Date post: | 17-Jul-2016 |
Category: |
Documents |
Upload: | syedmuzammilali |
View: | 215 times |
Download: | 1 times |
Corporate Governance and Returns on Information Technology Investment:
Evidence from an Emerging Market
Joanna L.Y. Ho* The Paul Merage School of Business
University of California Irvine, California
Anne Wu School of Accountancy
National Chengchi University, Taipei, Taiwan
Sean Xin Xu School of Business and Management
Hong Kong University of Science & Technology Clear Water Bay, Hong Kong
*Contacting author
We would like to thank Daphne Chang for her research assistance. This is a preliminary draft. Please do not quote without consent of the authors.
1
Corporate Governance and Returns on Information Technology Investment:
Evidence from an Emerging Market
Abstract Prior studies have reported mixed findings on the impact of corporate IT investment on firm performance. This study investigates the role of corporate governance, an important management control system, in the IT investment-firm performance relationship in the Taiwanese high-tech industries. Two specific corporate governance factors, i.e., board independence and foreign ownership, are explored across firms of different sizes and in industries whose degrees of competitiveness run a wide gamut. Our results show a positive moderating effect of board independence on the IT investment-firm performance relationship, especially when competition intensifies. Furthermore, we find that the association between foreign ownership and firm performance is negative but not significant. Yet, foreign ownership is a significant positive moderator for small firms, suggesting that foreign investors may bring IT expertise to help small firms reap the benefits of using IT. Keywords: IT investment; Board independence; Foreign ownership; Firm performance; Firm size;
Industry competitiveness
2
1. Introduction
Companies invest substantial financial resources in information technology (IT), and the
significant trend of IT investment is expected to continue, especially in emerging markets (IDC
2006).1 However, it is puzzling that despite significant IT investment, there is no consistent
evidence of IT value. A meta-analysis of 66 prior studies (Kohli and Devaraj 2003) shows that
IT investment improves, reduces, or has no effect on firm performance. Particularly in emerging
markets, positive IT returns are yet to be documented (Dewan and Kraemer 2000). The mixed
evidence on IT returns has triggered significant debate on the IT investment-firm performance
relationship (see Brynjolfsson et al. 2002 for a literature review).
Recent research has largely agreed that IT investments seldom act alone, but work in
conjunction with other corporate resources to create value for shareholders (e.g., Brynjolfsson et
al. 2002; Bresnahan et al. 2002; Wade and Hulland 2004). However, corporate governance
structures, one of the important corporate resources, have received little attention in the literature
on IT returns (e.g., Bresnahan et al. 2002; Brynjolfsson et al. 2002). Corporate governance2
should theoretically play a significant role in the IT investment-firm performance (IT-
performance) relationship. The Committee of Sponsoring Organizations of the Treadway
Commission (1992) argues that there is a strong relationship between management control
philosophy and corporate governance. In the context of IT investment, shareholders expect that
utilizing IT can improve firm performance by increasing operational efficiencies or productivity
(e.g., Hitt and Brynjolfsson 1996; Devaraj and Kohli 2003; Banker et al. 2006). However, IT
managers may over-invest to accumulate excessive IT resources for their own interests (e.g., 1 In the U.S., corporate IT investment amounts to $417.4 billion in 2005, representing 3.35% of GDP or about half of the total capital spending in the private sector; worldwide IT spending is projected to increase at a compound annual growth rate of 6% to reach $1.5 trillion in 2010 (IDC 2006). Among all the regions in the world, the Asia/Pacific Region is experiencing the highest growth rate of IT spending, about 9% in 2006 (IDC 2006). 2 Corporate governance refers to “the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment” (Shleifer and Vishny 1997, p.737).
3
Kauffman and Li 2003; Carr 2004), and they may imitate each other even if imitation deviates
from the optimal investment decision (e.g., Scharfstein and Stein 1990; Graham 1999).
Corporate governance may be important because it can mitigate the above agency problems (e.g.,
La Porta et al. 2000; Bushman and Smith 2001).
Specifically, this study addresses the role of corporate governance in the high-tech
industries in Taiwan, an emerging market. We focus on this research setting for both theoretical
and empirical reasons. Theoretically, corporations in emerging markets generally face weak
judicial systems to protect shareholders’ rights (e.g., Gugler et al. 2003). This may lead these
companies to resort to corporate governance mechanisms (e.g., boards of directors) to align
managerial interests with those of shareholders (e.g., Shleifer and Vishny 1997; La Porta et al.
2000). Moreover, prior research in emerging markets reports wide firm-level heterogeneity in
the forms and the effectiveness of corporate governance (Fan and Wong 2005). Therefore,
corporate governance could be a useful lens to understand different IT returns in emerging
markets.
Empirically, starting in 2001, the Taiwanese government conducted annual surveys of the
top 2,000 companies to better understand their IT investment (e.g., software, hardware, and
training costs). Corporate governance information for all Taiwanese publicly listed firms was
also made available in a Corporate Database. These Taiwanese datasets provide a unique
opportunity to use archival data to assess the role of corporate governance in IT returns.3 This is
different from a number of previous studies using primary survey data to explore the IT-
performance relationship, which could involve biases (Banker et al. 2006). Using a meta-
analysis, Kohli and Devarj (2003) find that studies utilizing primary survey data are more likely
3 In contrast, information concerning IT investment and corporate governance of U.S. companies are not available in public datasets.
4
to report positive IT performance impacts than those using archival data. We use the Taiwanese
datasets partly because they can attenuate the above biases and partly because the Taiwanese
high-tech industries play a leading role in the world market.4
We first investigate board independence—i.e., the presence of outside directors on the
corporate board. It is widely argued that independent boards can better perform a monitoring
role, and thus can reduce agency problems in IT investment (e.g., Kaplan and Reishus 1990;
Dalton et al. 1998; Klein 2002). We extend this line of research by examining whether or not
board independence is contingent on industry competition. Theoretical research implies that
competition may shape the role of corporate governance. That is, competition by itself might
serve as a bonding mechanism to mitigate agency problems, or intense competition may make
the monitoring mechanisms of corporate boards more desirable (e.g., Scharfstein 1988; Schmidt
1997). These competing perspectives motivate us to investigate the role of board independence
across industries.
In addition, considering the specific features of emerging markets, we investigate one
dimension of a firm’s ownership structure, foreign ownership—i.e., the share of foreign
investment in a firm’s common stock. Foreign ownership may play an important role in our
research setting because firms in emerging markets typically lack IT experience and management
expertise, a significant barrier to IT returns (Dewan and Kraemer 2000). Therefore, foreign
investors’ knowledge and experience with global technology may benefit companies in emerging
markets in deploying IT, which is the so-called “spillover” (of IT management expertise) (Aitken
and Harrison 1999). However, foreign investors’ general experience in the host countries may
be domain specific, i.e., not compatible with local management style or not suitable for the local
4 Global outsourcing has made Taiwan the largest producer of desktop PCs, notebooks, displays, and motherboards (Dedrick and Kraemer 2005). To illustrate, by 2004, Taiwan had accounted for over 70% of the world’s notebook PC market (Yang 2006).
5
environment, and may not help improve firm performance. Furthermore, Short (1994) argues
that foreign ownership may increase ownership dispersal and, therefore, exacerbate agency
problems. Motivated by such a tension between IT expertise spillover and exacerbated agency
problems, we attempt to explore the role of foreign ownership in IT returns.
Using a sample of 719 Taiwanese high-tech companies from 2001 to 2005, we find a
positive moderating role of board independence, i.e., the higher the board independence, the
more positive the IT-performance relation. This moderation effect becomes even stronger in
more competitive industries, suggesting that the value of corporate boards’ monitoring function
increases as competition intensifies. Also, we find the impact of foreign ownership on the IT-
performance relationship only in small firms, indicating that foreign investors can mitigate
resource disadvantages of small firms by bringing IT expertise to help them manage IT more
effectively.
Our study adds to the extant literature on IT returns in three different ways. First, the
results offer insights into why the mixed findings are reported in the prior literature on the IT-
performance relationship. Our results show an insignificant relationship between firm
performance and IT investment per se. That relationship, however, is moderated by corporate
governance. Clearly, a better monitoring system can help companies achieve better financial
performance through IT investment. Also, foreign ownership may have a spillover effect to help
small firms attain IT benefits. These results suggest that future research should incorporate
corporate governance in exploring how IT investment affects firm performance.
Second, this work has an important implication for research on board composition.
Based on a review of 54 empirical studies, Dalton et al. (1998) report no association between
board composition and financial performance. Therefore, they conclude “we are not optimistic
6
that future research in the general areas of board composition/financial performance…would be
fruitful. Also, the evidence would not seem to provide much confidence in further examinations
of possible moderating influences on those relationships” (p.284). Yet, this study finds that
while the direct relationship of board independence to firm performance is not clear, the
interaction between board independence and IT investment is significantly associated with better
firm performance. As such, we believe that investigating the role of board composition in
specific investment activities should be a promising avenue for future research.
Third, our results show contingency effects of firm size and industry competition in
examining corporate governance. In particular, we find that the moderation effect captured by
board independence is most robust in intensely competitive industries. One implication is that
controlling principle/agent conflicts through board monitoring may be more beneficial for firms
in more competitive industries. The relationship between competition and monitoring has
received attention in modeling papers (see Karuna 2006 for a literature review), but the literature
fall short of empirical evidence. Our work thus contributes to this burgeoning literature by
relating effects of corporate governance to industry competition.
The rest of this paper is organized as follows. Section 2 discusses related literature and
hypotheses development. Section 3 describes the method, including our sample data and models,
followed by empirical results in Section 4. Lastly, Section 5 provides a summary and conclusion.
2. Hypotheses Development
This study explores two factors concerning corporate governance, board independence
and foreign ownership, which may affect the IT-performance relationship in the Taiwanese
context.
7
2.1. Board independence
The principal/agent literature suggests that managers often act in their self-interest at the
expense of the firm’s value (e.g., Feltham and Xie 1994; Lambert 2001). That is, while
shareholders want managers to work to improve firm performance, managers may maximize
their own utility by over-consuming company resources or selecting suboptimal investments
(e.g., Fama and Jensen 1983).
In our research context, managers may make excessive IT investments, because a large
IT department brings power and high salary (Carr 2004). In addition, the literature of career
concerns suggests that managers may intentionally imitate others’ investment decisions to
enhance their professional reputations (e.g., Scharfstein and Stein 1990; Graham 1999). These
agency problems may occur, because evaluating managers’ IT investment decisions requires
technology-related information (e.g., costs, benefits, risks, implementation complexity, and
future development of the technology, and the company’s own technical capability), but
shareholders often lack such information. As such, shareholders may evaluate IT managers’
decisions by industry consensus (Kauffman and Li 2003). Consequently, managers may take the
advantage of asymmetric information to support their investment decisions (e.g., Fama and
Jensen 1983), and they tend to follow others to maintain their reputations rather than assessing
economic benefits for their companies (Graham 1999). Also, firms in emerging markets seek to
catch up to the global industry by moving along the IT diffusion curve and have been
increasingly allocating financial resources to IT (IDC 2006). This further allows managers to
use the industry consensus of investing in IT to justify their investment decisions. These agency
problems may cause the IT-performance relationship to be insignificant and even negative.
8
The corporate board of directors serves as an important internal mechanism in making
significant managerial decisions and in limiting managerial inefficiencies (e.g., Daily and Dalton
1994; Zahra and Pearce II 1989). The corporate board directly monitors IT investment when the
investment volume is large (Klein 2002). This is supported by prior research on corporate
investment in research and development (R&D) (e.g., Baysinger et al. 1991). In addition, the
board of directors meets regularly with internal and external auditors to review the firm’s
financial statements, audit process, and internal controls. This creates an indirect monitoring
function for IT investment. Thus, the board’s monitoring function helps reduce the agency
problems.
The corporate board consists of inside directors (those employed by the company) and
outside directors (e.g., CEOs of other firms, investment bankers, former governmental officials,
major suppliers). An outsider-majority board is associated with high independence, while an
insider-majority board is associated with low independence (e.g., Dalton et al. 1998; Klein 2002).
It has been argued that an independent board can monitor more effectively (e.g., Fama and
Jensen 1983; Kaplan and Reishus 1990; Klein 2002). One possible explanation is that outside
directors are more objective and have access to external information that is less available to
inside directors (e.g., Sutton and Callahan 1987; Daily and Dalton 1994). Furthermore, Beasley
et al. (2000) argue that outside directors have incentives to protect their reputation and avoid
litigation. Prior studies have shown that an increase in board independence improves earnings
quality (e.g., Dechow et al. 1996), lowers the magnitude of abnormal accruals (e.g., Klein 2002),
and also decreases the likelihood of financial fraud (e.g., Beasley et al. 2000). These studies
support the notion of better monitoring by outsider-majority boards.
9
To sum up, the agency problems may weaken the IT-performance relationship; however,
board independence may reduce the agency problems and thus enhance the IT value. Hence, we
predict a positive moderation effect of board independence as follows:
H1: Board independence positively moderates the relationship between IT investment and firm performance.
2.2. Industry competitiveness and board independence
The above moderation effect (H1) may differ across industries with different degrees of
competitiveness. The industrial organization literature commonly uses industry concentration as
a proxy for industry competitiveness (e.g., Cohen and Levin 1989; Waring 1996). Schumpeter
(1970) stresses that a high level of industry concentration (and therefore low competitiveness)
increases a firm’s profitability due to insulation of competition, and thus decreases the marginal
value of new technology investment. In contrast, industry competitiveness increases the
marginal gains to the firm from new technology (e.g., Levin et al. 1987; Cohen and Levin 1989;
Morrison 1997). Melville et al. (2004) also argue that firms in highly concentrated industries, by
attaining monopoly rents, may achieve superior performance without IT-enhanced efficiency
gains. Conversely, under competitive regimes, firms need to rely on other approaches to greater
profitability, e.g., using IT to increase efficiencies in business processes (e.g., Melville et al.
2004; Banker et al. 2006).
However, as discussed earlier, agency problems must be alleviated in order to convert
IT’s potential benefits to realized performance improvement. This is particularly important for
firms facing intense competition (e.g., Hermalin 1992; Schmidt 1997). That is, the marginal
value to the firm of reducing agency problems is higher when the firm is operating in more
competitive industries than in less competitive industries. Hence, the board’s monitoring would
10
benefit firms more as industry competitiveness becomes more intensive.
In another stream of research on managerial incentives, several theoretical papers imply
that increased competition may work as a bonding mechanism to decrease agency problems, by
providing more performance evaluation information (e.g., Nalebuff and Stiglitz 1983). This is
because intense competition makes firm performance more sensitive to uneconomic investments.
However, other studies suggest that the effect of competition may depend on managers’ utility
function (e.g., Karuna 2006). For instance, Scharfstein (1988) shows that competition actually
increases managerial actions for self-interest when the marginal utility gained from self-interest-
seeking is strictly positive. Taken together, the predicted relationships between competition and
managerial incentives are not clear.
More importantly, it is difficult for shareholders to relate changes in firm performance to
IT investment since disclosure of IT investment is not a requirement. This may weaken the
effect of competition, if any, to constrain principal/agent conflicts in IT investment. In fact,
Abrahamson and Rosenkopf (1997) find evidence suggesting that, under competition, companies
tend to make IT investments by following suit, without a thorough cost-benefit analysis. This
underscores the need for corporate board monitoring. Therefore, we expect the monitoring
function of independent boards, in the IT-performance context, to be more important in more
competitive industries.
H2: The moderation effect of board independence in the IT investment-firm performance relationship is more positive for more competitive industries than for less competitive industries.
2.3. Foreign ownership
Associated with the global sourcing of computers, the shipment of the Taiwanese high-
tech industries has increased significantly on the world platform, which has attracted foreign
11
investors to the Taiwanese market.5 Foreign ownership may influence firm performance in
different ways.
Foreign ownership may be negatively related to firm performance due to ownership
dispersal (e.g., Short 1994). The entry of foreign ownership increases the number of
shareholders, which dilutes share ownership and thereby lessens the incentive for each
shareholder to monitor the agents, consequently lowering firm performance (e.g., Boardman et al.
1997). Also, when foreign investors provide managerial suggestions to top management, their
involvement may not benefit the firm due to their limited knowledge of the local environment,
culture, and management style. Furthermore, Simunic (1980) reports a positive relationship
between foreign ownership and auditing fees. Consequently, firm performance may be lowered
by these additional agency costs, suggesting a negative relation between foreign ownership and
firm performance.
Conversely, Aitken and Harrison (1999) point out that foreign ownership may bring
about a positive moderation (“spillover”) effect in IT investment. Foreign ownership may offer
expertise for deploying IT resources, which helps bridge the gap of IT performance impacts
between developing and developed markets (e.g., Dewan and Kraemer 2000). In the IT
investment setting, foreign investors’ knowledge and experience is technical in nature, which is
not subject to the agency costs associated with their advice in general management. This view
falls into the scope of literature on technological knowledge spillover and foreign investment in
underdeveloped economies (e.g., Aitken and Harrison 1999). For instance, prior research
proposes that “in small economies, especially underdeveloped economies, most technological
knowledge will come from abroad” (Braga and Willmore 1991, p.431), and empirical evidence
5 Starting in 1991, the Taiwanese government approved the Qualified Foreign Institutional Investors (QFII) program, allowing qualified foreign investors to participate in the Taiwanese securities market. Over the past decade, foreign ownership has gained increasing importance in Taiwanese high-tech industries (Chiang and Kuo 2006).
12
supports that foreign investment is one approach to knowledge transfer (e.g., Blomström 1986;
Braga and Willmore 1991; Aitken and Harrison 1999). This discussion leads us to expect a
positive moderating effect of foreign ownership on the IT value.
H3: Foreign ownership positively moderates the relationship between IT investment and firm performance.
2.4. Firm size and foreign ownership
The outcome of new technology deployment may be contingent on certain organizational
characteristics such as firm size (e.g., Damanpour 1996). Compared to small firms, large firms
have more expertise in deploying IT resources, because they can afford to hire major consulting
firms to provide specialized services and offer expert advice (Damanpour 1996). Large firms are
also obliged to have more comprehensive internal controls and auditing systems in place (e.g.,
Klein 2002), which may mitigate agency problems. Therefore, the impacts of foreign ownership
(both the agency costs and the “spillover” benefits) are expected to be insignificant for large
firms.
In contrast, smaller firms are afflicted with “resource poverty” in deploying IT (e.g.,
Harrison et al. 1997; Thong 1999). These small firms generally lack IT knowledge and skills,
which is a significant barrier to IT returns. Furthermore, CEOs and directors of small firms are
usually less constrained by organizational systems and structures and therefore may have more
discretion compared to their large firm counterparts (e.g., Eisenhardt and Schoonhoven 1990;
Daily and Dalton 1994). The weak internal control system and lack of a comprehensive audit
system may curtail potential returns on IT investment. Since foreign investors can contribute
their IT expertise and enhance board structure, foreign ownership may strengthen a small firm’s
ability to generate value from deploying IT. Indeed, Aitken and Harrison (1999) find that the
13
spillover effect of foreign ownership is significant for small firms but not large firms. As such,
we expect that small firms benefit more from the “spillover” of IT expertise from foreign
ownership than large firms do. This discussion leads to our last hypothesis:
H4: The moderation effect of foreign ownership in the IT investment-firm performance relationship is more positive for small firms than for large firms.
3. Model Development
3.1. Data
Our sample consists of 719 manufacturing firms in the electronics industry in Taiwan
from 2001 to 2005. The distribution of industry sectors is shown in Table 1. We obtain data
concerning corporate IT investment from a database created by the Taiwanese Institute for
Information Industry, which is funded by the Taiwanese government. Since 2001, to set IT
policies, the Taiwanese government has asked the Taiwanese Institute for Information Industry
to collect IT investments data from the top 2,000 companies. The data is collected through face-
to-face interviews and then verified for accuracy. On average, the response rate is about 45
percent.
We collect financial performance information from the Financial Report Database
compiled by the Taiwan Economic Journal (TEJ), which contains data extracted from the firms’
annual financial reports. Also, we obtain information on firms’ ownership structures and board
composition from the TEJ’s Corporate Database. These two databases cover listed companies on
the Taiwan Stock Exchange, and therefore our sample contains only public companies.
_________________________
Insert Table 1 about here __________________________
14
3.1.1. Variables in hypotheses
Below, we define the variables involved in hypotheses 1-4:
Firm performance (ROA). Firm performance is measured by return on assets, ROA, the
ratio of net income to the year-end book value of total assets (e.g., Johnstone and Bedard 2004;
Fan and Wong 2005; Verrecchia and Weber 2006). This is a widely used proxy to examine how
firm performance relates to IT investment (e.g., Kohli and Devaraj 2003). Efficiency gains from
using IT may lead to higher ROA (e.g., Melville et al. 2004), but uneconomic IT investment,
expensed or capitalized, may lower ROA. Therefore, the relationship between ROA and IT is
shaped by how the uneconomic investment is controlled.
IT investment (ITINV). The amount of IT investment includes hardware, software, and
costs concerning maintenance, personnel, and training. This proxy covers all major components
of corporate IT spending (e.g., Hitt and Brynjolfsson 1996; Bresnahan et al. 2002). Following
previous research, we divide a firm’s IT investment by its total assets for size adjustment.
Board independence (BIND). As described earlier, board independence increases with
the proportion of outside directors. In this study we use three proxy measures. First, outside
directors include affiliated and non-affiliated directors. Affiliated directors include past
managers, relatives of current managers, affiliated attorneys, and consultants to the firm, while
non-affiliated directors are those without such affiliations (e.g., Klein 1998). A conservative
measure is the proportion of non-affiliated directors (NONAFF), which excludes the proportion
of affiliated directors (AFF) (e.g., Yermack 1996; Klein 1998). Other research, however, argues
that affiliated directors may often be substantially independent, and thus both affiliated and non-
affiliated directors should be considered (e.g., Bhagat and Black 2002). Therefore, following
prior studies (e.g., Dalton et al. 1998; Bhagat and Black 2002), we also use the proportion of all
15
outside directors (OUT), i.e., the sum of AFF and NONAFF, to measure board independence.
Third, considering that the presence of affiliated and non-affiliated directors may result in
different degrees of board independence (Klein 1998), we create a composite index by factor
analysis (e.g., Bens 2002), which assigns weights to NONAFF and AFF and gives us the third
proxy: )65.014.1( AFFNONAFFBIND ×+×= . It is clear that OUT treats NONAFF and AFF
as equally important in establishing board independence, while BIND assigns a relatively higher
weight to NONAFF. Below, we first present regression results based on BIND, and then show
the results based on NONAFF and OUT in sensitivity analysis. As will be discussed later, results
based on the three measures are highly consistent.
Foreign ownership (FOR). We measure FOR by the percentage of common stock owned
by foreign investors. This measure is used in the literature on technological knowledge spillover
with foreign ownership (e.g., Blomström 1986; Aitken and Harrison 1999).
Industry competitiveness (CR4). We compute the four-firm concentration ratio (CR4) for
each firm, which equals the percentage of total sales in the firm’s industry sector accounted for
by the four largest firms in the same sector. CR4 is conversely related to industry
competitiveness (e.g., Cohen and Levin 1989; Waring 1996).
Firm size (ASSETS). Following previous studies, we use total assets to measure firm size
(e.g., Johnstone and Bedard 2004; Engel et al. 2006).
3.1.2. Control variables
To evaluate cross-sectional associations between firm performance and board
composition, we need to control for lagged firm performance (e.g., Klein 1998). In addition, we
need to consider other factors that may explain firm performance. We draw upon the empirical
literature in economics that has modeled firm performance and explanatory factors (e.g., Waring
16
1996). We select a set of key factors based on further review of prior studies on IT investment
and firm performance (e.g., Hitt and Brynjolfsson 1996; Kohli and Devaraj 2003). These control
variables are as follows:
Lagged firm performance (ROA-1). A firm’s board composition could be a function of its
past performance. Weisbach (1988), based on observations in a 12-year period of time, found
that inside directors tended to leave the corporate board after a firm experienced a decrease in
stock market returns. However, Weisbach reported low magnitude of the impact of past
performance on board composition. To be conservative and to control for possible endogeneity
problems between board composition and firm performance, we follow prior research (e.g.,
Klein 1998; Peng 2004) by including the one-year lagged performance (ROA-1). This method is
appropriate because multiple regressions allow for correlations among the explanatory variables
(Klein 1998). We expect a positive coefficient of ROA-1 since performance measures are
positively correlated over time.
Firm growth (GROWTH). Prior studies on the IT-performance relationship have shown a
positive association between sales growth and ROA (e.g., Hitt and Brynjolfsson 1996).
Following the literature (e.g., Chaney and Philipich 2002), we measure GROWTH using the one-
year growth rate of sales and expect a positive coefficient of GROWTH.
Leverage (LEV). A firm’s performance may be influenced by its leverage ratio because
debt financing has both costs and benefits (e.g., Flannery and Rangan 2006). We measure LEV
as the book value of total debt divided by the book value of total equity (Flannery and Rangan
2006). Shareholders may have incentives to expropriate bondholders’ wealth by investing in
high-risk, high-return projects. As such, bondholders may demand higher rents by increasing the
costs of debt (Jensen and Meckling 1976). Conversely, debt financing may “urge” managers to
17
maintain their promises to pay out future cash flows. Thus, increased leverage may reduce cash
flows available for spending at the discretion of managers (Jensen 1986). Since previous
research on IT investment reports a negative coefficient for leverage (e.g., Hitt and Brynjolfsson
1996), we expect a negative relationship between firm performance and LEV.
Market share (SHARE). Following prior studies, we also include SHARE, i.e., the ratio
of a firm’s sales to the total industry sale. The literature contains mixed findings regarding both
the sign and the statistical significance of the market share-firm performance relationship (see a
review by Szymanski et al. 1993). This may be because of the inclusion of some proxy variables
for power, which may weaken the statistical association between market share and firm
performance (e.g., Hitt and Brynjolfsson 1996). In addition, firms with high market shares may
derive no extra negotiation power benefits, and as a result, the efforts devoted to increasing
market share (e.g., marketing costs and capacity build-up) may not pay off (e.g., Szymanski et al.
1993). Hence, we cannot predict the sign of SHARE.
Research and development intensity (RD). Research and development intensity is viewed
as a proxy for a firm’s intangible capital, which may enhance firm performance (e.g., Li and
Wong 2005). We measure RD as a firm’s R&D expenditures divided by sales (e.g., Callen et al.
2005; Darrough and Rangan 2005). While some studies show positive market valuation of R&D
expenditures (Hall 1993), recent research on the IT-performance relationship reports R&D to be
negatively related to performance (e.g., Bharadwaj et al. 1999). Thus, a priori the relationship
between R&D intensity and firm performance is not clear.
3.2. Regression Model
Our model builds on a previous cross-sectional analysis regressing firm performance on
board structure and lagged performance (Klein 1998). We extend the previous model in two
18
ways. First, we study two factors concerning corporate governance—board independence and
foreign ownership, both used as independent variables. Second, we add IT investment as a new
independent variable, which is the point of interest in this study. Our regression model is as
follows:
= β1 ITINV + β2 BIND + β3 (ITINV×BIND) + β4 FOR + β5 (ITINV×FOR)
+ β6 ROA-1 + β7 GROWTH + β8 LEV + β9 SHARE + β10 RD + β0 (1)
where )10,...2,1( =kxk is the kth explanatory variable (in vector terms, =],...,[ 1021 xxx [ITINV,
BIND, ITINV×BIND, FOR, ITINV×FOR, ROA-1, GROWTH, LEV, SHARE, ]RD ),
)10,...2,1( =kkβ is the corresponding regression coefficient, and β0 is an intercept.
The model specifies direct associations of firm performance (ROA) with IT investment
(ITINV) and the two corporate governance variables (BIND and FOR). Following the literature,
we expect the coefficients of IT investment (β1), board independence (β2), and foreign ownership
(β4) to be insignificant (e.g., Kohli and Devaraj 2003), insignificant (e.g., Dalton et al. 1998;
Bhagat and Black 2002), and significant and negative (e.g., Short 1994), respectively. To
evaluate the moderation of corporate governance, the model incorporates interaction terms
between IT investment and corporate governance (ITINV×BIND and ITINV×FOR). The model
also controls for lagged performance (ROA-1) and other factors (GROWTH, LEV, SHARE, and
RD) as discussed above.
According to this model, the relationship between firm performance and IT investment is
a function of board independence (BIND) and foreign ownership (FOR), expressed by the
following first-order derivative:
0
10
1ββ += ∑
=kkk xROA
FORBINDITINVROA
531 βββ ++=∂∂
19
where β1 represents the direct IT-performance relationship, and β3 and β5 represent moderating
effects of board independence and foreign ownership, respectively. According to the hypotheses
proposed earlier, β3 is expected to be significantly positive (H1) and β5 is to be significantly
positive (H3).
In H2 and H4, we are concerned with differential moderating effects of corporate
governance across industry sectors and size classes. Chow test (Chow 1960) can be used to test
for whether the relationships in a linear model remain stable across sub-samples. We employ the
dummy variable regression approach to Chow test (Gujarati 1988). Specifically, we create a
dummy C: C=1 for firms with CR4 above the sample median; C=0 for firms with CR4 below the
sample median. Then, we estimate the following regression model:
where )10,...2,1( =kxk denotes the explanatory variable as before (defined in Model (1)), and Ckβ
and Ck−1β are regression coefficients for less competitive industries (C=1) and more competitive
industries (C=0), respectively. The dummy variable regression also specifies the differential
intercept (by CC 0α ), which indicates how much the intercept term in the less competitive
industries differs from the base intercept ( 0α ). To see the implications of Model (2), we obtain:
where C3β and C−1
3β gauge the moderating effect of board independence in less competitive
industries and more competitive industries, respectively. As stated in H2, we expect C−13β to be
more positive than C3β .
)2())(1()( 00
10
1
110
1 CxCxCROA C
kk
Ck
kk
Ck ααββ ++−+= ∑∑
=
−
=
FORBINDITINVROA Custries titive indMore compe
FORBINDITINVROA Custries titive indLess compe
CCC
CCC
−−− ++=∂∂=
++=∂∂=
15
13
11
531
:)0(
:)1(
βββ
βββ
20
We create another dummy A: A=1 for firms with ASSETS above the sample median; A=0
for firms with ASSETS below the sample median. Then, we estimate the following model:
where )10,...2,1( =kxk is the explanatory variable as before, and Akβ and A
k−1β are regression
coefficients for large firms and small firms, respectively. AA 0γ is the differential intercept and
0γ is the base intercept. We obtain:
where A5β and A−1
5β represent the moderating effect of foreign ownership for large firms and
small firms, respectively. H4 proposes that A−15β is more positive than A
5β .
4. Results
4.1. Summary Statistics
Table 2 summarizes the descriptive statistics of the key variables on both the full sample
and sub-samples. Table 3 shows Pearson correlations. The ANOVA results in Table 2 show that
small firms in our sample have greater ROA, consistent with the nature of the electronics industry
(Kraemer and Dedrick 1998). The production of commodity electronic products (e.g., modems,
motherboards, flat-panel displays) is characterized by asset-intensive, high-volume, but low-
return manufacturing. In contrast, applications providers and firms focused on design and
services enjoy high profitability, while their operations require relatively fewer assets.
The ANOVA also shows that foreign investors target firms in less competitive industries,
possibly because investments in these firms are less risky. The mean board independence is also
higher in less competitive industries. It might be because firms in these industries are more
FORBINDITINVROA A firms Small
FORBINDITINVROA A firms Large
AAA
AAA
−−− ++=∂∂=
++=∂∂=
15
13
11
531
:)0(
:)1(
βββ
βββ
)3())(1()( 00
10
1
110
1 AxAxAROA A
kk
Ak
kk
Ak γγββ ++−+= ∑∑
=
−
=
21
inclined to rely on board independence to immunize themselves against managerial incentives
for self-interest (Nalebuff and Stiglitz 1983). Or, it may simply be that BIND is positively
correlated with FOR (see Table 3) and the mean FOR is higher in less competitive industries.
Because our focus is on how BIND moderates IT value and the corporate governance variables
are correlated with several control variables (Table 3), it is difficult to draw conclusions based
solely on the univariate analysis and we need to proceed to a multivariate regression.
Table 2 also shows that small firms have higher BIND, consistent with prior research
showing a negative relationship between firm size and audit committee independence (Klein
2002). Klein discusses that “[l]arger firms have stronger internal controls systems than smaller
firms…If the firms’ internal controls act as in-house monitoring mechanisms, then larger firms
require less alternative monitoring” (Klein 2002, p.440). A similar rationale can be used to
explain the observation in our sample. Also, our results show that firms operating in less
competitive industries or having a large size have higher GROWTH than those that are in more
competitive industries or are smaller in size.
In Table 3 we note two high correlations. First, as expected, our dependent variable
(ROA) is highly correlated (r=0.835, p<0.01) with a control variable ROA-1. Second, a control
variable SHARE is highly correlated (r=0.724, p<0.01) with CR4, which is used to divide the
sample into less and more competitive industries. The high correlation is intuitive; on average,
firms in more concentrated industries have higher market shares.
Finally, it is worth noting that the pair-wise correlations between the explanatory
variables in our regression model are not high, with the largest absolute value equaling 0.204
(between BIND and RD). Hence, multicollinearity is not likely to be a problem. We also
22
calculate variance inflation factors, which in our study are all below 1.5, suggesting no harmful
mutilcollinearity (Greene 2000).
__________________________________
Insert Tables 2 and 3 about here __________________________________
4.2. Full-Sample Results
Table 4 presents estimates for Model (1), showing the relationships of ROA with the ten
explanatory variables on the full sample. The overall model is significant (F= 221.07, p<0.01).
The coefficient of ITINV is insignificant (t=0.27), supporting that the direct relationship between
IT investment per se and firm performance is not clear (Kohli and Devaraj 2003).
The coefficient of BIND is insignificant (t=-0.62), consistent with the non-correlation
between board independence and firm performance reported by previous studies (e.g., Dalton et
al. 1998; Bhagat and Black 2002). Nonetheless, we find a positive and significant interaction
between ITINV and BIND (β3=0.084, t=3.84), indicating that the IT investment-ROA relation
would be more positive for firms with higher board independence. This finding supports H1.
The coefficient of ITINV×FOR is insignificant (t=0.78), thus showing no support for H3.
The coefficient of FOR is also insignificant (t=-0.89), which is different from our expectation. A
plausible explanation is that large firms, with sound internal controls and comprehensive
auditing systems in place, do not suffer from agency costs due to foreign investment. This
dilutes the influence of FOR on the full sample. To further check this explanation, we need to
look into the sample split by firm size, which is discussed in Section 4.4.
_________________________
Insert Table 4 about here _________________________
23
Among the controls, the coefficients of ROA-1 and GROWTH are significant and positive,
consistent with our expectations. Debt financing has both benefits and costs, and here we find a
significant and negative coefficient of LEV (-0.013, p<0.01). This is consistent with results in
previous studies using different samples (e.g., Agrawal and Knoeber 1996; Hitt and Brynjolfsson
1996). SHARE is insignificant, possibly because GROWTH taps increased negotiation power.
This interpretation is consistent with evidence in prior research (e.g., Hitt and Brynjolfsson 1996).
Another explanation is that firms in our sample may lack buying power (Szymanski et al. 1993).
Finally, the coefficient of RD is insignificant. This may be because ROA does not effectively
capture impacts of the intangible assets proxied by R&D expenditures.
4.3. More Competitive versus Less Competitive Industries
As shown in Table 5, estimates of Model (2) include two sets of regression coefficients,
i.e., for less competitive industries (C=1) and for more competitive industries (C=0). The
significant Chow test (11.21, p<0.01) rejects equality between the two sets of coefficients. A
notable difference is that the significant coefficient of ITINV×BIND in the full-sample (Table 4)
becomes insignificant in less competitive industries ( C3β =0.0093, t=0.27), while remains
significant and of higher magnitude ( C−13β =0.17, t=4.60) in more competitive industries. This
difference ( C−13β vs. C
3β ) provides support for H2.
In less competitive industries, there is no significant direct or indirect IT-performance
relationship through interacting with corporate governance. These results suggest that in less
competitive industries, firm profitability may result from sources other than new technologies,
such as monopoly rents. By contrast, firms operating in more competitive industries need to rely
more heavily on new technologies to improve performance. The significant interaction term
24
between ITINV and BIND (0.17, p<0.01) shows that, in conjunction with board independence, IT
investment creates significant value in intensely competitive industries.
Together with the ANOVA in Table 2, these regression results suggest an interesting
paradox: The mean board independence is lower in more competitive industries (Table 2), which
seems to be consistent with the rationale that competition, in general, could serve as a bonding
mechanism to monitor managerial incentives (e.g., Nalebuff and Stiglitz 1983). Nonetheless, as
shown in Table 5, the bonding mechanism of competition does not seem to be effective in the
context of IT investment. In the more competitive industries, the coefficient of ITINV is positive,
but numerically small and insignificant. IT investment relates to higher ROA only through
interacting with board independence. A plausible explanation, as stated in hypothesis
development, is that shareholders lack information about IT investment, which limits their ability
to assess managers’ investment decisions. Alternatively, shareholders need to rely on corporate
boards to monitor IT investment.
As to the controls, we observe qualitatively similar results except for the insignificant
LEV in less competitive industries. Since firms in less competitive industries generally possess
greater free cash flow, they benefit more from the control function of debt financing (Jensen
1986). The increased control function thus weakens the negative influence of LEV.
_________________________
Insert Table 5 about here _________________________
4.4. Large Firms versus Small Firms
Table 6 shows estimates for Model (3), including regression coefficients for large and
small firms. The significant Chow test (2.27, p<0.01) rejects equality between the two sets of
coefficients. Again, we find insignificant coefficients of ITINV and BIND. Consistent with the
25
full sample, the coefficient of ITINV×BIND is significant for both large and small firms. We
conclude that the moderating effect of board independence is robust across size classes.
The coefficient of FOR is negative, significant for small firms (t=-2.13) and insignificant
for large firms (t=-0.37). This suggests that compared with large firms, small firms suffer from
the agency problem attributed to dispersed ownership as suggested by Short (1994). The
interaction ITINV×FOR is significant and positive for small firms ( ,26.015 =−Aβ t=1.84) and
insignificant for large firms ( ,024.05 =Aβ t=0.45). This difference ( A−15β vs. A
5β ) supports H4
that small (instead of large) Taiwanese firms benefit from the IT expertise “spilled over” from
foreign investors.
_________________________
Insert Table 6 about here _________________________
4.5. Sensitivity Analysis
We conduct the following tests to examine the robustness of the results to measures and
model specifications. These additional tests demonstrate the robustness of our results.
Alternative measures for board independence. Earlier, we discussed alternative
measures for board independence (OUT and NONAFF). Panel A of Table 7 presents the results
of regressions using NONAFF as a proxy for board independence. As seen in Panel A, the
coefficients of all independent variables are virtually the same. The independent variable that
appears to be most affected by using the alternative proxy is ITINV×BIND, which turns out to be
significant at the 5% level for small firms (compared with 10% in Table 6). Panel B of Table 7
presents the results of regressions using OUT as a proxy for board independence. The results are
highly consistent with the results in Tables 4-6.
26
Alternative measure for industry competitiveness. We use an alternative measure for
industry competitiveness, entry costs (ENTCOST), which refer to sunk costs that must be
incurred for operating in one industry (Porter 1985). The higher the entry costs, the higher the
industry concentration and the lower the industry competitiveness. Following the literature (e.g.,
Karuna 2006), ENTCOST is defined as the weighted average gross value of property, plant, and
equipment for firms in one industry, weighted by each firm’s market share in the industry. We
rank all firms in our sample by ENTCOST and divide them at the median. Firms with ENTCOST
above median form the less competitive industries, and firms with ENTCOST below median
form the more competitive industries. As seen in Panel C of Table 7, we observe qualitatively
similar results using this alternative measure for industry competitiveness (Table 5).
Industry effects. Next, we control for industry effects. We re-estimate regression (1) by
including industry dummies as additional controls. Only one industry dummy (the systematic
product sector) has a significant coefficient (at 10% significance level), and jointly the industry
dummies are insignificant. We then add the industry dummies into regressions (2) and (3) and
find no significant industry dummies. As shown in Panel D of Table 7, including industry
dummies produces highly consistent results for all regressions.
Two-year lagged performance. We examine the sensitivity of our regressions (1)-(3) by
adding an additional control variable, the two-year lagged performance (ROA-2). As shown in
Panel E of Table 7, the coefficient of ROA-2 is significant only for large firms (in the “A=1”
column). The coefficients of all the other independent variables are qualitatively similar to those
reported in Tables 4-6. Also, for small firms (in the “A=0” column), the significance level of the
coefficient for ITINV×BIND increases (i.e., from 10% in Table 6 to 5%).
27
Partial models. We also estimate partial models. We check for the robustness of
regressions (1)-(3) by dropping the moderation effects (BIND, ITINV×BIND, FOR,
ITINV×FOR). The coefficient of ITINV remains insignificant in each regression (not reported),
consistent with the finding that the direct IT -performance relationship is not clear.
We then re-estimate regressions (1)-(3) with all control variables dropped. As reported in
Panel F of Table 7, the significance levels of two coefficients are affected by the removal of
controls, while there were no significant coefficients with a change in sign. The coefficient of
FOR becomes significant at the 10% level for small firms (compared with 5% in Table 6), and
the coefficient of ITINV×BIND becomes significant at the 1% level for large firms (compared
with 5% in Table 6). Taken together, we observe qualitatively similar results.
Dummy variable regression. Finally, we consider the robustness of the dummy variable
regression. We use the sample mean of CR4 as the cutoff to classify more and less competitive
industries. We obtain estimates qualitatively similar to those reported in Table 5 (where the
sample median of CR4 is used as the cutoff). We do not use the sample mean of ASSETS to
classify large and small firms because this variable is skewed.
________________________
Insert Table 7 about here ________________________
5. Concluding Remarks
Information technology has the potential to improve firm performance. Yet, researchers
are puzzled by the mixed empirical findings concerning returns on IT investment. Using a
sample of manufacturers in high-tech industries in Taiwan, we investigate how corporate
governance shapes the IT investment-firm performance relationship. We find moderating effects
28
of board independence and foreign ownership on the IT value on firm performance. The
moderation effects differ in significance across size classes and industries.
Specifically, we find no moderating effect of corporate governance on the IT investment-
firm performance relationship in less competitive industries. In contrast, in more competitive
industries, there is a significant interaction effect between IT investment and board independence,
which is positively related to firm performance. Information technologies help firms improve
performance under competition, while the economic returns on IT investment are positively
related to board independence. This supports the conjecture that the corporate board plays a
monitoring role in new technology investment. Our findings show that board monitoring
benefits intensely competitive industries more. Clearly, the relationships among IT returns,
board independence, and industry competitiveness are rather complex. While we have not
completely disentangled the effects, our results offer insights into the role of corporate
governance in achieving returns on IT investment.
Moreover, we find a negative direct association between foreign ownership and firm
performance for small firms, possibly due to additional agency costs incurred and foreign
investors’ limited local knowledge in offering advice for general management. However, we
observe a significant and positive interaction between IT investment and foreign ownership in
the regression for small firms. Foreign investors may inject IT expertise that is likely to be
applicable to both developed and emerging markets. Hence, foreign ownership helps small firms
more effectively deploy IT. As an implication for research on corporate governance, our results
suggest that an analysis of governance structures needs to be situated within specific contexts
such as firm size and industry competitiveness. Before discussing the managerial implications of
our results, we briefly point out some limitations of our study.
29
First, this study is based on a sample from an emerging market, and thus questions must
be raised about the generalizability of the findings to developed markets, e.g., the U.S. Second,
this study is focused on two specific corporate governance factors, and we view our work as an
initial attempt to analyze the moderation of governance structures on IT value. One line of
future inquiry is to investigate other dimensions of corporate governance (e.g., board size, major
stockholders’ ownership, deviation between major shareholders’ control rights and cash claim
rights). Third, although our analysis of spillover follows a common method in economics (e.g.,
Blomström 1986; Braga and Willmore 1991; Aitken and Harrison 1999), future studies can
cross-validate our results and delve deeper into these issues by using other methodologies. For
example, field research may detail how foreign investors are involved in making IT investment
decisions. Also, more studies are needed to better understand what type of knowledge and
experience needs to be adapted to the local environment and culture in emerging markets and
what type can be more directly transferred from foreign ownership. Notwithstanding these
limitations, our study makes significant contributions to both research and practices.
As an important implication for firms considering IT investment, our results suggest that
competitive firms should be more proactive in using IT to increase their business value. Yet IT
alone does not hold the answer to performance improvement and must be combined with better
and more objective monitoring to reap the benefits. Our results also imply that foreign
ownership can help smaller firms transcend barriers to IT return. Managers need to realize that,
with foreign investors involved, their firms are more likely to attain performance improvement
through IT. However, the evidence should not be interpreted as recommending no IT investment
at firms with little or no foreign ownership, because there are approaches other than spillover that
are considered in acquiring IT-related knowledge. Overall, an important message to managers is
30
that management control systems play a crucial role in IT impacts across different structures of
markets and firms.
In conclusion, the contribution of this study lies in its evaluation of the moderation
effects of corporate governance in IT value creation. Our study provides one theoretical lens
through which different outcomes of IT are analyzed. The evidence also shows the usefulness of
studying the role of corporate governance in a specific investment activity, rather than linking it
directly to overall firm performance. These directions may lead to fruitful areas of future inquiry.
31
Table 1. Industry Sectors
Industry sector Frequency % of sample Electronic components 135 18.8 Photoelectric products 131 18.2 Motherboard scheme 110 15.3 Integrated circuits 95 13.2 Electronic channel 81 11.3 Software applications 37 5.1 Network modem 34 4.7 General electronics 30 4.2 Consumer electronics 26 3.6 Communication technologies 21 2.9 Systematic product and others 19 2.7 Total 719 100.0
32
Table 2. Descriptive Statistics Less competitive vs. more competitive industries Large firms vs. small firms
Full sample CR4 above
median CR4 below
median ANOVA ASSETS above
median ASSETS below
median ANOVA Mean S.D. Mean S.D. Mean S.D. t-statistic Mean S.D. Mean S.D. t-statistic
ROA 0.151 0.114 0.158 0.118 0.152 0.120 0.66 0.132 0.104 0.170 0.121 -4.61 ***ITINV 0.037 0.617 0.004 0.005 0.094 1.033 -1.39 0.003 0.05 0.070 0.871 -1.48 BIND 16.348 18.749 28.117 17.453 1.511 4.440 27.48 *** 14.474 17.715 18.217 19.572 -2.69 ***FOR 5.490 9.726 7.010 10.744 4.175 8.857 3.56 *** 7.920 10.922 3.066 7.645 6.90 ***ROA-1 0.169 0.122 0.166 0.124 0.184 0.128 -1.74 * 0.145 0.108 0.193 0.131 -5.34 ***GROWTH 0.111 0.363 0.148 0.299 0.063 0.408 2.81 ** 0.150 0.380 0.073 0.342 2.84 ***LEV 0.673 0.618 0.607 0.617 0.737 0.631 -2.53 ** 0.681 0.561 0.665 0.670 0.33 SHARE 0.322 1.729 0.382 1.888 0.201 0.427 1.75 * 0.346 1.607 0.299 1.844 0.36 RD 0.028 0.041 0.024 0.042 0.040 0.043 -4.44 *** 0.025 0.033 0.032 0.047 -2.09 **ASSETS 9.583 23.711 9.493 18.842 10.773 32.434 -0.61 17.380 31.705 1.807 0.812 9.30 ***CR4 0.208 0.069 0.223 0.087 0.188 0.001 7.64 *** 0.202 0.406 0.214 0.088 -2.47 **
***p<0.01; **p<0.05; *p<0.10.
Data from the Taiwanese Institute for Information Industry: ITINV = annual IT spending (hardware, software and costs concerning maintenance, personnel and training) divided by total assets
Data from TEJ’s Corporate Database: BIND = 1.14×NONAFF+0.65×AFF
NONAFF = proportion of non-affiliated directors (%) AFF = proportion of affiliated directors (%) FOR = proportion of common stock owned by foreign investors (%)
Data from TEJ’s Financial Report Database: ROA = net income ÷ yearend book value of total assets
ROA-1 = ROA in the previous year GROWTH = one-year growth rate of sales
LEV = book value of total debt ÷ book value of total assets SHARE = sales ÷ total industry sales (industry sectors shown in Table 1)
RD = research and development expenses ÷ sales ASSETS = total assets (US$ billion, 2001 constant dollar)
CR4 = sales of the four largest firms in the industry ÷ total industry sales
33
Table 3. Pearson Correlations
ROA ITINV BIND FOR ROA-1 GROWTH LEV SHARE RD ASSETS CR4ROA 1 ITINV -0.002 1 BIND 0.114*** -0.047 1 FOR -0.010 -0.029 0.084** 1 ROA-1 0.835*** -0.009 0.077** -0.008 1 GROWTH 0.185*** -0.037 0.074** -0.007 -0.043 1 LEV -0.168*** 0.003 -0.112*** -0.120*** -0.159*** 0.147*** 1 SHARE 0.017 -0.008 -0.059 0.011 0.028 -0.002 0.013 1 RD 0.178*** -0.026 -0.204*** 0.043 0.199*** 0.017 -0.082** 0.087** 1 ASSETS -0.149*** -0.018 -0.109 0.308*** -0.119*** 0.014 -0.016 0.138*** 0.093** 1 CR4 0.008 -0.016 0.110*** -0.012 -0.006 -0.034 -0.109*** 0.724*** 0.046 -0.043 1 ***p<0.01; **p<0.05; *p<0.10. See Table 2 for variable definitions.
34
Table 4. Regression Results: Full Sample
= β1 ITINV + β2 BIND + β3 (ITINV×BIND) + β4 FOR + β5 (ITINV×FOR)
+ β6 ROA-1 + β7 GROWTH + β8 LEV + β9 SHARE + β10 RD + β0
Coefficient βk Predicted Sign (t-statistic)
ITINV ? 0.0011 (0.27)
BIND ? -0.000085 (-0.62) ITINV × BIND + 0.084*** (3.84) FOR – -0.00020 (-0.89) ITINV × FOR + 0.010 (0.78) Controls ROA-1 + 0.77*** (42.49) GROWTH + 0.073*** (12.23) LEV – -0.013*** (-3.64) SHARE ? -0.00066 (-0.53) RD ? 0.018 (0.34) Intercept 0.019*** (3.43) F-statistic 221.07*** Adjusted R2 0.754
***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. See Table 2 for variable definitions.
0
10
1ββ += ∑
=kkk xROA
35
Table 5. Regression Results:
More Competitive Industries vs. Less Competitive Industries
Less competitive industries (C=1)
More competitive industries (C=0)
Coefficient Ckβ Coefficient C
k−1β
(t-statistic) (t-statistic) ITINV -0.066 0.0032 (-0.67) (0.72) BIND 0.00014 -0.00021 (0.58) (-1.07) ITINV × BIND 0.0093 0.17*** (0.27) (4.60) FOR -0.00079 -0.000067 (-1.50) (-0.17) ITINV × FOR 0.107 -0.027 (0.74) (-0.52) Controls ROA-1 0.72*** 0.79*** (18.06) (36.01) GROWTH 0.069*** 0.095*** (6.67) (10.53) LEV -0.012 -0.014*** (-1.47) (-3.10) SHARE 0.00080 -0.00049 (0.42) (-1.48) RD -0.10 0.013 (-0.64) (0.20) Base intercept 0.0084 0.0084 (1.10) (1.10) Differential intercept 0.026** -- (2.03) Chow test 11.21*** F-statistic 94.88*** Adjusted R2 0.764
***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. C=1 for firms with above-median CR4; C=0 for firms with below-median CR4. See Table 2 for variable definitions.
C
kk
Ck
kk
Ck CxCxCROA 00
10
1
110
1))(1()( ααββ ++−+= ∑∑
=
−
=
36
Table 6. Regression Results:
Large Firms vs. Small Firms
Large Firms
(A=1) Small Firms
(A=0) Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) ITINV -0.0097 -0.0091 (-0.27) (-1.34) BIND -0.000013 -0.00016 (-0.08) (-0.65) ITINV × BIND 0.082** 0.059* (2.33) (1.84) FOR -0.00012 -0.0023** (-0.37) (-2.13) ITINV × FOR 0.024 0.26* (0.45) (1.84) Controls ROA-1 0.81*** 0.70*** (34.57) (24.36) GROWTH 0.061*** 0.094*** (8.42) (9.12) LEV -0.014*** -0.011** (-2.96) (-2.03) SHARE -0.0015 0.00058 (-0.80) (0.35) RD 0.011 -0.0065 (0.15) (-0.08) Base intercept 0.037*** 0.037*** (3.82) (3.82) Differential intercept -0.025** -- (-2.13) Chow test 2.27*** F-statistic 108.56*** Adjusted R2 0.759
***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. A=1 for firms with above-median ASSETS; A=0 for firms with below-median ASSETS.See Table 2 for variable definitions.
A
kk
Ak
kk
Ak AxAxAROA 00
10
1
110
1))(1()( γγββ ++−+= ∑∑
=
−
=
37
Table 7. Sensitivity Analysis
Panel A: Use NONAFF to measure board independence Full Sample C=1† C=0† A=1‡ A=0‡
Coefficient kβ Coefficient Ckβ Coefficient C
k−1β Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) (t-statistic) (t-statistic) (t-statistic) ITINV 0.00079 -0.060 0.0012 -0.027 -0.0088
(0.19) (-0.61) (0.27) (-0.77) (-1.30) NONAFF -0.00019 0.00012 -0.00037 0.0000094 -0.00040
(-0.72) (0.26) (-0.99) (0.03) (-0.92) ITINV × NONAFF 0.16*** 0.063 0.24*** 0.13** 0.15**
(3.47) (0.91) (3.51) (1.94) (2.20) FOR -0.00020 -0.00072 -0.00018 -0.00019 -0.0023**
(-0.87) (-1.37) (-0.45) (-0.63) (-2.13) ITINV × FOR 0.012 0.098 0.019 0.049 0.25*
(0.90) (0.68) (0.38) (0.98) (1.81) Controls in Tables 4-6 Included Included Included Included Included
Panel B: Use OUT to measure board independence Full Sample C=1† C=0† A=1‡ A=0‡
Coefficient kβ Coefficient Ckβ Coefficient C
k−1β Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) (t-statistic) (t-statistic) (t-statistic) ITINV 0.0010 -0.070 0.0031 -0.010 -0.0097
(0.25) (-0.71) (0.69) (-0.28) (-1.44) OUT -0.00016 0.000091 -0.00022 -0.000037 -0.00027
(-0.86) (0.29) (-0.79) (-0.16) (-0.84) ITINV × OUT 0.10*** 0.0041 0.19*** 0.093** 0.070*
(3.65) (0.08) (4.29) (2.12) (1.66) FOR -0.00019 -0.00079 0.000093 -0.00011 -0.0024**
(-0.85) (-1.15) (0.23) (-0.35) (-2.18) ITINV × FOR 0.010 0.11 -0.023 0.024 0.27*
(0.78) (0.78) (-0.44) (0.46) (1.95) Controls in Tables 4-6 Included Included Included Included Included † “C=1” and “C=0” denote the two subsamples as defined in Model (2), i.e., less competitive and more competitive industries, respectively. ‡ “A=1” and “A=0” denote the two subsamples as defined in Model (3), i.e., large firms and small firms, respectively. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. NONAFF and OUT are the shares of non-affiliated and outside directors in the board, respectively; see Table 2 for variable definitions.
38
Table 7. Sensitivity Analysis (continued)
Panel C: Use ENTCOST to measure industry competitiveness Less Competitive Industries
(ENTCOST above median) More Competitive Industries
(ENTCOST below median)
Coefficient Coefficient (t-statistic) (t-statistic) ITINV -0.040 0.0018
(-0.08) (0.44) BIND -0.00013 -0.0000090
(-0.55) (-0.05) ITINV × BIND 0.031 0.13 ***
(0.81) (4.27) FOR -0.00038 -0.00030
(-1.10) (-0.76) ITINV × FOR 0.041 0.0081
(0.81) (0.62) Controls in Tables 4-6 Included Included
Panel D: Control for industry effects Full Sample C=1† C=0† A=1‡ A=0‡
Coefficient kβ Coefficient Ckβ Coefficient C
k−1β Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) (t-statistic) (t-statistic) (t-statistic) ITINV 0.0012 -0.072 0.0035 -0.013 -0.0090
(0.28) (-0.71) (0.78) (-0.37) (-1.29) BIND -0.00011 0.00016 -0.00024 -0.000015 -0.00019
(-0.82) (0.66) (-1.21) (-0.09) (-0.80) ITINV × BIND 0.091*** 0.0092 0.17*** 0.078** 0.064*
(3.91) (0.24) (4.45) (2.20) (1.80) FOR -0.00017 -0.00079 -0.00016 -0.000066 -0.0024**
(-0.72) (-1.47) (-0.39) (-0.21) (-2.15) ITINV × FOR 0.0092 0.115 -0.032 0.029 0.25*
(0.72) (0.77) (-0.62) (0.56) (1.76) Controls in Tables 4-6 Included Included Included Included Included Industry dummies Included Included Included Included Included † “C=1” and “C=0” denote the two subsamples as defined in Model (2), i.e., less competitive and more competitive industries, respectively. ‡ “A=1” and “A=0” denote the two subsamples as defined in Model (3), i.e., large firms and small firms, respectively. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. ENTCOST = weighed average gross value of property, plant and equipment for firms in one industry; see Table 2 for variable definitions.
39
Table 7. Sensitivity Analysis (continued)
Panel E: Include lagged 2-year performance measure Full Sample C=1† C=0† A=1‡ A=0‡
Coefficient kβ Coefficient Ckβ Coefficient C
k−1β Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) (t-statistic) (t-statistic) (t-statistic) ITINV 0.00095 -0.066 0.0033 -0.012 -0.0086
(0.23) (-0.67) (0.73) (-0.32) (-1.27) BIND -0.00011 0.00014 -0.00021 -0.000083 -0.000059
(-0.75) (0.56) (-0.99) (-0.46) (-0.23) ITINV × BIND 0.083** 0.0094 0.17*** 0.073** 0.063**
(3.72) (0.27) (4.56) (2.06) (1.96) FOR -0.00021 -0.00079 -0.000071 -0.00013 -0.0023**
(-0.91) (-1.49) (-0.17) (-0.42) (-2.10) ITINV × FOR 0.010 0.107 -0.027 0.026 0.26*
(0.78) (0.74) (-0.52) (0.49) (1.84) Controls in Tables 4-6 Included Included Included Included Included ROA-2 0.014 -0.0013 -0.0047 0.066* -0.0039 (0.59) (-0.03) (-0.15) (1.93) (-1.09)
Panel F: Estimate partial models Full Sample C=1† C=0† A=1‡ A=0‡
Coefficient kβ Coefficient Ckβ Coefficient C
k−1β Coefficient A
kβ Coefficient Ak−1β
(t-statistic) (t-statistic) (t-statistic) (t-statistic) (t-statistic) ITINV 0.0011 -0.124 0.0039 0.048 -0.020
(0.13) (-0.63) (0.43) (0.67) (-1.54) BIND 0.00011 0.00058 0.00010 0.00022 -0.00036
(0.42) (1.23) (0.27) (0.67) (-0.79) ITINV × BIND 0.19*** 0.11 0.22*** 0.26*** 0.10*
(4.51) (1.58) (3.09) (3.73) (1.66) FOR -0.00025 -0.0016 -0.00029 0.00038 -0.0039*
(-0.56) (-1.52) (0.37) (0.63) (-1.82) ITINV × FOR -0.0038 0.18 -0.083 -0.070 0.47*
(-0.15) (0.53) (-0.79) (-0.69) (1.74) Controls in Tables 4-6 Dropped Dropped Dropped Dropped Dropped † “C=1” and “C=0” denote the two subsamples as defined in Model (2), i.e., less competitive and more competitive industries, respectively. ‡ “A=1” and “A=0” denote the two subsamples as defined in Model (3), i.e., large firms and small firms, respectively. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.
40
References:
Abrahamson, E., Rosenkopf, L., 1997. Social network effects on the extent of innovation diffusion: A computer simulation. Organization Science 8 (3), 289-309.
Agrawal, A., Knoeber, C.R., 1996. Firm performance and mechanisms to control agency problems between managers and shareholders. Journal of Financial & Quantitative Analysis 31(3), 377-390.
Aitken, B.J., Harrison, G., 1999. Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review 89 (3), 605-618.
Banker, R.D., Bardhan, I.R., Chang, H., Liu, S., 2006. Plant information systems, manufacturing capabilities, and plant performance. MIS Quarterly 30 (2), 315-337.
Baysinger, B.D., Kosnik, R.E. Turk., T.A, 1991. Effects of board and ownership structure on corporate R&D strategy. Academy of Management Journal 34(1), 205-214.
Beasley, M.S., Carcello, J.V., Hermanson, D.R., Lapides, P.D., 2000.Fraudulent financial reporting: Consideration of industry traits and corporate governance mechanisms. Accounting Horizons 14(4), 441-454.
Bens, D.A., 2002. The determinants of the amount of information disclosed about corporate restructurings. Journal of Accounting Research 40 (1), 1-20.
Bhagat, S., Black, B., 2002. The non-correlation between board independence and long-term firm performance. Journal of Corporation Law 27 (2), 231-273.
Bharadwaj, A.S., Bharadwaj, S.G., Konsynski, B.R., 1999. Information technology effects on firm performance as measured by Tobin's q. Management Science 45 (7), 1008-1024.
Blomström, M., 1986. Foreign investment and productive efficiencies: The case of Mexico. Journal of Industrial Economics 35 (1), 97-110.
Boardman, A.E., Shapiro, D.M., Vining, A.R., 1997. The role of agency costs in explaining the superior performance of foreign MNE subsidiaries. International Business Review 6 (3), 295-317.
Braga, H., Willmore, L., 1991. Technological imports and technological effort: An analysis of their determinants in Brazilian firms. Journal of Industrial Economics 39 (4), 421-432.
Bresnahan, T.F., Brynjolfsson, E., Hitt, L.M., 2002. Information technology, workplace organization, and the demand for skilled labor: Firm-level evidence. Quarterly Journal of Economics 117 (1), 339-376.
Brynjolfsson, E., Hitt, L.M., Yang, S., 2002. Intangible assets: Computers and organizational capital. Brookings Papers on Economic Activity: Macroeconomics (1), 137-199.
Bushman, R.J., Smith, A.J., 2001. Financial accounting information and corporate governance. Journal of Accounting and Economics 32 (1-3), 237-333.
Callen, J.L., Hope, O.-K., Segal, D., 2005. Domestic and foreign earnings, stock return variability, and the impact of investor sophistication. Journal of Accounting Research 43 (3), 377-412.
Carr, N., 2004. Does IT Matter: Information Technology and the Corrosion of Competitive Advantage. Cambridge: Harvard Business School.
Chaney, P.K., and Philipich, K.L., 2002. Shredded reputation: The cost of audit failure. Journal of Accounting Research 40 (4), 1221-1245.
Chiang, Y., Kuo, C., 2006. Characteristics of Firms Owned by Foreigners on the Taiwanese Stock Market: An Analysis. International Journal of Management 23 (3), 721-728.
Chow, G.C., 1960. Tests of equality between sets of coefficients in two linear regressions. Econometrica 28 (3), 591-605.
41
Cohen, W.M., Levin, R.C., 1989. Empirical studies of innovation and market structure. In: Schmalensee, R., Willig, R. (Eds.). Handbook of Industrial Organization. Amsterdam: The Netherlands, 1059-1107.
Committee of Sponsoring Organizations of the Treadway Commission 1992. Daily, C. M., Dalton, D.R., 1994. Bankruptcy and corporate governance: The impact of board
composition and structure. Academy of Management Journal 37 (6), 1603-1617. Dalton, D.R., Daily, C.M., Ellstrand, A.E., Johnson, J.L., 1998. Meta-analytic reviews of board
composition, leadership structure, and financial performance. Strategic Management Journal 19 (3), 269-290.
Damanpour, F., 1996. Organizational complexity and innovation: Developing and testing multiple contingency models. Management Science 42(5), 693-716.
Darrough, M., Rangan, S., 2005. Do insiders manipulate earnings when they sell their shares in an initial public offering? Journal of Accounting Research 43 (1), 1-33.
Dechow, P.M., Sloan, R.G., Sweeney, A.P., 1996. Causes and consequences of earnings manipulation: An analysis of firms subject to enforcement actions by the SEC. Contemporary Accounting Research 13 (1), 1-36.
Dedrick, J., Kraemer, K.L., 2005. The impacts of IT on firm and industry structure: The personal computer industry. California Management Review 47 (3), 122-142.
Devaraj, S., Kohli, R., 2003. Performance impacts of information technology: Is actual usage the missing link? Management Science 49 (3), 273-289.
Dewan, S., Kraemer, K.L., 2000. Information technology and productivity: Evidence from country-level data. Management Science 46(4), 548-562.
Eisenhardt, K.M., Schoonhoven, C.B., 1990. Organizational growth: Linking founding team, strategy, environment, and growth among U.S. semiconductor ventures, 1978-1988. Administrative Science Quarterly 35 (3), 504-529.
Engel, E., Hayes, R.M., Wang, X., 2006. The Sarbanes-Oxley Act and firms' going-private decisions. Journal of Accounting and Economics In Press, Corrected Proof.
Fama, E.F., Jensen, M.C., 1983. Agency problems and residual claims. Journal of Labor Economics 26 (2), 327-349.
Fan, J.P. H., Wong, T.J., 2005. Do external auditors perform a corporate governance role in emerging markets? Evidence from East Asia. Journal of Accounting Research 43 (1), 35-72.
Feltham, G., Xie, J., 1994. Performance measure congruity and diversity in multi-task principal/agent relations. The Accounting Review 69 (3), 429-453.
Flannery, M.J., Rangan, K.P., 2006. Partial adjustment toward target capital structures. Journal of Financial Economics 79 (3), 469-506.
Graham, J., 1999. Herding among investment newsletters: Theory and evidence. Journal of Finance 54(1), 237-269.
Greene, W.H., 2000. Econometric Analysis. Upper Saddle River, NJ: Prentice Hall. Gugler, K., Mueller, D.C., Yurtoglu, B.B., 2003. The impact of corporate governance on
investment returns in developed and developing countries. The Economic Journal 113, F511-F539.
Gujarati, D.N., 1988. Basic Econometrics. 2 ed. New York NY: McGraw-Hill. Hall, B.H., 1993. The stock market's valuation of R&D investment during the 1980's. American
Economic Review 83 (2), 259-264.
42
Harrison, D.A., Mykytyn, P.P., Jr., Riemenschneider, C.K., 1997. Executive decisions about adoption of information technology in small businesses: Theory and empirical tests. Information Systems Research 8(2), 171-196.
Hermalin, B., 1992. The effects of competition on executive behavior. Rand Journal of Economics 23(3), 350-365.
Hitt, L.M., Brynjolfsson, E., 1996. Productivity, business profitability, and consumer surplus: Three different measures of information technology value. MIS Quarterly 20 (2), 121-142.
IDC, 2006. Worldwide IT Spending 2006-2010 Forecast by Vertical Market. International Data Corporation, Framingham, MA.
Jensen, M.C., 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76 (2), 323-329.
Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3 (4), 305-360.
Johnstone, K.M., Bedard, J.C., 2004. Audit firm portfolio management decisions. Journal of Accounting Research 42 (4), 659-690.
Kaplan, S.N., Reishus, D., 1990. Outside directorships and corporate performance. Journal of Financial Economics 27 (2), 389-410.
Karuna, C., 2006. Industry product market competition and managerial incentives. Forthcoming, Journal of Accounting and Economics.
Kauffman, R.J., Li, X., 2003. Payoff externalities, informational cascades and managerial incentives: A theoretical framework for IT adoption herding. Proceedings of the 2003 INFORMS Conference on IS and Technology, Atlanta, GA.
Klein, A., 1998. Firm performance and board committee structure. Journal of Law and Economics 41 (1), 275-303.
Klein, A., 2002. Audit committee, board of director characteristics, and earnings management. Journal of Accounting and Economics 33 (3), 375-400.
Kohli, R., Devaraj, S., 2003. Measuring information technology payoff: A meta-analysis of structural variables in firm-level empirical research. Information Systems Research 14 (2), 127-145.
Kraemer, K.L., Dedrick, J., 1998. Globalization and increasing returns: Implications for the U.S. computer industry. Information Systems Research 9 (4), 303-322.
Lambert, R.A., 2001. Contracting theory and accounting. Journal of Accounting and Economics 32 (1-3), 3-87.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R.W., 2000. Investor protection and corporate governance. Journal of Financial Economics 58 (1), 3-27.
Levin, S.G., Levin, S.L., Meisel, J.B., 1987. A dynamic analysis of the adoption of a new technology: The case of optical scanners. Review of Economics and Statistics 69 (1), 12-17.
Li, F., F.M.H. Wong, 2005. Employee stock options, equity valuation, and the valuation of option grants using a warrant-pricing model. Journal of Accounting Research 43 (1), 97-132.
Melville, N., Kraemer, K.L., Gurbaxani, V., 2004. Information technology and organizational performance: An integrative model of it business value. MIS Quarterly 28(2), 283-322.
Morrison, C.J., 1997. Assessing the productivity of information technology equipment in U.S. manufacturing industries. Review of Economics and Statistics 79 (3), 471-481.
43
Nalebuff, B., Stiglitz, J., 1983. Information, competition, and markets. American Economic Review 73(2), 278-283.
Peng, M.W., 2004. Outside directors and firm performance during institutional transitions. Strategic Management Journal 25 (5), 453-471.
Porter, M.E. 1985. Competitive advantage: Creating and sustaining superior performance. New York: Free Press.
Scharfstein, D., 1988. Product market competition and managerial slack. Rand Journal of Economics 19(1), 147-155.
Scharfstein, D., Stein, J., 1990. Herd behavior and investment. American Economic Review 80(3), 465-479.
Schmidt, K., 1997. Managerial incentives and product market competition. Review of Economic Studies 64(2), 191-213.
Schumpeter, J.A., 1970. Capitalism, Socialism, and Democracy, 3rd ed. New York: Columbia University Press.
Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance 52 (2), 737-783.
Short, H., 1994. Ownership, control, financial structure and the performance of firms. Journal of Economic Surveys 8 (3), 203-249.
Simunic, D.A., 1980. The pricing of audit services: Theory and evidence. Journal of Accounting Research 18 (1), 161-190.
Sutton, R.I., Callahan, A.L., 1987. The stigma of bankruptcy: Spoiled organizational image and its management. Academy of Management Journal 30 (3), 405-436.
Szymanski, D.M., Bharadwaj, S.G., Varadarajan, P.R., 1993. An analysis of the market share-profitability relationship. Journal of Marketing 57 (3), 1-18.
Thong, J.Y.L., 1999. An integrated model of information systems adoption in small businesses. Journal of Management Information Systems 15 (4), 187-214.
Verrecchia, R.E., Weber, J., 2006. Redacted disclosure. Journal of Accounting Research 44(4), 791-814.
Wade, M., Hulland, J., 2004. Review: The resource-based view and information systems research: Review, extension, and suggestions for future research. MIS Quarterly 28 (1), 107-142.
Waring, G.F., 1996. Industry differences in the persistence of firm-specific returns. American Economic Review 86 (5), 1253-1265.
Weisbach, M.S., 1988. Outside directors and CEO turnover. Journal of Financial Economics 20(1/2), 431-460.
Yang, Y., 2006. The Taiwanese notebook computer production network in China. Center on China's Transnational Relations Research Workshop.
Zahra, S.A., Pearce II, J.A., 1989. Board of directors and corporate financial performance: A review and integrative model. Journal of Management 15 (2), 291-334.