journal homepage: www.elsevier.com/locate/jmacro

Banking industry volatility and growth

Pei-Chien Lin a,⇑, Ho-Chuan (River) Huang b
a
Department of Industrial Economics, Tamkang University, Taiwan
b
Department of Banking and Finance, Tamkang University, Taiwan

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, we provide evidence that banking industry volatility may exert a negative
Received 10 August 2011 impact on growth in a more economically integrated world. By applying the augmented
Accepted 13 August 2012 difference-in-difference framework of Rajan and Zingales (1998) to the cross-country
Available online 5 September 2012
cross-industry data developed by Ciccone and Papaioannou (2009), complemented by
the Financial Development and Structure database of Beck et al. (2010), we show that over
JEL classification: the 1980–1999 period the banking sector volatility, measured as the standard deviation of
D31
the growth of private credit, has a negative impact on the growth of industries that are
O40
more externally financially dependent, and this finding is robust to various sensitivity
Keywords: tests. However, the detrimental growth effect of banking sector volatility disappears when
Banking industry volatility the sample is restricted to the relatively placid 1980s. Compared to the 1980s, the 1990s
Bank development are characterized by a more economically integrated world accompanied by more often
Growth unpredicted financial crises that disturb the banking sector. As such, our results imply that
in a more economically integrated world, the stability of bank development may be impor-
tant to long-run growth.
Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction

This paper aims to empirically investigate the impact of banking industry volatility on long-run economic growth. Follow-
ing the pioneering work of King and Levine (1993a,b), an enormous literature has been devoted to investigating the finance-
growth nexus.1 By and large, this line of research generally concludes that the depth of financial markets and institutions,
usually measured by the size of private credit relative to GDP, has a positive and significant effect on long-run growth. Further
effort has also been made toward establishing that the causality goes effectively from financial to economic growth, but not the
other way around, by analyzing industry- and firm-level data to clarify the mechanisms that are somewhat obscured in cross-
country studies (Demirgüç-Kunt and Maksimovic, 2002; Rajan and Zingales, 1998). However, along the process of financial
development, which entails a deepening of markets and services that channel savings to productive investments, allow risk
diversification and thus possibly lead to higher economic growth in the long run, the same process can also present weaknesses
as evidenced by systemic banking crises, cycles of booms and busts, and overall financial volatility. Whether intrinsic to the
process of development or induced by policy mistakes or external shocks, these elements of financial volatility (or fragility)
can hurt economic growth (Loayza and Rancière, 2006). Nevertheless, this speculation is barely examined in the finance-growth
nexus literature.

⇑ Corresponding author. Address: Department of Industrial Economics, Tamkang University, 151 Ying-Zhuan Road, Danshui Dist., New Taipei City 25137,
Taiwan. Tel.: +886 2 26215656x3340; fax: +886 2 26209731.
E-mail address: pclin@mail.tku.edu.tw (P.-C. Lin).
1
Please see Levine (2005) and Ang (2008) for complete surveys on this issue.

0164-0704/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jmacro.2012.08.004
1008 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

On the other hand, in the last few decades, volatility has become an independent field of inquiry in macroeconomics,
moving on from a second-order research area to currently ‘occupy a central position in development economics’ (Aizenman
and Pinto, 2005). What brings volatility into this prominence are the cross-country studies that follow the seminal paper of
Ramey and Ramey (1995) in examining the negative impact of volatility on growth, and the growth literature that includes
volatility based on the endogenous growth theory. As to the role of the financial sector in this stream of the literature, recent
theoretical works tend to illustrate the negative relationship between volatility and growth via the channel of financial
frictions, e.g., Kharroubi (2007), Aysan (2007) and Aghion et al. (2010), just to name a few.2 Empirically, most of the previous
studies that investigate the negative relationship between volatility and growth have focused on debating the relative
importance of macroeconomic policy volatility versus institutional uncertainty to the process of economic development, e.g.,
Aizenman and Marion (1993, 1999), Brunetti and Weder (1998), Acemoglu et al. (2003), Campos and Nugent (2003), Easterly
(2005), Hnatkovska and Loayza (2005), and Loayza et al. (2007), just to mention a few. By contrast, recent efforts have been
directed toward understanding the factors, such as a well-developed financial sector and institutional strength, that help
mitigate the negative effect of volatility, e.g., Aghion et al. (2009), Aghion et al. (2010) and Fatás and Mihov (2006). However,
since the financial system may even propagate economic variability or create risk of its own, rather than treating the financial
sector as a channel for mediating with other volatility, it would be more interesting to investigate the direct impact of volatility
that originated from the financial sector on economic growth.
However, through what mechanism can the volatility in the financial sector affect real economic activity? Based on an
asymmetric information theory of financial instability, Mishkin (1998) summarizes four categories of fundamental factors
that lead to financial volatility (or instability): an increase in interest rates, increases in uncertainty due to recession or polit-
ical instability, the asset market effect on nonfinancial firms’ balance sheets, and problems in the banking sector. Generally
speaking, asymmetric information between lenders and borrowers leads to two basic problems in the financial system:
adverse selection and moral hazard. Specifically, rising interest rates and uncertainty will worsen the problems of adverse
selection as well as moral hazard and consequently drive up the possibility of lending leading to bad credit. Under such
circumstances, lenders will want to make fewer loans, possibly leading to a sharp decline in lending that will result in a sub-
stantial decline in investment and aggregate economic activity. On the other hand, the decline in net worth as a result of a
stock market decline makes lenders less willing to lend because the net worth of firms serves a similar role to collateral, and
when the value of the collateral declines, it provides less protection to lenders so that losses from loans are likely to be more
severe.3 As to the problems in the banking sector, banks have a very important role to play in financial markets since they are
well suited to engage in information-producing activities that facilitate productive investment for the economy. Thus, a decline
in the ability of banks to engage in financial intermediaries and make loans will lead directly to a decline in investment and
aggregate economic activity.
From the above analysis, one can see that financial instability occurs when shocks to the financial system interfere with
information flows so that the financial system can no longer do its job of channeling funds to those with productive invest-
ment opportunities and hence is likely to result in a contraction of output. As such, in this paper, we will revisit the question
about the role of the financial sector in economic growth by extending the set of variables that characterize the process of
financial development. In contrast to most of the previous empirical studies that only use the size of the banking system,
such as the ratio of private credit to GDP, as a regressor to predict growth performance, we additionally consider the vola-
tility in the banking sector, measured by the standard deviation of the growth of private credit, and argue that it can also be a
key characteristic of the financial sector that matters for long-term economic performance, especially in a more economically
integrated world. Moreover, while much of the existing research relies on cross-country analysis to identify volatility as an
impediment to growth, it is of interest to present complementary evidence that is derived from more disaggregated data. In
addition, as the financial system may even propagate economic variability or create risk of its own, it would be more rea-
sonable to treat the financial sector as the source of volatility rather than a channel for mediating with other volatility.
For the reasons presented above, in this study, we particularly focus on the impacts of banking sector volatility on growth
by employing the difference-in-difference approach of Rajan and Zingales (1998, thereafter RZ) to cross-country and cross-
industry data, which are mainly obtained from Ciccone and Papaioannou (2009) and complemented by the Financial
Development and Structure Database constructed and recently updated by Beck et al. (2010).
Our empirical analysis for the 1980-1999 period reveals that when adding the interaction of banking industry volatility
and industrial external finance to the original RZ specification, the positive effect of bank development on sectoral growth via

2
Specifically, Kharroubi (2007) illustrates that normal volatility and abnormal volatility, which result from liquidity crises that are due to maturity
mismatches between assets and liabilities, have independent negative effects on the average growth rate. Moreover, Aysan (2007) employs a two-period
overlapping generations model with two types of technologies to predict that greater volatility induces financial intermediaries to charge higher interest rates,
and therefore increases the cost of borrowing associated with capital market imperfections, thus deterring people from obtaining and using more productive
technologies, which is detrimental to growth. Aghion et al. (2010) develop a model in which a novel propagation mechanism emerges such that tighter credit
affects the cyclical composition of investment and thus can lead to both higher volatility and lower mean growth.
3
Alternatively, according to the ‘financial accelerator framework’ developed by Bernanke et al. (1999), the pro-cyclicality of credit allocation amplifies small
shocks occurring in the real sectors of the economy. Specifically, agency problems make firms’ external funding costs depend on their collateral values. As such,
when the economy is booming, firms appear to be less risky with stronger balance sheets, and thus can borrow from banks with a lower external finance
premium, thereby inducing firms to make more new investments that contribute to economic growth. This mechanism which works in good times also works
in reverse in bad times. If uncertainty increases, banks respond by raising the average external finance premium because they expect more firms to go bankrupt.
Consequently, the higher average external finance premium causes the average level of investments to decline, and can thus induce lower economic growth.
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1009

the channel of industry external finance is attenuated. However, volatility in the banking system does exert a significantly
negative impact on the growth of those industries that are more externally financially dependent. Moreover, this result is
robust to various determinants of growth controlled, different measures of stock market development considered, alterna-
tive measures of financial volatility employed, and other macroeconomic volatilities incorporated. However, as the sample is
restricted to the relatively placid 1980s, we find that the negative impact of banking industry volatility is not statistically
significant, while bank development continues to exert a positive growth effect as suggested by the previous literature.
How do we reconcile these seemingly inconsistent outcomes? Compared to the 1980s, the 1990s were characterized by
increasing international capital flows, especially for those capital flows to developing countries, thus making the economies
around the world more integrated and also more often accompanied by unpredicted financial crises which made the global
financial markets more turbulent. As a result, the major message emerging from our research is that as the world is becom-
ing more and more economically integrated, the stable development of the banking sector may be important to determine
long-run economic growth. This finding complements the earlier literature in two respects. First, it extends the literature on
the finance-growth nexus by additionally considering the volatility of the banking sector as an important factor in determin-
ing long-run economic growth. Second, it contributes to the volatility-growth literature by confirming the possible
detrimental effect of banking industry volatility on growth using a cross-country cross-industry data set.
The rest of this paper is structured as follows. Section 2 presents our empirical models, which extend from the difference-
in-difference methodology introduced by Rajan and Zingales (1998), to assess the causal effect of banking sector volatility on
growth. In particular, we add an interaction of a measure of industry-level external financial dependence and a country-level
indicator of banking industry volatility to an otherwise standard RZ specification. Section 3 explains the data sources and the
construction of relevant variables. Section 4 presents our main results. Section 5 provides some robustness checks. Lastly,
Section 6 concludes.

2. Empirical strategy

To examine whether volatility in the banking system exerts a negative impact on the growth of industry output via the
channel of external financial dependence, we employ the difference-in-difference approach of Rajan and Zingales (1998). In
their seminal paper, Rajan and Zingales (1998) use this framework to assess if industries with higher dependence on external
finance grow faster in countries with better development in financial institutions and markets. Hence, their framework has
suggested that one way to check whether a channel is at work is to see whether industries that might be most affected by a
channel grow differentially (more quickly or more slowly depending on the nature of the effect) in countries where that
channel is likely to be more operative. Accordingly, the concern with the causality problem can be greatly mitigated through
this mechanism.4
Since we are interested in the differential effects of financial volatility on the growth of industries in different countries,
based on the RZ framework, we add an interaction term between a measure of banking sector volatility and a measure of
external financial dependence to examine whether volatility in the banking system has a disproportionately larger effect
on the growth of industries that are relatively more dependent on external finance. As such, our benchmark regression is
specified as follows:

Dlnyi;k;19801999 ¼ a0 þ a1 Industryi þ a2 Countryk þ b1 ðExternal Financei  Bank Dev elopmentk Þ

þ b2 ðExternal Financei  Volatility of Bank Dev elopmentk Þ þ b3 lnyi;k;1980 þ b4 Other Controlsi;k
þ i;k ð1Þ

where the subscripts i and k denote the ith industry and the kth country, respectively. The dependent variable Dlnyi;k;19801999
is the annualized growth rate of value added for industry i in country k during the 1980-1999 period. Moreover, industry and
country dummies are included to capture the industry- and country-specific effects, respectively. By including these dummy
variables, the problems of omitted variable bias or model specification, which are common to cross-country regressions, can
be significantly mitigated.
Other main explanatory variables include an interaction of External Financei (the measure of external financial depen-
dence of the ith industry) and Bank Dev elopmentk (the indicator of development in the banking industry of the kth country),
and the interaction between External Financei and Volatility of Bank Dev elopmentk (the indicator of banking industry volatil-
ity of the kth country). The former is included to examine whether development in the banking sector exerts a causal effect
in stimulating industrial growth in value-added through the role of external financial dependence, while the latter, which is
our main interest, is employed to test whether sectors that are more externally financially dependent are more vulnerable in
countries where the banking sector is more volatile. As such, we are particularly interested in the sign and statistical signif-
icance of the parameter b2 . In addition, we also consider the ‘catch-up’ effect, as implied by neoclassical growth theory, by
including the initial conditions, i.e., the country-industry (log) value added in 1980 (lnyi;k;1980 ) correspondingly, in the model.

4
Recent papers that employ Rajan and Zingales (1998) approach include Cetorelli and Gambera (2001), Beck and Levine (2002), Claessens and Laeven (2003,
2005), Braun and Larrain (2005), Larrain (2006), Kroszner et al. (2007), Beck et al. (2008), Gupta and Yuan (2009) and Levchenko et al. (2009), to name a few.
1010 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

We also control for other determinants of industry growth, which is suggested by the previous literature and indicated as
Other Controls in Eq. (1), to check the robustness of our results. First of all, we examine whether the negative impact of finan-
cial volatility on growth in industries with higher external financial dependence is sensitive to control for the role of human
capital as suggested by Ciccone and Papaioannou (2009). As such, we augment the specification by an interaction between a
measure of industry-level human capital intensity and a country-level of human capital. Secondly, as the important role of
capital stock on growth has been emphasized by the neoclassical growth theory, we also control for this possibility by adding
an interaction between industry capital intensity and the country-level capital-output ratio. Finally, empirical studies by
Levchenko (2007) and Nunn (2007) show that countries with good contract enforcement (property rights) institutions
specialize in complex, contract-intensive industries. As such, we add an interaction between Nunn’s measure of industry
contract intensity and a country-level rule of law index. Finally, i;k is the error term.
Moreover, the recent literature has also assessed the comparative role of financial structure, i.e., a credit-market (debt)
oriented versus a capital-market (equity) oriented financial system, in economic growth, e.g., Beck and Levine (2002,
2004), Demirgüç-Kunt and Maksimovic (2002), and Levine (2002), just to name a few of the studies. Based on this line of
the literature, we extend our benchmark model by including two additional interaction terms, i.e., External Finance interact-
ing respectively with Stock Market Development and Volatility of Stock Market Development, to account for the impact of stock
market development on the growth of industries with higher external financial dependence. Specifically, the regression is
specified as:

Dlnyi;k;19801999 ¼ a0 þ a1 Industryi þ a2 Countryk þ b1 ðExternal Financei  Bank Dev elopmentk Þ

þ b2 ðExternal Financei  Volatility of Bank Dev elopmentk Þ þ b3 ðExternal Financei
 Stock Market Dev elopmentk Þ þ b4 ðExternal Financei
 Volatility of Stock Market Dev elopmentk Þ þ b5 lnyi;k;1980 þ b6 Other Controlsi;k þ i;k ð2Þ

b3 and b4 respectively represent the impact of the stock market, either measured by size or by its volatility, on the growth of
industries that are more externally financially dependent. By this augmented regression, we can further test for the robust-
ness of our finding obtained from the benchmark regression.

3. Data sources and description

The data used in this study are mainly obtained from Ciccone and Papaioannou (2009), complemented by the Financial
Development and Structure Database constructed and recently updated by Beck et al. (2010). For illustrative purposes, the
relevant variables are categorized into country-industry-specific, industry-specific and country-specific, respectively. In
what follows we will describe these variables in more detail.

3.1. Country-industry-specific variables

In our benchmark specifications, the dependent variable is the annual growth rate of real value added for 28 manufac-
turing industries at the three-digit ISIC level for a large number of countries over the 1980–1999 period. To maintain the
quality of data, we follow Ciccone and Papaioannou (2009) to restrict our sample to countries with data for more than
ten industries, and also require at least five years of data in the 1980s and in the 1990s for each country-industry. Moreover,
the United States is excluded from the sample since it is used as the industry benchmark. In this manner, we turn out to have
a sample of 44 countries and 1084 observations for industry value-added growth. In addition, the initial condition, measured
as the logarithm of country-industry real value added in 1980, is also included to capture the ‘catch-up’ effect as implied by
the neoclassical growth model.

3.2. Industry-specific variables

Due to the limited availability of industry data for most countries, the variables at the industry level are usually con-
structed from the industry data of a benchmark country, which is usually the US due to the detail and quality of US statistics.
First of all, to assess the impact of the banking sector on industrial growth, an important channel is through industry reliance
on external finance. In this study, the series of external financial dependence (External Financial Dependence), defined
as the difference between industry investment and industry cash flow relative to industry investment, is originally con-
structed by Laeven et al. (2002) using US data from COMPUSTAT at the three-digit ISIC level for the 1980–1989 period.
Moreover, other industry characteristics are also further controlled in our empirical analysis. The first indicator is human
capital intensity (Human Capital Intensity), defined as the average years of schooling of each corresponding three-digit
industry in 1980 and constructed by Ciccone and Papaioannou (2009) based on US data. The second indicator is physical cap-
ital intensity (Capital Intensity) measured as the ratio of industry capital stock to industrial value added from the NBER
Manufacturing Database (Bartelsman and Gray, 1996). The third measure is industry contract intensity (Contract Inten-
sity) which is constructed by Nunn (2007) using US input-output tables. Specifically, contract intensity is defined as the
cost-weighted proportion of an industry’s inputs that are highly differentiated and can therefore be expected to require
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1011

relationship-specific investments in the production process.5 The last measure for the industrial characteristic is industrial
R&D intensity which is measured by direct R&D expenditure as a percentage of gross output for each corresponding US industry
over the 1991–1997 period, and the source of the data is the ANBERD and STAN databases of OECD.

3.3. Country-specific variables

The depth of country-level bank development (i.e., the size of the banking system) is measured by the domestic credit to
the private sector relative to GDP in 1980 (Private Credit). Correspondingly, banking sector volatility is measured by the
standard deviation of the growth of the private credit by deposit money banks and other financial institutions relative to GDP
(Volatility of Private Credit) over the 1980–1999 period.6 For the robustness check, we additionally consider three
alternative measures of financial volatility over the 1980–1999 period. The first one is the high-low range of the growth of
financial development over the 1980–1999 period as in Ang (2011a). The second alternative measure is the standard deviation
of the growth of deposit money bank assets over GDP (Bank Asset (A)). The last alternative measure is the standard deviation
of the growth of deposit money bank assets over the sum of deposit money banks’ and the central bank’s assets (Bank Asset
(B)).
To check for the robustness of the impact of financial volatility on the industrial growth of output and employment via the
channel of external financial dependence, we further consider adding three alternative indicators of stock market develop-
ment to our analysis. The first is the stock market capitalization to GDP (Stock Market Capitalization), which is
measured as the ratio of the value of listed shares to GDP, and is a measure of the relative size of the stock markets to
the economy. The second one is the Stock Market Turnover Ratio, which is measured as the ratio of the value of total
shares traded to market capitalization and usually serves as an indicator of the liquidity (or activity) of a stock market
relative to its size. The third one is the stock market total value traded to GDP (Stock Market Total Value Traded), which
is the ratio of total shares traded on the stock exchange to GDP, and is another measure of the degree of liquidity that stock
markets provide to the economy. The source of the data used to construct these financial variables is the Financial Develop-
ment and Structure Database of Beck et al. (2010).
Furthermore, other country-level characteristics, which interact with the previously mentioned industry characteristics,
are also controlled for in our empirical model. These include the country-level human capital (Human Capital Level),
which is measured as the average years of schooling of the population aged 25 and over in 1980, obtained from the latest
update of the Barro and Lee (2011) database, the index of property rights protection (Rule of Law Index), which is proxied
by the index of the rule of law from Knack and Keefer (1995) and IRIS on a scale of 0 to 6 in 1984 (with higher values denot-
ing higher protection), and the physical capital-GDP ratio (Capital-Output Ratio) in 1980.
In addition, as recent studies have highlighted that the growth experience of both developed and developing countries
can be explained by an R&D-driven growth model, we thus increase the robustness of our result by controlling for innovative
activity in the regressions. In line with Ha and Howitt (2007) and Ang and Madsen (2011), the following measures of R&D
activity (in intensive terms) are used: (1) the ratio of real R&D expenditure to real GDP, (2) the share of R&D workers in total
labor employment, and (3) the share of R&D workers in total labor employment adjusted by human capital per worker. 7 The
cross country data for R&D expenditure over GDP and the share of R&D workers in total employment are obtained from Leder-
man and Saenz (2005), while the human capital per worker is measured as the average years of schooling of the population aged
25 and over in 1980 from Barro and Lee (2011) as previously mentioned.
Lastly, we also interact the External Financial Dependence variable with other macroeconomic variables as well as
their volatilities to check whether the previously tested detrimental effect of banking sector volatility on industrial growth
still holds. Specifically, we consider: (1) the level of economic development (Koren and Tenreyro, 2007), which is measured
by the (logarithm of) real GDP per capita, (2) the government size (Fatás and Mihov, 2001; Andrés et al., 2008), which is mea-
sured by the ratio of government spending over GDP, (3) the trade openness (di Giovanni and Levchenko, 2009), which is
measured as the ratio of imports plus exports divided by GDP, (4) the inflation rate, and (5) the degree of financial liberal-
ization (Ang, 2011b), which is measured by an index capturing several types of policy changes in the financial environment.
The levels of these macroeconomic variables are measured in 1980 and are taken directly from the 2009 World Bank World
Development Indicators Database, while the measure of financial liberalization is obtained from the data recently compiled
by Abiad et al. (2010). Moreover, the corresponding volatilities of these variables are measured by the standard deviations of
their respective growth over the 1980–1999 period.
Table 1 provides summary statistics for some of the relevant variables, and Table 2 displays detailed (country-level)
figures of the preferred measures of bank development (Private Credit) and the corresponding measure of volatility.

5
For details about the construction of these variables, please refer to the data appendix of Ciccone and Papaioannou (2009).
6
In general, volatility is an oscillation around a particular economic variable and is usually designed to capture the variability of macroeconomic fluctuations
around a constant mean or a time-varying trend, with the former usually measured by a standard deviation of that economic variable over some historical
period, and the latter, following the real business cycle literature, consisting of the standard deviation of the gap between the realized values and the trend of
that economic variable. Other more subtle measurements of volatility include using GARCH models to construct smooth (time-varying) measures of volatility
or calculating the standard deviation of the residuals obtained from a regression of prediction. Due to the fact that our data are short in terms of the time
dimension, we employ the first concept to measure volatility.
7
Specifically, if we let N denote the number of R&D workers, L represent total labor employment, and h denote human capital per worker, then the share of
R&D workers in total labor employment adjusted by human capital per worker is measured as N=hL.
1012 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

Table 1
Summary statistics.

Country-industry level variables Mean Std. Dev. Max Min Obs.

Value Added Growthi,k 0.0181 0.0530 0.1978 0.3729 1084
Initial Value of Value Addedi,k 18.7098 2.3185 24.3937 10.6375 1120
Industry level variables
External Financial Dependencei 0.2702 0.3499 1.1400 0.4500 1131
Human Capital Intensityi 11.6165 0.8290 13.2040 10.1380 1131
Capital Intensityi 1.4375 0.6996 3.1935 0.4433 1131
Contract intensityi 0.4947 0.1987 0.8587 0.0577 1131
Country level variables
Private Creditk 0.4771 0.2592 1.3231 0.0884 1068
Volatility of Private Creditk 0.1123 0.0882 0.6043 0.0203 1103
Human Capital Levelk 5.8525 2.4786 11.5862 1.2410 1131
Capital-Output Ratiok 1.6079 0.6372 2.9000 0.5400 1070
Rule of Law Indexk 3.7470 1.5629 6.0000 1.0000 1107

Note:The level of bank development is measured by the size of the private credit in 1980. The volatility of bank development is calculated by the standard
deviation of the growth of the private credit by deposit money banks and other financial institutions over GDP during the 1980–1999 period, and the data
set is taken from Beck et al. (2010). All other variables are taken from Ciccone and Papaioannou (2009).

Table 2
Indicators of financial development and volatility.

Country Private credit Volatility of private credit Country Private credit Volatility of private credit
Australia 0.2569 0.0593 Jordan 0.5097 0.0704
Austria 0.7336 0.0210 Japan 1.3231 0.0203
Belgium 0.2964 0.1244 Kenya 0.2948 0.0696
Bolivia 0.1709 0.1771 Korea, Republic of 0.5072 0.0504
Canada 0.6691 0.1762 Kuwait 0.3765 0.1887
Chile 0.4685 0.1154 Sri Lanka 0.1717 0.2314
China 0.5344 0.1339 Morocco 0.2703 0.2026
Cameroon 0.2954 0.0927 Mexico 0.1938 0.1828
Costa Rica 0.2789 0.0511 Malta 0.2996 0.0578
Cyprus 0.6148 0.0689 Malaysia 0.4901 0.1119
Ecuador 0.2253 0.1557 Netherlands 0.9025 0.1182
Spain 0.7649 0.0457 Norway 0.5120 0.0618
Finland 0.4756 0.0834 Panama 0.5811 0.0855
France 1.0213 0.0337 Portugal 0.7316 0.0942
United Kingdom 0.2763 0.0964 Senegal 0.4273 0.0976
Indonesia 0.0884 0.1531 Singapore 0.8104 0.0514
India 0.2399 0.0368 Sweden 0.7547 0.0615
Ireland 0.4235 0.0384 Trinidad and Tobago 0.2872 0.0973
Iran 0.4375 0.0939 Turkey 0.1359 0.1739
Israel 0.7081 0.0625 Uruguay 0.3724 0.1759
Italy 0.5599 0.0434 South Africa 0.5564 0.1047

Note:The level and volatility of banks’ development are respectively measured by the ratio of domestic private credit to GDP in 1980, and the standard
deviation of the growth of the private credit by deposit money banks and other financial institutions over GDP during the 1980-1999 period. The data set is
taken from Beck et al. (2010).

4. Empirical results

4.1. Benchmark results

Table 3 reports the unconditional and conditional estimates for the effects of bank development and banking industry
volatility on annual growth in more compared to less external financially dependent industries. Columns (1) and (2) present
the unconditional results. These results are obtained by regressing the average growth of sectoral value added on the inter-
action term of the industrial external financial dependence (External Financial Dependence) and country-level bank
development (Private Credit), and the interaction term of the industrial external financial dependence and the coun-
try-level volatility of bank development (Volatility of Private Credit), controlled only for the respective initial con-
ditions.8 Country and industry dummies are also included to control for unobserved country-specific and industry-specific

8
Industrial employment growth is also used as an alternative measure of industrial growth. As the results from regressing employment growth on other
determinants of growth are very similar to those from the set using industrial value-added growth as the dependent variable, to save some space, we do not
report the outcomes for industrial employment growth here, but the results are available upon request.
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1013

Table 3
Benchmark results.

Note: The dependent variables are the growth of real value added during the 1980-1999 period for each 3-digit ISIC industry in each country. External
Financial Dependence is the difference between industry investment and industry cash flow relative to industry investment. The level and volatility of
bank development are respectively measured by the ratio of domestic private credit to GDP in 1980, and the standard deviation of the growth of the private
credit by deposit money banks and other financial institutions over GDP during the 1980-1999 period. The human capital interaction is country-level
average years of schooling aged 25 and over in 1980 multiplied by industry-level schooling intensity. The physical capital interaction is the product of
industry physical capital intensity and the country-level physical capital to output ratio in 1980. The rule of law interaction is the product of industry
contract intensity and a country-level measure of the rule of law in 1984. Differential growth measures the difference in sectoral growth between an industry
at the 75th percentile level of external financial dependence with respect to an industry at the 25th percentile level when it is located in a country at the
75th percentile of the volatility of bank development rather than in one at the 25th percentile. Country and industry fixed effects are included in all
regressions, but not reported. All results are obtained from the OLS procedure and the heteroskedasticity-robust standard errors are reported in the
parentheses.
⁄⁄⁄
significance at the 1 % level.
⁄⁄
significance at the 5% level.
⁄
significance at the 10% level.

effects. All these regressions are estimated by ordinary least squares (OLS), complemented by endogeneity tests for the vari-
ables of bank development and volatility in the banking sector to provide statistical guidance for the requirement to correct
for the problem of endogeneity using the instrumental variable technique. 9 First of all, for comparison purposes, we replicate
the work of Rajan and Zingales (1998) by only considering the interaction of External Financial Dependence and Private
Credit to investigate whether industries that are more external financially dependent grow disproportionately faster in
countries with better developed banking systems. By focusing on column (1) of Table 3, the estimated coefficient for the
financial interaction term is 0.0325 and statistically significant at the 5% level, and we thus verify the results of Rajan and
Zingales (1998).
We next examine the differential growth effect of volatility in the banking sector on industrial value-added growth, which
is our main interest, by adding one more interaction of External Financial Dependence and Volatility of Private
Credit to the RZ’s specification. As can be seen in column (2) of Table 3, after including this additional interaction term, the
positive growth impact of bank development on industrial output growth is not statistically significant. Instead, the esti-
mated coefficient for the interaction term of External Financial Dependence and Volatility of Private Credit
is negative (0.2410) and statistically significant at the 1% level, thereby indicating that volatility in the banking sector is
detrimental to industrial output growth over the study period, particularly for those industries that are more dependent
on external finance. Moreover, the row of Differential Growth in Table 3 reports the economic magnitude of this detrimental
effect caused by the volatility of the banking sector. It presents an annual growth differential of 0.59% between the industry
at the 75th percentile (transport equipment) and the 25th percentile (petroleum refineries) of external financial dependence
in a country with the volatility of the banking sector at the 75th percentile (China) compared with a country at the 25th
percentile (Malta).
Next, we further investigate the robustness of our results by controlling alternative determinants of growth as suggested
by the growth theory and previous empirical studies. Technically, this is done by adding the interaction terms of human
capital, physical capital and the rule of law in country k with corresponding intensity, i.e., human capital intensity, physical
capital intensity, and contract intensity, in industry i. We first add these controlling variables one at a time to the uncondi-
tional model and the conditional results are reported in columns (3)–(5) of Table 3. As shown in columns (3)–(5) of Table 3,

9
The instrumental variables used to perform the endogeneity test include the legal origins of the country and the year of independence.
1014 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

Table 4
Financial structure and financial volatilities on sector growth.

Note: The dependent variables are the growth of real value added during the 1980-1999 period for each 3-digit ISIC industry in each country. The level and
volatility of bank development are respectively measured by the ratio of domestic private credit to GDP in 1980 and the standard deviation of the growth of
the private credit by deposit money banks and other financial institutions over GDP during the 1980 -1999 period. The level and volatility of stock market
development are, respectively, measured by the initial level measured in 1980 and the standard deviation of the growth of (i) stock market capitalization
over GDP, (ii) the stock market turnover ratio, and (iii) stock market total value traded over GDP, over the 1980 -1999 period. All results are obtained using
the OLS procedure and the heteroskedasticity-robust standard errors are reported in the parentheses.
⁄⁄⁄
significance at the 1 % level.
⁄⁄
significance at the 5% level.
⁄
significance at the 10% level.

the coefficients estimated for these additional interaction terms are mostly statistically insignificant in the regressions,
except for the human capital interaction. Nevertheless, the sizes and significance levels of the coefficients for the banking
sector volatility interaction remain very similar to the unconditional model reported in column (2) of Table 3, and thus
provide additional support for the tested hypothesis of the detrimental growth effect caused by banking sector volatility
on industries with higher external financial reliance.
Finally, we extend our model to simultaneously control for all the interaction terms of human capital, physical capital and
the rule of law. Column (6) of Table 3 shows the robustness of these estimated coefficients, with particular interest being
paid to the substantial negative effect of volatility in the banking sector on industrial growth, especially for those industries
with more external financial dependence. To sum up, our conditional results from controlling additional determinants of
growth have confirmed the previous finding of the detrimental growth effect of banking sector volatility on more external
financially dependent industries, and thus lend more support to the hypothesis that volatility in the banking sector can have
a harmful impact on the growth of industries that are more externally financially dependent.

4.2. Control for the impacts from stock markets

So far, we have been focusing on financial measurement based on the development of the banking sector, i.e., the relative
size of private credit to GDP. However, recent studies, such as Beck and Levine (2002, 2004), Demirgüç-Kunt and Maksimovic
(2002), and Levine (2002), have also assessed the comparative role of the financial structure, i.e., credit-market (debt)
oriented versus capital-market (equity) oriented financial systems, in economic growth. The general message delivered from
this line of research is that for the economy to have access to a well-developed financial system is important for economic
growth, while the precise composition of the financial system is of second-order importance. To take this argument into ac-
count, we add two additional interaction terms, i.e., External Finance respectively interacts with Stock Market Development
and Volatility of Stock Market Development, to our empirical models.
In practice, the development of stock markets is measured by the stock market capitalization to GDP (Stock Market
Capitalization), the stock market turnover ratio (Stock Market Turnover Ratio) and the stock market total value
traded to GDP (Stock Market Total Value Traded), respectively. The estimated results are reported in Table 4, with
odd columns reporting unconditional specifications while even columns report the results after simultaneously controlling
for the interaction terms for human capital, physical capital and the rule of law. The results reported in the first two rows of
Table 4 show that the development of financial intermediaries and the stock market generally have no significant impact on
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1015

Table 5
Alternative volatility measures of bank development.

Note: The dependent variables are the growth of real value added during the 1980-1999 period for each 3-digit ISIC industry in each country. The levels
of bank development are respectively measured by (1) the ratio of domestic private credit to GDP, (2) deposit money bank assets over GDP, and (3)
deposit money bank assets over the sum of deposit money banks’ and the central bank’s assets, in 1980. The volatility of bank development referred to
as the range of private credit is measured by the range of growth of private credit by deposit money banks and other financial institutions over GDP
during the 1980-1999 period. The volatility of bank development referred to as the standard deviation of bank assets (A) is the standard deviation of the
growth of deposit money bank assets over GDP, over the 1980 -1999 period. The volatility of bank development referred to as the standard deviation of
the bank assets ratio (B) is the standard deviation of the growth of deposit money bank assets over the sum of deposit money banks’ and the central
bank’s assets, over the 1980 -1999 period. All results are obtained using the OLS procedure and the heteroskedasticity-robust standard errors are
reported in the parentheses.
⁄⁄⁄
denote significance at the 1% level.
⁄⁄
denote significance at the 5% level.
⁄
denote significance at the 10% level.

industrial growth, for the coefficients estimated for the interaction between External Financial Dependence and
Private Credit and the interaction between External Financial Dependence and Stock Market Development
are statistically insignificant in most of the regressions, except for the unconditional case where the stock market develop-
ment is measured by the Stock Market Turnover Ratio.
We next focus on exploring whether the volatilities of the banking sector and stock markets, which is our main interest,
significantly affect the value-added growth of industries that are more externally financially dependent. As such, in addition
to the Volatility of Private Credit as in the previous section, we independently contain an extra interaction term of
industrial-level External Financial Dependence and three country-level measures of the volatility of stock market
development, i.e., the standard deviations of the growth of Stock Market Capitalization, Stock Market Total Value
Traded, and Stock Market Turnover Ratio over the 1980–1999 period, respectively. As shown in the third row of Table 4,
the results continue to yield empirical support for the detrimental growth effect of banking sector volatility on the growth of
more externally financially dependent industries, for the coefficients estimated for the interaction between industry
External Financial Dependence and country-level Volatility of Private Credit enter negatively and significantly
at the 1% level in all specifications.
By contrast, the volatility of the stock market, as shown in the fourth row of Table 4, plays no significant role in affect-
ing industrial growth, as the estimated coefficients for the interaction between External Financial Dependence and
Volatility of Stock Market Development (no matter in what measures) are statistically insignificant in all
specifications. To sum up, our results show that the negative growth effect from the volatility of bank development is
robustly confirmed even when the impacts from the stock market are considered. On the other hand, the growth
effects from the development of the stock market and the volatility of the stock market, however, are statistically
insignificant.

5. Robustness check

In this section, we perform some robustness checks for our main finding, i.e., the volatility in the banking system is
detrimental to the growth of industries that are more externally financially dependent. For that purpose, we first consider
alternative proxies of banking industry volatility. Specifically, we consider three alternative measures of financial volatility:
(i) the high-low range of the growth of Private Credit to GDP over the 1980–1999 period, (ii) the standard deviation of
the growth of deposit money bank assets over GDP (Bank Asset (A)), and (iii) the standard deviation of the growth of
deposit money bank assets over the sum of deposit money banks’ and the central bank’s assets (Bank Asset (B)),
1016 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

Table 6
Considering R&D activity.

Note: The dependent variables are the growth of real value added during the 1980 -1999 period for each 3-digit ISIC industry in each country. The level and
volatility of bank development are respectively measured by the ratio of domestic private credit to GDP in 1980, and the standard deviation of the growth of
the private credit by deposit money banks and other financial institutions over GDP during the 1980 -1999 period. The measures of R&D activity are
respectively defined by (1) the ratio of R&D expenditure to GDP, (2) the share of R&D workers in total labor employment, and (3) the share of R&D workers
in total labor employment adjusted by human capital per worker, which is measured as educational attainment, in 1980. Differential growth measures the
difference in sectoral growth between an industry at the 75th percentile level of external financial dependence with respect to an industry at the 25th
percentile level when it is located in a country at the 75th percentile of the volatility of bank development rather than in one at the 25th percentile. Country
and industry fixed effects are included in all regressions but not reported. All results are obtained using the OLS procedure and the heteroskedasticity-
robust standard errors are reported in the parentheses.
⁄⁄⁄
denote significance at the 1% level.
⁄⁄
denote significance at the 5% level.
⁄
denote significance at the 10% level.

respectively over the 1980–1999 period.10 The results reported in Table 5 show that all the coefficients for the interaction be-
tween External Financial Dependence and Volatility of Bank Development are negative in nature and statistically
significant at the 10% level or better in all regressions. This outcome demonstrates that, even with alternative indicators of bank-
ing industry volatility, volatility in the banking system continues to exert a negative impact on the value-added growth of
industries with higher external financial dependence. Consequently, the results from this sensitivity test have demonstrated
that the detrimental growth effect of banking sector volatility on industries with higher external financial dependence is not
driven by the particular indicator of volatility of bank development (Private Credit) used.
Secondly, as the recent R&D-based endogenous growth model has emphasized the role that R&D activity has played in a
country’s growth, we thus further control for innovative activity in the regressions to check if our main result will change
with this concern being additionally considered. Following Ha and Howitt (2007) and Ang and Madsen (2011), we add three
alternative measures of R&D activity, i.e., (i) the ratio of real R&D expenditure to real GDP, (ii) the share of R&D workers in
total labor employment, and (iii) the share of R&D workers in total labor employment adjusted by human capital per worker,
to our models and the results are summarized in Table 6. As shown in Table 6, we can see that the country with higher inno-
vative activity will be beneficial to the value-added growth of those industries with higher R&D intensity, regardless of
whether other determinants of growth are included or not. Most importantly, even with this additional consideration, the
coefficients for the interaction of industrial External Financial Dependence and country-level Volatility of Pri-
vate Credit are still all negative and statistically significant at the 10% level or better, thus again suggesting that volatility
in the banking system exerts a negative impact on sectoral output growth via the channel of external finance. Accordingly,
the results from this sensitivity check have shown that even with innovative activity considered, the detrimental growth
effect of financial volatility on industrial growth still holds and thus provides additional robustness to our main result.
Thirdly, we further investigate if the negative impact of banking industry volatility on industrial growth continues to hold
when other macroeconomic determinants along with their corresponding volatilities are controlled. In particular, we
augment our specifications with the following five macro-factors: (i) the level of economic development, (ii) the government
size, (iii) the trade openness, (iv) the inflation rate, and (v) the degree of financial liberalization. The results are summarized

10
One referee has suggested that we use principal components analysis (PCA) to construct a summary measure of financial development using bank assets or
stock market measures, along the lines of Ang and McKibbin (2007). However, we are not able to use the PCA technique to create a synthetic indicator of
financial development and volatility for the following reasons: (1) for most of the country, the sample contains discontinuous observations (i.e., missing values)
for some financial indicators over the 1980–1999 period, and (2) as the measurement of financial volatility is calculated as the standard deviation of the growth
of the financial indicator over the study period, we only can obtain one observation for each country, thus rendering the technique of principal components
analysis to construct a synthetic indicator of financial volatility for each country using time series data inapplicable.
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1017

Table 7
Controlling for countries’ characteristics.

Note: The dependent variables are the growth of real value added during the 1980-1999 period for each 3-digit ISIC industry in each country. The
macroeconomic variables considered include: (1) (log) GDP per capita; (2) government spending measured by government purchases over GDP; (3) the
degree of openness measured by the sum of imports and exports divided by GDP; (4) the annual inflation rate over the study period; and (5) the degree of
financial liberalization obtained from Abiad, Detragiache, E. and Tressel, T. (2010). The level of volatility for each macroeconomic variable is measured by
the standard deviation of the growth of the corresponding variable. Differential growth measures the difference in sectoral growth between an industry at
the 75th percentile level of external financial dependence with respect to an industry at the 25th percentile level when it is located in a country at the 75th
percentile of the volatility of bank development rather than in one at the 25th percentile. Country and industry fixed effects are included in all regressions
but not reported. All results are obtained from the OLS procedure and the heteroskedasticity-robust standard errors are reported in the parentheses.
⁄⁄⁄
denote significance at the 1% level.
⁄⁄
denote significance at the 5% level.
⁄
denote significance at the 10% level.

in Table 7. Specifically, column (1) presents the regression augmented with per capita GDP and its volatility, column (2) with
government spending and its volatility added, column (3) with trade openness and its volatility included, column (4) with
inflation and its volatility covered, column (5) with financial liberalization and its volatility considered, and column (6) with
all macroeconomic variables and their corresponding volatilities simultaneously controlled. As shown in Table 7, apart from
the level of economic development and the degrees of openness and financial liberalization that exert a positive impact on
value-added growth, other macroeconomic variables and their volatilities exhibit no statistically significant impact on indus-
try output growth, However, the coefficients for the interaction of industrial External Financial Dependence and coun-
try-level Volatility of Private Credit are all negative and statistically significant at the 1% level, and thus again
robustly confirm the detrimental growth effect of banking sector volatility on more externally financially dependent
industries.
Lastly, we re-estimate the main results in the alternative 1980–1989 period to see how financial volatility in the 1980s,
which excludes several major financial crises in the 1990s, affects industrial growth.11 The unconditional and conditional esti-
mates are reported in Table 8. As can be seen, when we restrict the study period to the relatively placid 1980s, the volatility
from the banking sector does not exert any significant effect on industrial growth, regardless of whether other determinants
of growth are controlled or not. Why is this so? Compared to the 1980s, the 1990s were characterized by increasing interna-

11
The major financial crises that occurred in the 1990s included the Western European exchange rate mechanism crisis of 1992, the Mexican crisis of 1994–
1995, the East Asian financial crisis of 1997–1998, the Russian crisis of 1998 and the Brazilian crisis of 1998–1999.
1018 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019

Table 8
Alternative sample period (1980s).

(1) (2) (3)

⁄⁄⁄ ⁄⁄⁄
External Financial Dependencei  Private Creditk 0.0627 0.0480 0.0394⁄⁄
(0.0174) (0.0178) (0.0184)
External Financial Dependencei Volatility of Private Creditk 0.0577 0.0588
(0.0882) (0.0925)
Initial conditionsi,k 0.0197⁄⁄⁄ 0.0165⁄⁄⁄ 0.0145⁄⁄⁄
(0.0033) (0.0024) (0.0026)
All Controls Yes
Country dummies Yes Yes Yes
Industry dummies Yes Yes Yes
2 0.3717 0.3717 0.4106
R
Observations 1553 1472 1281

Note: The dependent variables are the growth of real value added during the 1980–1999 period for each 3-digit ISIC industry in each country. External
Financial Dependence is the difference between industry investment and industry cash flow relative to industry investment. The level and volatility of
bank development are respectively measured by the ratio of domestic private credit to GDP in 1980, and the standard deviation of the growth of private
credit by deposit money banks and other financial institutions over GDP during the 1980–1989 period. As such, the sample is restricted to the study period
of the 1980s. All results are obtained from the OLS procedure and the heteroskedasticity-robust standard errors are reported in the parentheses.
⁄
Significance at the 10% level.
⁄⁄
Significance at the 5% level.
⁄⁄⁄
Significance at the 1% level.

tional capital flows, especially for those capital flows to developing countries, thus enabling the developing countries to more
fully integrate themselves into the global economy. However, the more integrated global economy was also more often accom-
panied by unpredicted financial crises which made the global financial markets more turbulent. As a result, the outcome of no
significant effect of banking industry volatility on industrial growth in the 1980s may imply that the main result of the negative
impact of banking industry volatility on industrial growth for the 1980–1999 period may be driven by the financial crises result-
ing from the more integrated global economy in the 1990s, thereby indicating that our result may be sensitive to the choice of
study period.

6. Conclusion

Most of the existing literature studying the relationship between the financial sector and economic growth has focused on
the impacts of bank development or the structure of the financial sector on economic growth. However, as the banking sys-
tem can be one of the sources of volatility by propagating economic variability or creating risk of its own, which may defer
the investment decision and thus potential growth, it would be interesting to further investigate the role of banking industry
volatility in long-run economic growth. To cope with this concern, we employ the framework of Rajan and Zingales (1998) by
additionally considering an interaction term of country-level banking sector volatility and industry reliance on external fi-
nance. The data used consist of the cross-country cross-industry data constructed by Ciccone and Papaioannou (2009), com-
plemented by the Financial Development and Structure Database of Beck et al. (2010).
Our benchmark result for the 1980–1999 period shows that the positive effect of bank development, measured as the rel-
ative size of domestic private credit to GDP, on sectoral growth via the channel of industry external finance is attenuated when
banking sector volatility is considered. On the other hand, volatility in the banking sector, measured as the standard deviation
of the growth of private credit, has a significant negative impact on the growth of industries that are more externally finan-
cially dependent. This finding is robust to various determinants of growth controlled, different indices of stock market devel-
opment considered, alternative measures of banking industry volatility employed, and other macroeconomic volatilities
incorporated. However, when we restrict the sample to the relatively placid 1980s, the negative impact of banking industry
volatility is not statistically significant while bank development still exerts a positive growth effect as suggested by the pre-
vious literature. Our interpretation of these seemingly conflicting results is as follows. Compared to the 1980s, the 1990s are
characterized by a more economically integrated world more often accompanied by unpredicted financial crises that disturb
the banking sector. As a result, the outcome of no significant effect of banking industry volatility on industrial growth in the
1980s may imply that the main result of the negative impact of banking industry volatility on industrial growth for the 1980–
1999 period may be driven by the financial crises resulting from more integrated economies around the world in the 1990s As
such, an implication that emerges from this research is that the policymakers should closely monitor the stability of bank
development in a more and more economically integrated world to ensure long-run economic growth.

Acknowledgments

We are grateful to Elias Papaioannou for making publicly available the computer code and data used in this article. We
also thank Chris Papageorgiou (the editor) and one anonymous referee for comments and suggestions that have led to sub-
stantial improvements in this paper. The second author is also grateful to financial support from the National Science Council
(Taiwan, ROC) via Grant Number NSC 97-2410-H-032-002-MY3. Any remaining errors are our own responsibility.
 P.-C. Lin, H.-C. (River) Huang / Journal of Macroeconomics 34 (2012) 1007–1019 1019

References

Abiad, A., Detragiache, E., Tressel, T., 2010. A new database of financial reform. IMF Staff Paper 57, 281–302.
Acemoglu, D., Johnson, S., Robinson, J., Thaicharoen, Y., 2003. Institutional causes, macroeconomic symptoms: volatility, crises and growth. Journal of
Monetary Economics 50, 49–123.
Aghion, P., Angeletos, G., Banerjee, A., Manova, K., 2010. Volatility and growth: credit constraints and the composition of investment. Journal of Monetary
Economics 57, 246–265.
Aghion, P., Bacchetta, P., Rancière, R., Rogoff, K., 2009. Exchange rate volatility and productivity growth: the role of financial development. Journal of
Monetary Economics 56, 494–513.
Aizenman, J., Pinto, B., 2005. Overview. In: Aizenmann, J., Pinto, B. (Eds.), Managing Economic Volatility and Crises. Cambridge University Press, Cambridge,
United Kingdom, pp. 1–44.
Aizenman, J., Marion, N., 1993. Policy uncertainty, persistence and growth. Review of International Economics 1, 145–163.
Aizenman, J., Marion, N., 1999. Volatility and investment: interpreting evidence from developing countries. Economica 66, 157–179.
Andrés, J., Domeńech, R., Fatás, A., 2008. The stabilizing role of government size. Journal of Economic Dynamics and Control 32, 571–593.
Ang, J.B., 2008. A survey of recent developments in the literature of finance and growth. Journal of Economic Surveys 22, 536–576.
Ang, J.B., 2011a. Finance and consumption volatility: evidence from India. Journal of International Money and Finance 30, 947–964.
Ang, J.B., 2011b. Financial development, liberalization and technological deepening. European Economic Review 55, 688–701.
Ang, J.B., Madsen, J.B., 2011. Can second-generation endogenous growth models explain productivity trends and knowledge production in the Asian miracle
economies? Review of Economic Studies 93, 1360–1373.
Ang, J.B., McKibbin, W.J., 2007. Financial liberalization, financial sector development and growth: evidence from Malaysia. Journal of Development
Economics 84, 215–233.
Aysan, A.F., 2007. The effects of volatility on growth and financial development through capital market imperfections. METU Studies in Development 34, 1–
18.
Bartelsman, E.J., Gray, W.B., 1996. The NBER Productivity Manufacturing Database. National Bureau of Economic Research Technical Working Paper No. 205.
Barro, R.J., Lee, J-W., 2011. A New Data Set of Educational Attainment in the World, 1950–2010. NBER Working Paper No. 15902 (revised).
Beck, T., Demirgüç-Kunt, A., Laeven, L., Levine, R., 2008. Finance, firm size, and growth. Journal of Money, Credit and Banking 40, 1379–1405.
Beck, T., Demirgüç-Kunt, A., Levine, R., 2010. Financial institutions and markets across countries and over time: the updated financial development and
structure database. World Bank Economic Review 24, 77–92.
Beck, T., Levine, R., 2002. Industry growth and capital allocation: does having a market- or bank-based system matter? Journal of Financial Economics 64,
147–180.
Beck, T., Levine, R., 2004. Stock markets, banks and growth: panel evidence. Journal of Banking & Finance 28, 423–442.
Bernanke, B., Gertler, M., Gilchrist, S., 1999. The financial accelerator in a quantitative business cycle framework. In: Taylor, J.B., Woodford, M. (Eds.),
Handbook of Macroeconomics, vol. 1, Elsevier Science B.V., pp. 1341–1393.
Braun, M., Larrain, B., 2005. Finance and the business cycle: international inter-industry evidence. Journal of Finance 60, 1097–1128.
Brunetti, A., Weder, B., 1998. Investment and institutional uncertainty: a comparative study of different uncertainty measures. Review of World Economics
(Weltwirtschaftliches Archiv) 134, 513–533.
Cetorelli, N., Gambera, M., 2001. Banking market structure, financial dependence and growth: international evidence from industry data. Journal of Finance
56, 617–648.
Ciccone, A., Papaioannou, E., 2009. Human capital, the structure of production, and growth. Review of Economics and Statistics 91, 66–82.
Claessens, S., Laeven, L., 2003. Financial development, property rights, and growth. Journal of Finance 58, 2401–2436.
Claessens, S., Laeven, L., 2005. Financial dependence, banking sector competition, and economic growth. Journal of the European Economic Association 3,
179–207.
Campos, N.F., Nugent, J.B., 2003. Aggregate investment and political instability: an econometric investigation. Economica 70, 533–549.
Demirgüç-Kunt, A., Maksimovic, V., 2002. Funding growth in bank-based and market-based financial systems: evidence from firm level data. Journal of
Financial Economics 65, 337–363.
di Giovanni, J., Levchenko, A.A., 2009. Trade openness and volatility. Review of Economics and Statistics 91, 558–585.
Easterly, W., 2005. National policies and economic growth: a reappraisal. In: Aghion, P., Durlauf, S.N. (Eds.), Handbook of Economic Growth, vol. 1. Elsevier,
Amsterdam, pp. 1015–1059.
Fatás, A., Mihov, I., 2001. Government size and automatic stabilizers: international and intranational evidence. Journal of International Economics 55, 3–28.
Fatás, A., Mihov, I., 2006. The macroeconomic effects of fiscal rules in the US states. Journal of Public Economics 90, 101–117.
Gupta, N., Yuan, K., 2009. On the growth effect of stock market liberalizations. Review of Financial Studies 22, 4715–4752.
Ha, J., Howitt, P., 2007. Accounting for trends in productivity and R&D: A Schumpeterian critique of semi-endogenous growth theory. Journal of Money,
Credit and Banking 39, 733–774.
Hnatkovska, V., Loayza, N., 2005. Volatility and growth. In: Aizenman, J., Pinto, B. (Eds.), Managing Economic Volatility and Crises. Cambridge University
Press, Cambridge, United Kingdom, pp. 65–100.
Kharroubi, E., 2007. Crises, volatility and growth. World Bank Economic Review 21, 439–460.
King, R.G., Levine, R., 1993a. Finance, entrepreneurships, and growth-theory and evidence. Journal of Monetary Economics 32, 513–542.
King, R.G., Levine, R., 1993b. Finance and growth: Schumpeter might be right. Quarterly Journal of Economics 108, 717–737.
Knack, S., Keefer, P., 1995. Institutions and economic performance: cross-country tests using alternative institutional measures. Economics and Politics 7,
207–227.
Koren, M., Tenreyro, S., 2007. Volatility and development. Quarterly Journal of Economics 122, 243–287.
Kroszner, R.S., Laeven, L., Klingebiel, D., 2007. Banking crises, financial dependence, and growth. Journal of Financial Economics 84, 187–228.
Laeven, L., Klingebiel, D., Kroszner, R.S., 2002. Financial Crises, Financial Dependence, and Industry Growth. World Bank Policy Research Working Paper No.
2855.
Larrain, B., 2006. Do banks affect the level and composition of industrial volatility? Journal of Finance 61, 1897–1925.
Lederman, D., Saenz, L., 2005. Innovation and Development around the World, 1960–2000. World Bank Policy Research Working Paper No. 3774.
Levchenko, A.A., 2007. Institutional quality and international trade. Review of Economic Studies 74, 791–819.
Levchenko, A.A., Ranciere, R., Thoenig, M., 2009. Growth and risk at the industry level: the real effects of financial liberalization. Journal of Development
Economics 89, 210–222.
Levine, R., 2002. Bank-based or market-based financial systems: which is better? Journal of Financial Intermediation 11, 398–428.
Levine, R., 2005. Finance and growth: theory and evidence. In: Aghion, P., Durlauf, S.N. (Eds.), Handbook of Economic Growth, vol. 1A. Elsevier, Amsterdam,
pp. 865–934.
Loayza, N.V., Rancière, R., 2006. Financial development, financial fragility, and growth. Journal of Money, Credit and Banking 38, 1051–1076.
Loayza, N.V., Rancière, R., Servén, L., Ventura, J., 2007. Macroeconomic volatility and welfare in developing countries: an introduction. World Bank Economic
Review 21, 343–357.
Mishkin, F.S., 1998. International Capital Movements, Financial Volatility, and Financial Instability. NBER Working Paper 6390.
Nunn, N., 2007. Relationship-specificity, incomplete contracts and the pattern of trade. Quarterly Journal of Economics 122, 569–600.
Rajan, R.G., Zingales, L., 1998. Financial dependence and growth. American Economic Review 88, 559–586.
Ramey, G., Ramey, V.A., 1995. Cross-country evidence on the link between volatility and growth. American Economic Review 85, 1138–1151.