International Journal of Hospitality Management 29 (2010) 72–82

Contents lists available at ScienceDirect

International Journal of Hospitality Management

journal homepage: www.elsevier.com/locate/ijhosman

Impacts of positive and negative corporate social responsibility activities on

company performance in the hospitality industry
Kyung Ho Kang a,*, Seoki Lee b,1, Chang Huh c,2
a
School of Tourism and Hospitality Management, Temple University, 1700 North Broad St., Suite 201, Philadelphia, PA 19122, United States
b
School of Tourism and Hospitality Management, Temple University, 1700 North Broad St., Suite 201-F, Philadelphia, PA 19122, United States
c
College of Hospitality and Tourism Management, Niagara University, St. Vincent’s Hall, Room 304, NY 14109, United States

A R T I C L E I N F O A B S T R A C T

Keywords: In spite of growing concern for corporate social responsibility (CSR) in various industries including the
Corporate social responsibility hospitality industry, the relationship between CSR activities and financial performance is a rarely
Hospitality industry examined subject in the hospitality context. Especially, research measuring the separate impacts of
Financial performance
positive and negative CSR activities on companies’ financial performances remains, as yet, unconsidered.
Thus, this study examines different impacts of positive and negative CSR activities on financial
performance of hotel, casino, restaurant and airline companies, theoretically based on positivity and
negativity effects. Findings suggest mixed results across different industries and will contribute to
companies’ appropriate strategic decision-making for CSR activities by providing more precise
information regarding the impacts of each directional CSR activity on financial performance.
ß 2009 Elsevier Ltd. All rights reserved.

1. Introduction Hospitality Net,3 increased from 63 in 1999 to 139 in 2007 with

With a growing concern for corporate social responsibility
(CSR), leading companies in various industries, driven by
companies’ stakeholders, consumers, societies and governments,
are accelerating initiatives to demonstrate their CSR commit-
ments. For example, General Electric initiated ‘‘Ecomagination’’
(a neologism combining ecology with imagination) as its
strategic key word (Business Wire, 2005) and ever since has
continued to prepare extensive reports on environmental issues.
HSBC, a leader in global banking, announced a carbon neutral
initiative, showing its commitment to CSR for climate change
problems (BBC News, 2004), while in 2008, Wal-Mart launched
its new jewelry line called ‘‘Love, Earth’’ in which Wal-Mart only
uses gold, silver and diamonds from mines and manufacturers
that meet sustainability standards established by Wal-Mart
(Fibre2fashion, 2008).
Along with this general trend of public and corporate attention
to CSR issues, the hospitality industry has indicated an ever
increasing interest in CSR. For example, the number of CSR and
environmental news items on a hospitality industry web site,
* Corresponding author. Tel.: +1 215 204 5612; fax: +1 215 204 8705.
E-mail addresses: kyung.ho.kang@temple.edu (K.H. Kang), seokilee@temple.edu
(S. Lee), chuh@niagara.edu (C. Huh).
1
Tel.: +1 215 204 0543; fax: +1 215 204 8705.
2 3
Tel.: +1 716 286 8223; fax: +1 716 286 8277.
10.4% compound annual growth rate while the number of overall
news increased with just a 2.8% compound annual growth rate.
More leading hospitality companies, including Hilton, Starwood,
Choice Hotels, Starbucks and McDonald’s, exclusively provide CSR-
related reports. Also, at the 2008 Democratic National Convention
in Denver, sustainable and eco-friendly wood key cards debuted at
hotels, and following the International Air Transport Association’s
recognition of airline business’ social, economic and environ-
mental influences (International Air Transport Association, n.d.),
airlines including British Airways, SAS Group, Cathay Pacific and
Dragonair launched carbon-offsetting programs to help fund
environmental projects (Air Transport Action Group, n.d.; Arm-
strong, 2008; Kjelaggard, 2007).
In spite of rising interest in CSR and vigorous participation in
CSR activities in the hospitality industry, only a few studies have
been done on the impacts of these activities on financial
performance (for example, Lee and Park, 2009; Garcı́a and Armas,
2007). Moreover, despite the significant industry effect reported by
the CSR literature (Amato and Amato, 2007; Banerjee et al., 2003;
Brammer and Millington, 2005; Sweeny and Coughlan, 2008),
comparative research across industries in the hospitality field has
been rarely conducted. While Lee and Park (2009) examined
impacts of CSR activities on U.S. hotels’ and casinos’ financial
performances, their study measured CSR activities as an aggregate
Retrieved September 2, 2008, from www.hospitalitynet.org.

0278-4319/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhm.2009.05.006
 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82
activity, not differentiating impacts of positive and negative CSR
activities. This treatment may be problematic according to the
positivity and negativity effects arguing that magnitudes of
impacts from positive and negative activities can be different.
The purpose of this study, therefore, is to examine the separate
impacts of positive and negative CSR activities of each hospitality
industry (hotel, casinos, restaurant and airline industry) on both
profitability (short-term performance) and firm value (long-term
performance). Theoretical backgrounds justifying examining
separate impacts of each directional CSR activity are positivity
effect and negativity effect which have been studied in social
science, including psychology and applied to management fields
such as marketing (more detailed review is provided in Section
2.2).
Findings of this study are expected to contribute not only to the
existing CSR literature, but also to management in hospitality firms
by providing strategic insight for CSR’s beneficial impact on
increasing firm value by taking into account legitimate stake-
holders. More specifically, this study aims to contribute to
hospitality firms’ strategic decision-making processes by helping
to predict the varying impacts of positive and negative CSR
activities on financial performance in their respective industries,
and thereby help to develop more appropriate CSR strategies. For
example, if the study results confirm the positivity effect in hotel
industry, the primary action for hotels should be taken to increase
positive CSR activities rather than reduce negative CSR activities,
while if the negativity effect holds, firms should first focus on
reducing negative CSR activities, then move on increasing positive
CSR activities. Otherwise, by just enhancing or eliminating CRS
activities as a whole, firms may sacrifice valuable opportunities for
reaping rewards from more precise decisions.
The study next reviews relevant literature, followed by
methodology and data. Results and discussion are provided, and
limitations and suggestions for future research conclude the study.
2. Literature review
2.1. CSR and financial performance
A modern concept of CSR has evolved since the 1950s,
formalized in the 1960s and proliferated in the 1970s (Carroll,
1999). Based on various studies from the CSR literature (Carroll,
1999; Engardio et al., 2007; Hart, 1995; Holme and Watts, 2000;
McWilliams and Siegel, 2001; Nicolau, 2008; Tsoutsoura, 2004),
CSR can be broadly defined as the activities making companies
good citizens who contribute to society’s welfare beyond their
own self interests. Throughout the past several decades,
numerous aspects of CSR have been the subject of investigation
in academic and business literature, and according to the
framework of Schwartz and Carroll (2003), economic, legal and
ethical domains can be epitomized as the most common
components of CSR.
One aspect of CSR interesting to many financial economists is
the economic domain: financial impact of CSR for profit-seeking
corporations. Regarding the relationship between companies’ CSR
activities and their performances (especially, financial perfor-
mance), the literature presents three assertions.
The first group of researchers, based on the viewpoint of
Friedman (1970), has found a negative relationship between CSR
activities and financial performance as measured by, for example,
stock price changes (Vance, 1975), excess return (Wright and
Ferris, 1997), or analysts’ earnings-per-share forecasts (Cordeiro
and Sarkis, 1997). Friedman argued that managements are selected
by the stockholders as agents and their sole responsibility is acting
on behalf of the principals’ best interests. From Friedman’s
perspective, the one and only social responsibility of business is
73
to use its resources and engage in activities designed to increase
profits and wealth of owners. Any other activities disturbing the
optimal allocation of scarce resources to alternative uses exert an
adverse influence on firm performance.
The second group argued for positive impact from companies’
CRS activities on financial performance (Aragón-Correa et al., 2008;
Bird et al., 2007; Bragdon and Marlin, 1972; Grave and Waddock,
1994; Hart and Ahuja, 1996; Heinze, 1976; Judge and Douglas,
1998; Klassen and McLaughlin, 1996; Nicolau, 2008; Orlitzky et al.,
2003; Pava and Krusz, 1996; Preston and O’Bannon, 1997; Russo
and Fouts, 1997; Sturdivant and Ginter, 1977; Waddock and Grave,
1997). This group’s assertion, based on stakeholder theory
(Freeman, 1984), suggests that firms expand the scope of
consideration in their decision-making and activities beyond
shareholders to several other constituencies with interests, such as
customers, employees, suppliers and communities. The second
group asserts that CSR activities, which encompass all legitimate
stakeholders’ implicit claims as stakeholder theory suggests, can
improve firm value by (1) immediate cost saving, (2) enhancement
of firm reputation, and (3) dissuasion of future action by regulatory
bodies including governments which might impose significant
costs on the firm (Bird et al., 2007).
A third group has supported no particular relationship between
CSR activities and financial performance (Abbott and Monsen,
1979; Alexander and Buchholz, 1978; Aupperle et al., 1985; Teoh
et al., 1999), partially arguing for the existence of too many
confounding factors for researchers to uncover a particular impact
from CSR on firm performance.
Seemingly contradictory themes between Friedman’s (1970)
viewpoint and the stakeholder theory arise from the assumption
that CSR, which considers the interests of a broad spectrum of
stakeholders (suggested by stakeholder theory), is in fact
detrimental to value maximization activities of the firm (asserted
by Friedman’s viewpoint). However, Jensen (2001) attempted to
reconcile the potential conflict between these two viewpoints by
proposing enlightened stakeholder theory, which asserts that a
firm cannot maximize its long-term value if it ignores the interests
of diverse stakeholders. And, according to Post et al. (2002), a
firm’s capacity that generates sustainable wealth over time and its
long-term value are determined by the relationship with both
internal and external stakeholders. CSR, if it contributes to
enhancing firm value, can be an appropriate corporate strategy as
the stakeholder theory suggests, not an exploitation of share-
holders’ wealth to benefit other parties, as Friedman (1970)
worried.
In the hospitality industry, examination of CSR activities’
impacts on financial performance seems insufficient, considering
the industry’s growing interest in CRS and firms’ massive
commitments. Kirk (1995) studied hotel environmental policies
and activities in the U.K. by employing a survey method, and
concluded that the hotel industry reacted to environmental issues
only for direct financial rewards (e.g., energy and waste manage-
ment) and governmental requirements. However, Kirk’s study
focused only on the environmental issue, and did not attempt to
specifically relate hotels’ actual environmental activities to their
financial performances. In a recent study, Garcı́a and Armas (2007)
found a positive relationship between hotel companies’ CSR
activities and return on assets (ROA). However, in measuring CSR
activities, they collected data based on managers’ opinions, which
might be biased, reflecting lower construct validity. Nicolau (2008)
employed an event study that examined abnormal returns (ARs) of
two Spanish hotels with 26 CSR-related announcements during the
1996–2006 period. The study found positive ARs and concluded
that CSR is considered value-added to hotel firms. Another study
(Lee and Park, 2009) measured impacts of CRS activities separately
for hotel and casino companies in terms of profitability (ROE and
74 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82
ROA) and firm value (AMV: average market value). Lee and Park
found a different relationship between CSR activities and profit-
ability (and firm value) for hotel companies (positive relationship)
and casino companies (no relationship). However, their study
measured CSR as one aggregate variable that combines positive
and negative CSR activities together, thereby commingling the
possible individual impact of two such different (i.e., positive and
negative) activities.
2.2. Negativity effect vs. positivity effect
Psychology asserts two different types of attributional biases or
tendencies (i.e., negativity effect and positivity effect) when
evaluating behaviors or processing impressions. A group of social
science researchers examined the negativity effect, arguing that
people place more weight on negative information than positive
information when evaluating a target (Fiske, 1980; Hamilton and
Huffman, 1971; Klein, 1996; Lampel and Anderson, 1968). Some
scholars contributed to expanding this negativity bias to a product
level by conducting studies into the context of word-of-mouth
communication (Herr et al., 1991), time pressure or distraction
(Wright, 1974), and brand defection (Winchester et al., 2008). An
underlying logic of the negativity effect is that negative informa-
tion is a more useful diagnostic for categorizing and is given greater
weight than positive information (Ahluwalia et al., 2000).
On the other hand, another group of researchers supported the
positivity effect under specific conditions. Charles et al. (2001)
compared the negative and the positive affect in four different
generations and found the negative affect decreases with age in all
generations while the positive affect remains stable, which means
older individuals may be more likely than the younger ones to pay
attention to positive stimuli than to negative ones. Similarly,
Mather and Carstensen (2005) found the positivity effect in older
adults. They argued that such positivity effect exists because older
adults focus more on emotion regulation and cognitive control
mechanisms that enhance positive and diminish negative infor-
mation. Folkes and Patrick (2003) examined the positivity effect in
the service context and found that when the consumer has little
experience with a service, positive information leads to inference
of assimilation to a greater extent than negative information does.
According to Folkes and Kamins (1999), the positivity effect is also
associated with a firm’s ethical behavior; the study results showed
that a superior product attribute enhances attitudes toward
ethically behaving firms more than toward unethically behaving
firms.
According to the cue-diagnosticity approach (Skowronski and
Carlston, 1987), negativity effect pertains to the domain of
morality, and the positivity effect relates to the domain of ability.
Unethical behaviors and superior abilities are perceived as more
diagnostic than moderate behaviors or abilities because extreme
behaviors are generally seen as characteristic only of people who
have extreme traits. Conversely moderate behaviors are consid-
ered characteristic both of people who have extreme traits and
people who have moderate traits.
Based on differently weighted perceptions for positive or
negative information in a certain condition, the likelihood of an
asymmetric impression exists for a firm’s positive and negative
CSR activities. For example, as the cue-diagnosticity approach
(Skowronski and Carlston, 1987) suggested, if CSR is perceived as
information related to morality (McWilliams and Siegel, 2001),
negative CSR may be given more weight (i.e., negativity effect). If
CSR is perceived as activity pertaining to ability, as McGuire et al.
(1988) suggested, then people may tend to place more weight on
positive CSR (i.e., positivity effect). According to Frooman’s (1997)
meta-analysis, the stock market reacts negatively to a firm’s
engagement in socially irresponsible or illicit activities, while there
is insufficient evidence exists to attribute socially responsible
activities’ impact on firm value. Similarly, Johnson (2003)
developed a CSR continuum framework which asserted that a
firm’s socially irresponsible or illegal activities hurt financial
performance, but socially responsible activities linked to no fiscal
advantages. Bird et al. (2007) investigated relationships between
firms’ positive and negative CSR activities for each stakeholder
group and equity performance, measured by market to book and
price to earnings ratio. They reported different directions and
magnitudes for each (i.e., positive and negative) CSR activity’s
impact on firm performance. For example, the findings showed
employment-related negative CSR activity influenced more
heavily than employment-related positive CSR activity. However,
the dataset included overall US companies in the S&P500 index,
and therefore did not provide industry specific examination. In
addition, measuring short-term financial performance such as ROE
and ROA other than market-based firm value (market to book and
price to earnings ratio) is needed for firms to decide appropriately
their periodic focus on CSR strategies. This study’s expectation is
that examination of separate impacts of each directional CSR
activity on short-term (profitability) and long-term financial
performance (firm value) in the hospitality context, including
four different industries, will provide more comprehensive and
industry specific information for the hospitality field. Findings will
help hospitality executives and managers develop more appro-
priate CSR strategies.
3. Methodology
3.1. Model
To examine the research question and accomplish this study’s
purpose, the analysis is conducted for two different dimensions:
accounting profitability to measure short-term financial perfor-
mance and firm value to measure long-term financial performance.
The study uses a pooled linear regression model where return on
assets (ROA) and return on equity (ROE) is adopted, in turn, as a
dependant variable for a model of each industry (i.e., hotel, casino,
restaurant and airline industries) to measure profitability, and
Tobin’s Q and price-earnings ratio (PER) is used, in turn, as a proxy
for firm value. The interest (main) variables are positive CSR
activity and negative CSR activity, while sales, leverage ratio, and
S&P500 index control for the effect of firm size, capital structure,
and market condition, respectively.
Represented models are:
ROEðROAÞt ¼ a0 þ a1 PCSRt þ a2 NCSRt þ a3 SIZEt
þ a4 LEVERAGEt þ a5 MARKETt þ et ;
PERðTobin’s Q Þt ¼ a0 þ a1 PCSRt þ a2 NCSRt þ a3 SIZEt
þ a4 LEVERAGEt þ a5 MARKETt þ et ;
where ROE represents return on equity measured by the log of the
sum of 1 and OBIDA (operational earnings before interest,
depreciation and amortization) divided by total shareholder’s
equity; ROA represents return on asset measured by the log of the
sum of 1 and OBIDA divided by total assets; PER represents price-
earnings ratio (P/E ratio) measured by the log of fiscal year end
price divided by OBIDA per share; Tobin’s Q represents an
approximate Tobin’s Q measured by the log of market value to
book value ratio; PCSR represents positive CSR activities measured
by KLD STAT; NCSR represents negative CSR activities measured by
KLD STAT; SIZE represents a firm size measured by the log of sales;
LEVERAGE represents a firm’s capital structure measured by debt-
to-assets ratio; MARKET represents market condition, measured by
 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82
average monthly S&P500 Index for each year, and subscript, t,
indicates the time period.
3.2. Dependant variable
The study examines financial performance by employing both
accounting number-based profitability and market-based firm
value. As a dependent variable, ROE(ROA) is used as a proxy for
profitability and PER(Tobin’s Q) is used as a proxy for firm value.
ROE(ROA) represents short-term financial performance of a firm by
measuring how a firm efficiently creates profits using its equity
(assets) during a fiscal year, and PER(Tobin’s Q) represents long-
term performance by reflecting investors’ perceptions from the
stock market relative to a firm’s accounting value – earnings-per-
share (book value).
Regarding a firm’s return, a component of ROE, ROA and PER,
the study adopted operational earnings before interest, deprecia-
tion and amortization to adequately gauge the fundamental power
of a firm’s operations instead of net income which can be distorted
by non-operating factors. To maintain compatibility in calculating
PER, the study made sure to collect fiscal year-end stock price of
each firm because some firms do not follow a calendar year as their
fiscal year. The study chose the approximate Tobin’s Q, suggested
by Chung and Pruitt (1994), to facilitate data collection and
computational simplicity. Approximate Tobin’s Q is defined as:
MVE þ PS þ DEBT
approximate Tobin’s Q ¼ ; because the market perceives the firm as too risky (Brealey and
TA
where MVE is the product of firm’s stock price and the number of
common stocks outstanding; PS represents the liquidating value of
outstanding preferred stock; DEBT is the value of short-term
liabilities net of short-term assets plus the book value of long-term
debt, and TA represents the book value of total assets.
To enhance normality by correcting positive skewness, log
transformation is employed for each dependent variable (more
detailed explanation appears in Section 4). For ROE and ROA, before
natural logarithmic transformation, 1 is added to each value to
maintain observations with negative values (Howell, 1992).
Therefore, in the study, exact form of ROE and ROA is the log of
(1 + ROE) and the log of (1 + ROA), and that of PER and Tobin’s Q is
the log of (PER) and the log of (Tobin’s Q), respectively.
3.3. Interest variables and control variables
In this study, PCSR (positive CSR) and NCSR (negative CSR) were
included in the model as interest variables, representing positive
CSR performance (i.e., the degree of socially responsible activities),
and negative CSR performance (i.e., the degree of socially
irresponsible activities) of each firm in a particular year. Positive
CSR is from positive ratings (strengths) provided by the KLD STAT
database encompassing seven major issue areas (i.e., community,
corporate governance, diversity, employee relations, environment,
human rights and product). For example, if the company has
consistently given over 1.5% of trailing three-year net earnings
before taxes (NEBT) to charity, or has otherwise been notably
generous in its giving, the company earns one point for its positive
CSR score in the community area, and if the company has made
notable progress in the promotion of women and minorities,
particularly to line positions with profit-and-loss responsibilities
in the corporation, the company’s positive CSR score increases by
one unit for the diversity area. PCSR in the study is a cumulative
score of positive CSR ratings (strengths) from KLD STAT. Negative
CSR is measured as a summed score of negative CSR ratings
(concerns) in seven major issue areas and controversial business
indicators which include negative ratings only from KLD STAT. For
example, companies that own and/or operate casinos are
75
inevitably given a negative rating in controversial business issues
related to gambling. To illustrate specific PCSR and NCSR data
examples, Marriot International’s PCSR for 2007 was 9, which
means 9 positive CSR ratings across seven key issue areas.
Similarly, Marriot International’s NCSR in 2007 was 7, which
represents 7 concerns across key issue areas.
This study uses three control variables in the linear regression
model: SIZE, LEVERAGE, and MARKET. SIZE, estimated by the log of
sales, controls for any systematic effect generated by different sizes
of firms in relationship to their financial performances. Banz
(1981) investigated the small firm effect which suggested that
smaller firms, on average, yield more expected common stock
return than larger firms. In contrast, according to Ball (1978), small
firms are riskier than larger firms that tend to provide more
information to the financial community. Less information asym-
metry between investors and managers in larger firms decreases
the cost of capital, which in turn increases firm value. Also, based
on the economy of scale, large firms perform better than small
ones, thereby proposing a positive relationship between firm size
and profitability (Chauvin and Hirschey, 1993). LEVERAGE
represents debt-to-assets ratio (total debt divided by total assets)
and controls for the effect driven by firm-specific capital structure.
On one hand, a firm can take advantage of increased debt due to the
fact that interest expense is tax deductible while dividends are not
(McConnell and Servaes, 1990). On the other hand, when a firm
increases its debt excessively, the firm’s equity return may dwindle
Myers, 2003). MARKET, computed by averaging monthly S&P500
index per respective year, is incorporated in the model to control
for the effect of general economic conditions in a specific year.
During economic boom (recession) periods, firms tend to perform
better (worse). Such correlation may confound the relationship
between firm performance and CSR activities, and is, therefore,
controlled in the model.
4. Data
The study collected CSR data from KLD Research and Analytics’
KLD STAT that covers approximately 80 indicators and provides
extensive CSR rating information for companies listed on the
S&P500 and Russell 3000 Indices. The KLD CSR ratings are available
from 1991 to 2007, which constitutes the sample period of this
study. To retrieve other annual accounting and financial data
including sales, income, total stockholders’ equity, total debt, and
stock price, this study uses the Compustat database, and to collect
monthly stock price, the study uses the web site, Yahoo! Finance.4
Before completing the main analysis, to check assumptions of
multiple linear regression model for the ordinary least squares
method (i.e., normality, homoscedasticity and independence of
dependent variable, linearity between response variable and
explanatory variables, multicollinearity among independent vari-
ables, and autocorrelation between residuals), the study conducted
a residual analysis in which, as diagnostics, studentized residuals,
Cook’s distance, DFFITS, DFBETAS, Jarque–Bera test statistics, Q–Q
plot, residual histogram, VIF value and Durbin–Watson d-statistics
were employed along with residual plots of independent variables.
To check normality, the study employed the Jarque–Bera test, a
hypotheses test considering skewness and kurtosis of a random
variable (Gujarati, 2006). Because the result of the Jarque–Bera test
showed test statistics of most of four dependant variables of each
industry, with positive skewness, exceeded the critical value at 5%
significance level, the study determined a natural logarithmic
transformation for each dependent variable. After transformation,
normalities of all 16 (four analyses for each of four industries)
4
Retrieved September 17, 2008, from www.finance.yahoo.com.
76 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82

dependent variables improved with their respective Jarque–Bera 5. Results

test statistics within critical values. Q–Q plot and residual
histogram were used to confirm normality. In addition, the study 5.1. Descriptive statistics
analyzed homoscedasticity and independence of each transformed
dependent variable and linearity of each dependent and indepen- Table 2 summarizes descriptive statistics of the data for hotel,
dent variable with studentized residual-by-predicted value plot casino, restaurant and airline industries. Statistics of explanatory
and residual plots of independent variables. To improve conditions variables (i.e., PCSR, NCSR, SIZE, LEVERAGE and MARKET) in each
of assumptions, the study detected outliers and influential cases panel represent those in the model that has the largest number of
using studentized residuals, Cook’s distance, DFFITS, and DFBETAS observations for each industry (e.g., ROA model in hotel and
of each observation. Observations whose studentized residuals Tobin’s Q model in airlines industry). Although the dataset for
values fell above 99 % of t-distribution were removed from original each model is not perfectly identical with the slightly different
dataset as outliers (Anderson et al., 2005). Influential cases with number of observations due to the data cleaning process, a large,
Cook’s distance values larger than 1 (Anderson et al., 2005) were common portion of datasets for the four models ensures just
eliminated. Any observation whose DFFITS or DFBETAS value was negligible difference between statistics of explanatory variables
greater than 2 or less than 2 was also excluded from dataset as an in each model. For example, mean (standard deviation) of PCSR in
influential case as SAS 9.1 user’s guide recommended. For example, ROA model for the hotel industry is 2.17 (2.32) and mean
no observation was eliminated for ROA model of hotels, and one (standard deviation) of PCSR in PER model in the hotel industry is
observation was removed from PER model’s original dataset 2.2 (2.34).
because it was beyond cutoff values for studentized residuals. For
ROA and Tobin’s Q model of hotels, 2 observations out of 46 were
rejected by applying decision rules for studentized residual, Cook’s Table 2
distance, DFFITS and DFBETAS. VIF value was measured to identify Summary of descriptive statistics.a.
the presence of multicollinearity and the study found that no Variable N Mean S.D. Minimum Maximum
independent variable’s VIF value throughout the 16 models (four
Panel I: Hotel industry
for four industries) was greater than 10: the cutoff value suggested
ROE 44 0.291 0.108 0.121 0.55
by Belsley et al. (1980). The study detected positive autocorrela- ROA 46 0.103 0.022 0.059 0.146
tions for all 16 models after conducting Durbin–Watson d test, and PER 45 2.133 0.397 1.17 3.08
adopted t-statistics using Newey–West standard errors for each Tobin’s Q 44 0.283 0.207 0.136 0.886
PCSR 46 2.174 2.321 0 9
regression coefficient to account for biases of the estimated
NCSR 46 2.13 1.424 0 7
variances of ordinary least squares (OLS) estimators as well as SIZE 46 8.061 1.183 5.63 9.472
possible heteroscedasticity (Gujarati, 2003) (more detailed expla- LEVERAGE 46 0.388 0.155 0.153 0.786
nation appears in Section 5). Table 1 summarizes assumption MARKET 46 1061 325.462 381.534 1478
check and data cleaning techniques. Panel II: Casino industry
Finally, 44, 46, 45, 44 observations (without violation of ROE 58 0.372 0.171 0.1 0.68
normality, homoscedasticity, independence, linearity and multi- ROA 55 0.105 0.046 0.027 0.234
collinearity) were selected for the hotel industry’s ROE, ROA, PER PER 48 1.783 0.474 0.329 2.896
Tobin’s Q 53 0.325 0.336 0.244 1.328
and Tobin’s Q model, respectively. For the casino industry, 58
PCSR 58 0.707 0.859 0 3
(ROE), 55 (ROA), 48 (PER) and 53 (Tobin’s Q) observations were NCSR 58 2.483 1.301 1 6
used. To measure the effect in the restaurant industry, 132 (ROE), SIZE 58 7.046 2.014 1.635 9.29
143 (ROA), 150 (PER) and 156 (Tobin’s Q) observations were LEVERAGE 58 0.534 0.157 0 0.764
MARKET 58 1191 248.786 460.664 1478
collected, and in the airline industry’s models, 60, 65, 65 and 70
observations were used for ROE, ROA, PER and Tobin’s Q model, Panel III: Restaurant industry
respectively. ROE 132 0.295 0.082 0.107 0.502
ROA 143 0.17 0.034 0.088 0.26
PER 150 2.066 0.45 0.851 3.358
Table 1 Tobin’s Q 156 0.592 0.417 0.496 1.8
Assumption check and data cleaning specification. PCSR 156 2.051 2.327 0 12
NCSR 156 2.28 1.634 0 8
Objective Technique Remedy SIZE 156 7.379 1.249 4.562 10.034
LEVERAGE 156 0.263 0.19 0 1.049
Panel I: Assumption check MARKET 156 1146 271.861 381.534 1478
Normality check Jarque–Bera test Log transformation for each
dependent variable Panel IV: Airline industry
Homoscedasticity check Studentized Using Newey–West ROE 60 0.348 0.169 0.02 0.732
residuals plot standard errors ROA 65 0.102 0.061 0.011 0.24
Independence check Studentized Improved by data cleaning PER 65 1.466 0.727 0.069 2.843
residuals plot Tobin’s Q 70 0.191 0.369 0.978 0.588
Linearity check Studentized Improved by data cleaning PCSR 70 2.471 2.125 0 8
residuals plot NCSR 70 2.243 1.69 0 7
Multicollinearity check VIF value Not violated in this study SIZE 70 8.23 1.017 5.888 9.85
measurement LEVERAGE 70 0.343 0.164 0.092 0.69
Autocorrelations check Durbin–Watson Using Newey–West MARKET 70 1015 372.727 381.534 1478
d test standard errors a
ROE represents return on equity measured by ln(1 + OBIDA/total shareholder’s
equity); ROA represents return on asset measured by ln(1 + OBIDA/total asset); PER
Objective Technique Cutoff value represents price-earnings ratio measured by ln(fiscal year end price/OBIDA per
share); Tobin’s Q is measured by ln(market value/book value); PCSR represents
Panel II: Data cleaning
positive CSR activities measured by KLD STAT; NCSR represents negative CSR
Outliers Studentized residuals analysis 99 % of t-distribution
activities measured by KLD STAT; SIZE represents a firm size measured by ln(sales);
Influential cases Cook’s distance value 1
LEVERAGE represents a firm’s capital structure measured by debt-to-asset ratio,
DFFITS 2, 2
and MARKET represents market condition measured by average monthly S&P500
DFBETAS 2, 2
Index for each year.
 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82 77

Across four industries, values of PERs are distributed within the
broadest range with larger means and standard deviations than the
other three dependant variables (i.e., ROE, ROA and Tobin’s Q).
Average values of monthly S&P500 index per a year vary from
381.534 to 1478 throughout the 17-year period (from 1991 to
2007). For hotel, restaurant and airline industries, the minimum
value of positive and negative CSR activities is commonly 0;
however, for the casino industry, the minimum value of negative
CSR activities is 1 because all firms in the casino industry were
given one point negative CSR rating as gambling is a controversial
business issue in the KLD STAT data base. The maximum values of
positive CSRs in four models in the hotel, restaurant and airline
industries are higher than those of negative CSRs, while in the casino
industry, the maximum values of negative CSRs are greater by 3.
The study also performed a Pearson correlation test to report
general bivariate correlations between study variables as provided
in Table 3. Similar to the case of descriptive statistics, correlation
values between independent variables are from the model with the
largest observations in each industry, and differences between
models are subtle. Across four industries a correlation between
ROE and ROA, and between PER and Tobin’s Q is positive at the 1%
significance level.
For the hotel industry, PCSR appears to have a significant and
positive correlation with firm values (i.e., PER and Tobin’s Q). In
the casino industry, NCSR positively correlates with ROE
representing short-term financial performance while negatively
correlating with Tobin’s Q representing firm value. From correla-
tion analysis in restaurant industry, PCSR commonly shows a
significant (at the 1% level) and positive correlation with both
profitability and firm value variables. NCSR appears to have a
positive correlation with only Tobin’s, although the magnitude
and significance level (5%) are smaller than those of PCSR. In the
airline industry, PCSR negatively correlates with profitability
(ROE) and positively correlates with firm value (PER and Tobin’s
Q), while NCSR appears to be positively correlated with profit-
ability (ROE) and negatively correlated with firm value (PER and
Tobin’s Q) at the 1% a level. PCSR and NCSR are negatively
correlated, which implies airline firms are clearly differentiated in
terms of direction of CSR.
5.2. Main analysis
In the 16 pooled regression results (four models for each of four
industries), goodness of fit of each models, except ROA model of

Table 3
Summary of Pearson’s correlation.

Variable ROE ROA PER Tobin’s Q PCSR NCSR SIZE LEVERAGE MARKET

Panel I: Hotel industry

ROE 1 0.55** 0.425** 0.098 0.063 0.144 0.105 0.423** 0.072
ROA 1 0.193 0.277 0.117 0.082 0.146 0.044 0.19
PER 1 0.752** 0.602** 0.107 0.373* 0.572** 0.067
Tobin’s Q 1 0.71** 0.029 0.431** 0.487** 0.152
PCSR 1 0.195 0.678** 0.636** 0.264
NCSR 1 0.2 0.192 0.235
SIZE 1 0.53** 0.02
LEVERAGE 1 0.187
MARKET 1

Panel II: Casinos industry

ROE 1 0.775** 0.47** 0.4** 0.05 0.295* 0.581** 0.59** 0.295*
ROA 1 0.144 0.08 0.071 0.036 0.578** 0.219 0.339*
PER 1 0.684** 0.177 0.038 0.036 0.309* 0.544**
Tobin’s Q 1 0.094 0.279* 0.441** 0.656** 0.125
PCSR 1 0.05 0.335* 0.121 0.193
NCSR 1 0.123 0.136 0.04
SIZE 1 0.457** 0.033
LEVERAGE 1 0.13
MARKET 1

Panel III: Restaurants industry

ROE 1 0.572** 0.19* 0.113 0.306** 0.053 0.599** 0.428** 0.048
ROA 1 0.233 0.625** 0.385** 0.14 0.487** 0.123 0.07
PER 1 0.778** 0.348** 0.104 0.003 0.56** 0.03
Tobin’s Q 1 0.492** 0.19* 0.252** 0.324** 0.031
PCSR 1 0.111 0.57** 0.051 0.166*
NCSR 1 0.303** 0.153 0.39**
SIZE 1 0.203* 0.053
LEVERAGE 1 0.081
MARKET 1

Panel IV: Airlines industry

ROE 1 0.541** 0.659** 0.2 0.446** 0.321** 0.221 0.177 0.136
ROA 1 0.183 0.253* 0.01 0.01 0.45** 0.617** 0.041
PER 1 0.69** 0.496** 0.649** 0.054 0.19 0.028
Tobin’s Q 1 0.437** 0.626** 0.319** 0.183 0.016
PCSR 1 0.42** 0.229 0.444** 0.065
NCSR 1 0.208 0.005 0.015
SIZE 1 0.092 0.125
LEVERAGE 1 0.047
MARKET 1

ROE represents return on equity; ROA represents return on asset; PER represents price-earnings ratio; Tobin’s Q is measured market value to book value ratio; PCSR represents
positive CSR activities measured by KLD STAT; NCSR represents negative CSR activities measured by KLD STAT; SIZE represents a firm size measured by log of sales; LEVERAGE
represents a firm’s capital structure measured by debt-to-asset ratio, and MARKET represents market condition measured by average monthly S&P500 Index for each year.
*
Significance level of 0.05.
**
Significance level of 0.01.
78 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82

the hotel industry, implies that each model explains a significant
portion of total variance; F-value confirms overall significance of
15 models at the 1% a level. Every VIF value smaller than 4 suggests
no substantial multicollinearity problem in examinations (Belsley
et al., 1980). The study estimates regression coefficients using
Newey–West heteroscedasticity and autocorrelation consistent
(HAC) standard errors to account for significant autocorrelations
across all models and possible heteroscedasticity (Gujarati, 2003).
Newey–West standard errors have been frequently used in
examinations with financial panel data to adjust for effects of
underestimation of OLS estimators’ true variances and consequent
inflation of t-values caused by serious autocorrelation (Henisz and
Zelner, 2001; Jamal and Yu, 2007; Yu, 2008). This study used
bandwidth or lag length of 2 for Newey–West standard errors
following the formula, bandwidth = INTEGER[4(time period/100)2/
9
] suggested by Newey and West (1994).
Table 4 presents results of the main analysis for the hotel
industry. While no statistically significant impacts are apparent
for both positive and negative CSR on ROE and ROA at the 5% a
level, positive CSR (PCSR) shows positive impact on both PER at
the 5% a level (t-value = 2.44) and Tobin’s Q at the 0.1% a level (t-
value = 3.96). Negative CSR (NCSR) appears to have no statis-
tically significant impact on both firm value measures. PCSR’s
coefficient, 0.11 (0.07) in PER (Tobin’s Q) model means that one
unit increase in positive CSR activity generates, on average, 11%
(7%) growth rate in price-earnings ratio (Tobin’s Q), given other
variables are held constant (Gujarati, 2006), hence, in this study,
PER (Tobin’s Q) represents a natural logarithm of price-earnings
ratio (Tobin’s Q). SIZE has a positive relationship with ROE at the
5% significance level, while LEVERAGE positively relates with
ROE at the 0.1% level. Although MARKET shows a significantly
negative impact on ROA at the 5% a level, its practical magnitude
seems negligible (absolute value of coefficient is less than
0.0001).
Results of regression analysis for casino firms presented in
Table 5 show no significant (at the 5% a level) impact of interest
variables (i.e., PCSR and NCSR) on both profitability and firm
value. Only control variables have some significant effects on
financial performance; SIZE has a positive relationship with ROE
and ROA at the 0.1% significance level while leverage shows a
negative relationship with Tobin’s Q at the 0.1% significance
level and appears to be negatively related with ROA at the 1% a
level.
The examination of the restaurant industry shows similar
results with the hotel industry regarding impacts of interest
variables; no significant impacts of PCSR and NCSR are found on
accounting profitability (i.e., ROE and ROA) while significantly
positive impacts of positive CSR (PCSR) are apparent on both PER
(t-value = 3.46) and Tobin’s Q (t-value = 3.43) (Table 6). SIZE shows
a positive relationship with only profitability at the 0.1% a level.
Additionally, LEVERAGE negatively relates to firm value, while it
shows a positive (negative) relationship with ROE (ROA). A
significant and positive relationship exists between MARKET
and ROA.
Results of the main analysis for airlines appear in Table 7, and
present a negative impact of PCSR on ROA at the 1% a level (t-
value = 2.99) and a negative impact of NCSR on market-based
firm values, PER (t-value = 3.11) at the 1% significance level and

Table 4
Summary of pooled regression analysis for hotel industry.

PCSR NCSR SIZE LEVERAGE MARKET

Panel I: ROE (ROEt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)

Coefficient 0.023 0.0075 0.0326 0.7472 >0.0001
t-Value 1.67 0.55 2.05* 6.22*** 1.20
VIF 3.7 1.19 2.09 2.97 1.48
n 44
Adj-Rq 0.3278
F-value 5.19**

Panel II: ROA (ROAt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0028 0.0007 0.0012 0.033 >0.0001
t-Value 0.86 0.17 0.23 0.99 2.39*
VIF 3.3 1.3 2.03 2.59 1.43
n 46
Adj-Rq 0.0145
F-value 0.87

Panel III: PER (PERt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.1136 0.0379 0.0417 0.3844 0.0002
t-Value 2.44* 0.8 0.68 0.76 1.28
VIF 3.3 1.3 2.02 2.65 1.45
n 45
Adj-Rq 0.4029
F-value 6.94***

Panel IV: Tobin’s Q (Tobin’s Qt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0725 0.0183 0.0129 0.0689 >0.0001
t-Value 3.96*** 1.06 0.51 0.24 0.34
VIF 3.56 1.32 2.03 2.86 1.52
n 44
Adj-Rq 0.4605
F-value 8.34***

ROE represents return on equity measured by ln(1 + OBIDA/total shareholder’s equity); ROA represents return on asset measured by ln(1 + OBIDA/total asset); PER represents
price-earnings ratio measured by ln(fiscal year end price/OBIDA per share); Tobin’s Q is measured by ln(market value/book value); PCSR represent positive CSR activities
measured by KLD STAT; NCSR represents negative CSR activities measured by KLD STAT; SIZE represents a firm size measured by ln(sales); LEVERAGE represents a firm’s
capital structure measured by debt-to-asset ratio; MARKET represents market condition measured by average monthly S&P500 Index for each year, and subscript t indicates
the time period. t-Values are calculated using Newey–West standard errors.
*
Significance level of 0.05.
**
Significance level of 0.01.
***
Significance level of 0.001.
 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82 79

Table 5
Summary of pooled regression analysis for casino industry.

PCSR NCSR SIZE LEVERAGE MARKET

Panel I: ROE (ROEt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)

Coefficient 0.0206 0.0249 0.0376 0.3727 0.0002
t-Value 1.14 1.64 4.43*** 1.71 2.57*
VIF 1.17 1.03 1.41 1.31 1.07
n 58
Adj-Rq 0.541
F-value 14.44***

Panel II: ROA (ROAt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0027 0.0066 0.0142 0.068 >0.0001
t-Value 0.42 1.24 5.06*** 1.62 3.09**
VIF 1.23 1.15 1.25 1.32 1.13
n 55
Adj-Rq 0.5033
F-value 11.95***

Panel III: PER (PERt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0179 0.0059 0.0228 0.8261 0.0009
t-Value 0.25 0.13 0.25 1.44 2.64*
VIF 1.68 1.04 1.68 1.18 1.20
n 48
Adj-Rq 0.2774
F-value 4.61**

Panel IV: Tobin’s Q (Tobin’s Qt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0107 0.0483 0.0351 1.2755 0.0001
t-Value 0.16 1.75 1.34 6.57*** 0.56
VIF 1.2 1.02 1.34 1.22 1.08
n 53
Adj-Rq 0.4605
F-value 9.88***

ROE represents return on equity measured by ln(1 + OBIDA/total shareholder’s equity); ROA represents return on asset measured by ln(1 + OBIDA/total asset); PER represents
price-earnings ratio measured by ln(fiscal year end price/OBIDA per share); Tobin’s Q is measured by ln(market value/book value); PCSR represent positive CSR activities
measured by KLD STAT; NCSR represents negative CSR activities measured by KLD STAT; SIZE represents a firm size measured by ln(sales); LEVERAGE represents a firm’s
capital structure measured by debt-to-asset ratio; MARKET represents market condition measured by average monthly S&P500 Index for each year, and subscript t indicates
the time period. t-Values are calculated using Newey–West standard errors.
*
Significance level of 0.05.
**
Significance level of 0.01.
***
Significance level of 0.001.

Tobin’s Q (t-value = 4.13) at the 0.1% level. According to the result (Tobin’s Q) growth rate after accounting for the effects of other
of the regression analysis, one unit increase in PCSR generates, on variables. SIZE has a significantly negative relationship with ROA
average, a 9.7% decrease of ROA growth rate, while one unit and Tobin’s Q, and LEVERAGE (MARKET) shows a significant and
increase in NCSR creates, on average, a 24.9% (9.8%) decrease of PER negative (positive) relationship only with ROA.

Table 6
Summary of pooled regression analysis for restaurant industry.

PCSR NCSR SIZE LEVERAGE MARKET

Panel I: ROE (ROEt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)

Coefficient 0.0001 0.0084 0.0373 0.1429 <0.0001
t-Value 0.02 1.58 4.24*** 5.12*** 2.71**
VIF 1.84 1.37 1.9 1.13 1.3
n 132
Adj-Rq 0.4456
F-value 22.05***

Panel II: ROA (ROAt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0012 0.0004 0.0142 0.0506 >0.0001
t-Value 0.42 0.11 3.57*** 2.14* 0.03
VIF 1.61 1.39 1.89 1.18 1.27
n 143
Adj-Rq 0.2891
F-value 12.55***

Panel III: PER (PERt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0782 0.0558 0.0651 1.2681 >0.0001
t-Value 3.46*** 1.88 0.96 3.81*** 0.69
VIF 1.56 1.39 1.78 1.1 1.27
n 150
Adj-Rq 0.4448
F-value 24.87***
80 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82

Table 6 (Continued )
PCSR NCSR SIZE LEVERAGE MARKET

Panel IV: Tobin’s Q (Tobin’s Qt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0828 0.04 0.0044 0.7289 0.0001
t-Value 3.34** 1.18 0.09 3.08** 0.75
VIF 1.58 1.35 1.74 1.1 1.25
n 156
Adj-Rq 0.3496
F-value 17.66***

ROE represents return on equity measured by ln(1 + OBIDA/total shareholder’s equity); ROA represents return on asset measured by ln(1 + OBIDA/total asset); PER represents
price-earnings ratio measured by ln(fiscal year end price/OBIDA per share); Tobin’s Q is measured by ln(market value/book value); PCSR represent positive CSR activities
measured by KLD STAT; NCSR represents negative CSR activities measured by KLD STAT; SIZE represents a firm size measured by ln(sales); LEVERAGE represents a firm’s
capital structure measured by debt-to-asset ratio; MARKET represents market condition measured by average monthly S&P500 Index for each year, and subscript t indicates
the time period. t-Values are calculated using Newey–West standard errors.
*
Significance level of 0.05.
**
Significance level of 0.01.
***
Significance level of 0.001.

Table 7
Summary of pooled regression analysis for airline industry.

PCSR NCSR SIZE LEVERAGE MARKET

Panel I: ROE (ROEt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)

Coefficient 0.0231 0.0232 0.0389 0.032 <0.0001
t-Value 1.37 1.09 1.15 0.13 1
VIF 1.83 1.56 1.25 1.32 1.03
n 60
Adj-Rq 0.2097
F-value 4.13**

Panel II: ROA (ROAt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.097 0.0003 0.0349 0.3095 <0.0001
t-Value 2.99** 0.08 5.98*** 8.28*** 2.83**
VIF 1.9 1.54 1.17 1.39 1.08
n 65
Adj-Rq 0.7611
F-value 41.78***

Panel III: PER (PERt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.0675 0.2494 0.0319 0.2943 0.0001
t-Value 1.23 3.11** 0.43 0.47 0.39
VIF 1.85 1.54 1.23 1.34 1.04
n 65
Adj-Rq 0.4355
F-value 10.87**

Panel IV: Tobin’s Q (Tobin’s Qt = a0 + a1 PCSRt + a2 NCSRt + a3 SIZEt + a4 LEVERAGEt + a5 MARKETt + et)
Coefficient 0.047 0.0986 0.1095 0.2154 <0.0001
t-Value 1.59 4.13*** 3.27** 0.79 0.48
VIF 1.85 1.47 1.22 1.34 1.03
n 70
Adj-Rq 0.477
F-value 13.59***

ROE represents return on equity measured by ln(1 + OBIDA/total shareholder’s equity); ROA represents return on asset measured by ln(1 + OBIDA/total asset); PER represents
price-earnings ratio measured by ln(fiscal year end price/OBIDA per share); Tobin’s Q is measured by ln(market value/book value); PCSR represent positive CSR activities
measured by KLD STAT; NCSR represents negative CSR activities measured by KLD STAT; SIZE represents a firm size measured by ln(sales); LEVERAGE represents a firm’s
capital structure measured by debt-to-asset ratio; MARKET represents market condition measured by average monthly S&P500 Index for each year, and subscript t indicates
the time period. t-Values are calculated using Newey–West standard errors.
*Significance level of 0.05.
**
Significance level of 0.01.
***
Significance level of 0.001.

6. Discussion
The main purpose of this study is to examine separate impacts
of positive and negative CSR activities on accounting profitability
and firm value across four industries in the hospitality field,
theoretically based on the positivity and negativity effects. The
study analyzed sampled hotel, casino, restaurant and airline firms
separately to investigate industry specificity of impact of positive
and negative CSR, and the result supports the literature asserting
industry idiosyncrasies in terms of CSR and the need for
comparative studies across industries (Amato and Amato, 2007;
Banerjee et al., 2003; Brammer and Millington, 2005; Sweeny and
Coughlan, 2008).
In summary, analysis for the hotel and the restaurant industry
shows a positive impact of positive CSR activities (and no
significant impact of negative CSR activities) on firm value
measured by PER and Tobin’s Q, while it does not reveal any
significant impact of positive and negative CSR on profitability. In
contrast, for the airline industry, the study found the existence of
a negative impact of positive CSR activities (and no significant
impact of negative CSR activities) on profitability; whereas,
negative CSR activities have a negative impact (and no significant
impact of positive CSR activities) on firm value. These findings
can contribute to managements’ strategic decision-making for
implementing CSR by providing evidence that CSR is helpful in
enhancing firm value through increasing socially responsible
 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82
activities (for hotels and restaurants) or decreasing socially
irresponsible activities (for airlines). According to the results, in
the casino industry, no particular relationship seems to exist
between CSR activity and financial performance. The findings of
positive impact of positive CSR and no effect of negative CSR on
firm value in the hotel industry supports the positivity effect
while the presence of negative impact of negative CSR (and no
significant impact of positive CSR) on firm value for airlines
supports the negativity effect. In addition, a negative impact of
positive CSR on ROA of airlines, found in this study, may imply a
case in which CSR is a detriment to short-term profit
maximization. This implication supports Friedman’s (1970)
caution that all business resources should be designed and used
to increase profit.
According to the study’s findings, hotels and restaurants, in
establishing a CSR strategy, should be cautious about the
possibility of the tendency that more weights are placed on
positive CSR activities than negative CSR activities by stakeholders
including consumers, employees, suppliers, communities and/or
investors that can influence firm value. Therefore, for hotels and
restaurants, concentrating on increasing positive CSR activities
rather than reducing negative ones would be a more appropriate
CSR strategy. More specifically, they should prepare CSR invest-
ment plans considering long-term effect that compensates the
implementation with firm value evaluated by market rather than
expecting short-term profitability.
On the other hand, airlines should attend to the negativity
effect in terms of CSR’s influence on firm value; negative CSR
activities rather than positive ones create a greater impression or
garner greater attention by stakeholders and investors in the
market. Thus, the more effective strategy for airlines would be to
focus on reducing negative CSR activities than enhancing
positive CSR activities to increase firm value. It seems that
more frequent positive CSR activities (in the study, the number of
positive CSR is 184 while the number of negative CSR is 160 in
the sampled airlines) has been offset by more influencing
negative CSR activities when market evaluates firm value of
airlines. Or, it is probable that entities that influence airlines’
firm value only respond negatively to airline firms’ socially
irresponsible activities while they are indifferent to socially
responsible activities. According to study findings, airlines may
have to reduce, especially, employee-related negative CSR
activities such as carrying out inadequate retirement benefit
program and poor union relations because employee relation-
ship concerns account for 42.9% of total negative CSR activities
(from KLD STAT rating). Additionally, findings suggest that an
improper CSR strategy would be investing in positive CSR
activities if airlines attempt to improve their short-term financial
performance such as ROA, because the result of this study
suggests that an additional positive CSR activity will reduce the
growth rate of ROA by 9.7%.
Regarding casinos, the study could not find any significant
relationship between positive and negative CSR activities, and
financial performance; stakeholders related with casinos’ profit-
ability and firm value may not be significantly influenced by CSR
activity. Possible reasons for the findings could be that customers
may not visit MGM Mirage more frequently because MGM
considers women and minorities more carefully (positive CSR
activity) nor gamble less money if MGM awards notably high
levels of compensation to its top management (negative CSR
activity). Also, employees may not be willing to remain and be
more loyal to a casino because the casino encourages employee
involvement (positive CSR) but may move to a competitor
company easily for higher compensation. Consequently, investors
and analysts may expect no influence of casinos’ CSR activities on
firm value.
81
7. Limitations and suggested future research
The study is not free from external validity problems caused by
a restricted sampling frame and small sample size. Due to the
feasibility of collecting CSR and financial data, the study includes in
the sample only publicly traded firms in US market. To enhance
generalizability, future research using increased numbers of
observations including international and private hospitality firms
is encouraged. In addition, this study used secondary data, which
might generate bias; future CSR research employing primary data
and qualitative data that can represent, for example, the
perspective of the firm, would complement this study. Regarding
CSR data, KLD STAT data places an equally weighted binary score
on CSR activities that may have dissimilar natures and different
amounts of influence across diverse industries. And, remaining
unclear is whether or not KLD STAT includes the economic domain
of CSR which is defined as activities intended to have either a direct
or indirect positive economic impact on the corporation (Schwartz
and Carroll, 2003). Future studies may develop a method that
provides its own weight for each CSR activity and employ data that
more clearly and comprehensively encompass CSR domains.
To address the mechanism of how CSR impacts financial
performance in a specific industry, examining relevant mediators
would be a valid research approach. For example, one hypothesis
may be that positive CSR activities enhance brand images of hotels,
and then, improved brand images influence hotels’ financial
performances by attracting more customers. Or, environmentally
unsound activities of restaurants may bring about a negative
relationship with community or government, which in turn
adversely impacts financial performance due to possible boycotts
or penalties. Investigating mediators can reveal specific key
vehicles of CSR in each industry which may have different business
structures, profit drivers and main stakeholder groups, thereby
helping to establish and implement more sophisticated CSR
strategy for a specific industry. This suggested future study could
be conducted using structural equation modeling (SEM) that may
provide a greater level of precision and depth in the analysis of the
determinants and consequences of each CSR measure.
References
Abbott, W.F., Monsen, R.J., 1979. On the measurement of corporate social respon-
sibility: self-reported disclosures as a method of measuring corporate social
involvement. Academy of Management Journal 22 (3), 501–515.
Ahluwalia, R., Burnkrant, R.E., Unnava, H.R., 2000. Consumer response to negative
publicity: the moderating role of commitment. Journal of Marketing Research
37 (2), 203–214.
Air Transport Action Group, n.d. Carbon Offsetting. Retrieved September 2, 2008,
from http://www.enviro.aero/Carbonoffsetting.aspx.
Alexander, G.J., Buchholz, R.A., 1978. Corporate social responsibility and stock
market performance. Academy of Management Journal 21 (3), 479–486.
Amato, L.H., Amato, C.H., 2007. The effects of firm size and industry on corporate
giving. Journal of Business Ethics 72 (3), 229–241.
Anderson, D.R., Sweeney, D.J., Williams, T.A., 2005. Statistics for Business and
Economics. South Western, Mason.
Aragón-Correa, J.A., Hurtado-Torres, N., Sharma, S., Garcı́a-Morales, V.J., 2008.
Environmental strategy and performance in small firms: a resource-based
perspective. Journal of Environmental Management 86 (1), 88–103.
Armstrong, D., 2008. Airline global-warming battle heats up fast. Retrieved
September 2, 2008, from http://www.aviation.com/technology/071206-
aircraft-emissions-environmental-battle.html.
Aupperle, K.E., Carroll, A.B., Hatfield, J.D., 1985. An empirical examination of the
relationship between corporate social responsibility and profitability. Academy
of Management Journal 28 (2), 446–463.
Ball, R., 1978. Anomalies in relationships between securities’ yields and yield-
surrogates. Journal of Financial Economics 6 (2–3), 103–126.
Banerjee, S.B., Lyer, E.S., Kashyap, R.K., 2003. Corporate environmentalism: ante-
cedents and influence of industry type. Journal of Marketing 67 (2), 106–122.
Banz, R.W., 1981. The relationship between return and market value of common
stocks. Journal of Financial Economics 9 (1), 3–18.
BBC News, 2004. HSBC bank to go carbon neutral. Retrieved September 1, 2008,
from http://news.bbc.co.uk/1/hi/business/4071503.stm.
Belsley, D.A., Kuh, E., Welsch, R.E., 1980. Regression Diagnostics: Identifying Influ-
ential Data and Sources of Collinearity. John Wiley & Sons, New York.
82 K.H. Kang et al. / International Journal of Hospitality Management 29 (2010) 72–82
Bird, R., Hall, A.D., Momente, F., Reggiani, F., 2007. What corporate social respon-
sibility activities are valued by the market? Journal of Business Ethics 76 (2),
189–206.
Bragdon, J.H., Marlin, J.T., 1972. Is pollution profitable? Risk Management 19 (4), 9–
18.
Brammer, S., Millington, A., 2005. Corporate reputation and philanthropy: an
empirical analysis. Journal of Business Ethics 61 (1), 29–44.
Brealey, R., Myers, S., 2003. Principles of Corporate Finance, 5th ed. McGraw-Hill/
Irwin, New York.
Business Wire, 2005. GE launches Ecomagination to develop environmental tech-
nologies: company-wide focus on addressing pressing challenges. Retrieved
September 1, 2008, from http://www.genewscenter.com/Content/Detail.
asp?ReleaseID=1992&NewsAreaID=2&MenuSea rchCategoryID=.
Carroll, A.B., 1999. Corporate social responsibility: evolution of a definitional
construct. Business and Society 38 (3), 268–295.
Charles, S.T., Reynolds, C.A., Gatz, M., 2001. Age-related differences and change in
positive and negative affect over 23 years. Journal of Personality and Social
Psychology 80 (1), 136–151.
Chauvin, K.W., Hirschey, M., 1993. Advertising, R&D expenditures and the market
value of the firm. Financial Management 22 (4), 128–140.
Chung, K.H., Pruitt, S.W., 1994. A simple approximation of Tobin’s q. Financial
Management 23 (3), 70–74.
Cordeiro, J.J., Sarkis, J., 1997. Environmental proactivism and firm performance:
evidence from security analyst earning forecast. Business Strategy and the
Environment 6 (2), 104–114.
Engardio, P., Capell, K., Carey, J., Hall, K., 2007. Beyond the green corporation;
imagine a world in which eco-friendly and socially responsible practices
actually help a company’s bottom line. It’s closer than you think. Business
Week 4019, 50.
Fibre2fashion, 2008. Wal-Mart launches Love, Earth jewelry. Retrieved September
1, 2008, from http://www.fibre2fashion.com/news/company-news/walmart/
newsdetails.aspx?news_id=59935.
Fiske, S.T., 1980. Attention and weight in person perception: the impact of negative
and extreme behavior. Journal of Personality and Social Psychology 38 (6), 889–
906.
Folkes, V.S., Kamins, M., 1999. Effects of information about firms’ ethical and
unethical actions on consumers’ attitudes. Journal of Consumer Psychology 8
(3), 243–259.
Folkes, V.S., Patrick, V.M., 2003. The positivity effect in perceptions of services: seen
one, seen them all? Journal of Consumer Research 30 (June), 125–137.
Freeman, R.E., 1984. Strategic Management: A Stakeholder Approach. Prentice Hall,
Englewood Cliffs.
Friedman, M., 1970. The Social responsibility of business is to increase its profits.
New York Times Magazine, September 13.
Frooman, J., 1997. Socially irresponsible and illegal behavior and shareholder
wealth: a meta-analysis of event studies. Business and Society 36 (3),
221–249.
Garcı́a, F., Armas, Y., 2007. Relationship between social-environmental responsi-
bility and performance in hotel firms. International Journal of Hospitality
Management 26 (4), 824–839.
Grave, S.B., Waddock, S.A., 1994. Institutional owners and corporate social perfor-
mance. Academy of Management Journal 37 (4), 1034–1046.
Gujarati, D.N., 2003. Basic Econometrics, 4th ed. McGraw-Hill, Boston.
Gujarati, D.N., 2006. Essentials of Econometrics, 3rd ed. McGraw-Hill, Boston.
Hamilton, D.L., Huffman, L.J., 1971. Generality of impression formation processes for
evaluative and nonevaluative judgments. Journal of Personality and Social
Psychology 20 (2), 200–207.
Hart, S.L., 1995. A natural-resource-based view of the firm. Academy of Manage-
ment Review 20 (4), 986–1014.
Hart, S.L., Ahuja, G., 1996. Does it pay to be green? An empirical examination of the
relationship between emission reduction and firm performance. Business
Strategy and the Environment 5 (1), 30–37.
Heinze, D.C., 1976. Financial correlates of a social measure. Akron Business and
Economic Review 7 (1), 48–51.
Henisz, W.J., Zelner, B.A., 2001. The institutional environment for telecommu-
nications investment. Journal of Economics & Management Strategy 10 (1),
123–147.
Herr, P.M., Kardes, F.R., Kim, J., 1991. Effects of word-of-mouth and product-
attribute information of persuasion: an accessibility–diagnosticity perspective.
Journal of Consumer Research 17 (4), 454–462.
Holme, R., Watts, P., 2000. Corporate social responsibility: Making good business
sense. Retrieved September 3, 2008, from http://www.wbcsd.org/DocRoot/
IunSPdIKvmYH5HjbN4XC/csr2000.pdf.
Howell, D.C., 1992. Statistical Methods for Psychology, 3rd ed. Duxbury Press,
Belmont.
International Air Transport Association, n.d. Environment: sustainability. Retrieved
September 2, 2008, from http://www.iata.org/whatwedo/environment/
sustainability.htm.
Jamal, A.M.M., Yu, H., 2007. Test of Taylor rule and policy implications. Atlantic
Economic Journal 35 (1), 121–122.
Jensen, M.C., 2001. Value maximisation, stakeholder theory, and the corporate
objective function. European Financial Management 7 (3), 297–317.
Johnson, H.H., 2003. Dose it pay to be good? Social responsibility and financial
performance. Business Horizons 46 (6), 34–40.
Judge, W.Q., Douglas, T.J., 1998. Performance implications of incorporating natural
environmental issues into the strategic planning process: an empirical assess-
ment. Journal of Management Studies 35 (2), 241–262.
Kirk, D., 1995. Environmental management in hotels. International Journal of
Contemporary Hospitality Management 7 (6), 3–8.
Kjelaggard, C., 2007. Flying eco-friendly. Retrieved September 2, 2008, from http://
www.aviation.com/technology/070803_environmentally_friendlier.html.
Klassen, R.D., McLaughlin, C.P., 1996. The impact of environmental management on
firm performance. Management Science 42 (8), 1199–1214.
Klein, J.G., 1996. Negativity in impression of presidential candidates revisited: the
1992 election. Personality and Social Psychology Bulletin 22 (March), 289–296.
Lampel, A.K., Anderson, N.H., 1968. Combining visual and verbal information in an
impression-formation task. Journal of Personality and Social Psychology 9 (1),
1–6.
Lee, S., Park, S., 2009. Do socially responsible activities help hotel and casino achieve
their financial goals? International Journal of Hospitality Management 28 (1),
105–112.
Mather, M., Carstensen, L.L., 2005. Aging and motivated cognition: the positivity
effect in attention and memory. Trends in Cognitive Sciences 9 (10), 496–502.
McConnell, J.J., Servaes, H., 1990. Additional evidence on equity ownership and
corporate value. Journal of Financial Economics 27 (2), 595–612.
McGuire, J.B., Sundgren, A., Schneeweis, T., 1988. Corporate social responsibility and
firm financial performance. Academy of Management Journal 31 (4), 854–872.
McWilliams, A., Siegel, D., 2001. Corporate social responsibility: a theory of the firm
perspective. The Academy of Management Review 26 (1), 117–127.
Newey, W.K., West, K.D., 1994. Automatic leg selection in covariance matrix
estimation. Review of Economic Studies 61 (4), 631–653.
Nicolau, J.L., 2008. Corporate social responsibility: worth-creating activities. Annals
of Tourism Research 35 (4), 990–1006.
Orlitzky, M., Schmidt, F.L., Rynes, S.L., 2003. Corporate social and financial perfor-
mance: a meta-analysis. Organization Studies 24 (3), 403–441.
Pava, M.L., Krusz, J., 1996. The association between corporate social-responsibility
and financial performance: the paradox of social cost. Journal of Business Ethics
15 (3), 321–357.
Post, J.E., Preston, L.E., Sachs, S., 2002. Managing the extended enterprise: the new
stakeholder view. California Management Review 45 (1), 6–28.
Preston, L.E., O’Bannon, D.P., 1997. The corporate social-financial performance
relationship: typology and analysis. Business and Society 36 (4), 419–429.
Russo, M.V., Fouts, P.A., 1997. A resource-based perspective on corporate environ-
mental performance and profitability. Academy of Management Journal 40 (3),
534–559.
Schwartz, M.S., Carroll, A.B., 2003. Corporate social responsibility: a three-domain
approach. Business Ethics Quarterly 13 (4), 503–530.
Skowronski, J.J., Carlston, D.E., 1987. Social judgment and social memory: the role of
diagnosticity in negativity, positivity and extreme biases. Journal of Personality
and Social Psychology 52 (4), 689–699.
Sturdivant, F.D., Ginter, J.L., 1977. Corporate social responsiveness: management
attitude and economic performance. California Management Review 19 (3),
30–39.
Sweeny, L., Coughlan, J., 2008. Do different industries report corporate social
responsibility differently? An investigation through the lens of stakeholder
theory. Journal of Marketing Communication 14 (2), 113–124.
Teoh, S.H., Welch, I., Wazzan, C.P., 1999. The effect of socially activist investment
policies on the financial markets: evidence from the South African boycott.
Journal of Business 72 (1), 35–89.
Tsoutsoura, M., 2004. Corporate social responsibility and financial performance.
Retrieved September 3, 2008, from University of California at Berkley, Hass
School of Business web site: http://www.haas.berkeley.edu/responsiblebusiness
/documents/FinalPaperonCSR_PDFII.pdf.
Yu, H., 2008. The impacts of stock price and country risk on the exchange rate in
Singapore. International Journal of Development Issues 7 (1), 56–61.
Vance, S.C., 1975. Are socially responsible corporations good investment risks?
Management Review 64 (8), 18–24.
Waddock, S.A., Grave, S.B., 1997. The corporate social performance-financial per-
formance link. Strategic Management Journal 18 (4), 303–319.
Winchester, M., Romaniuk, J., Nogomolova, S., 2008. Positive and negative brand
beliefs and brand defection/uptake. European Journal of Marketing 42 (5/6),
553–570.
Wright, P., 1974. The harassed decision maker: time pressure, distraction and use of
evidence. Journal of Applied Psychology 59 (5), 555–561.
Wright, P., Ferris, S.P., 1997. Agency conflict and corporate strategy: the effect of
divestment on corporate value. Strategic Management Journal 18 (1), 77–83.