Dividend Policy, Shareholder Rights, and Corporate Governance

Pornsit Jiraporn, Ph.D. Assistant Professor of Finance Great Valley School of Graduate Professional Studies Pennsylvania State University 30 E. Swedesford Road Malvern, PA 19355 Email: [email protected] Tel: (610) 725-5342

Yixi Ning, Ph.D. Assistant Professor of Finance School of Business Administration University of Houston – Victoria Sugar Land, TX 77479 Email: [email protected]

Date: September 18, 2006

Most of the empirical work was conducted while the first author was affiliated with Texas A&M International University (TAMIU) in Laredo, Texas. The first author thanks TAMIU for a summer research grant, which made this project possible. We thank an anonymous referee for helpful comments and suggestions. Naturally, all remaining errors are ours.

Dividend Policy, Shareholder Rights, and Corporate Governance Abstract Grounded in agency theory, this study explores agency costs as a determinant of dividend policy. Specifically, we examine how dividends are related to the strength of shareholder rights. The evidence reveals an inverse association between dividend payouts and shareholder rights, indicating that firms pay higher dividends where shareholder rights are more suppressed. This evidence is consistent with the substitution hypothesis (La Porta et al., 2000), which contends that firms with weak shareholder rights need to establish a reputation for not exploiting shareholders. As a result, these firms pay dividends more generously than do firms with strong shareholder rights. In other words, dividends substitute for shareholder rights. Finally, there is evidence that regulation influences the association between dividends and shareholder rights.

JEL Classification: G30, G32, G34 Keywords: dividend policy, shareholder rights, corporate governance

2

Dividend Policy, Shareholder Rights, and Corporate Governance Dividend payouts have been argued to mitigate agency conflicts by reducing the amount of “free” cash flow available to managers, who do not necessarily act in the best interests of the shareholders (Grossman and Hart, 1980; Easterbrook, 1984; Jensen, 1986). Furthermore, Easterbrook (1984) argues that dividends help alleviate agency conflicts by exposing firms to more frequent monitoring by the primary capital markets because paying dividends increases the probability that new common stock has to be issued more often. Grounded in agency theory, this article contends that dividend payouts are influenced by the severity of agency costs and agency costs, in turn, are related to the strength of shareholder rights (Gompers, Ishii, and Metrick, 2003). Therefore, there should be a relationship between dividend payouts and the strength of shareholder rights. We advance two competing hypotheses. The first hypothesis is primarily predicated upon the free cash flow hypothesis (Jensen, 1986). The managerial opportunism hypothesis argues that opportunistic managers are likely to retain cash within the firm so they may consume more perquisites, engage in empire building, and invest in projects and acquisitions that may enhance their personal prestige but not necessarily be beneficial to shareholders. Firms where shareholder rights are weak are prone to managerial opportunism because managers are given vast discretion with little monitoring from shareholders. If this is the case, then, managers of firms with weak shareholder rights should opportunistically attempt to retain cash within the firm rather than pay it out to shareholders. The empirical prediction of this hypothesis is that a positive association

3

should be observed between dividend payouts and the strength of shareholder rights – the weaker the shareholder rights, the less paid out in dividends. On the contrary, the substitution hypothesis is based on an argument made by La Porta, Lopez, Shleifer, and Vishny (2000).1 This view argues that dividends substitute for shareholder rights. This argument relies critically on the need for firms to raise money in the external capital markets, at least occasionally. To be able to raise external funds on attractive terms, a firm must establish a reputation for moderation in expropriating shareholders. One way to establish such a reputation is by paying dividends, which reduces what is left for expropriation (La Porta et al., 2000). A reputation for good treatment of shareholders is worth the most for firms with weak shareholder rights. As a result, the need for dividends to establish a reputation is the greatest for such firms. By contrast, for firms where shareholder rights are strong, the need for a reputation mechanism is weaker, and, thus, so is the need to pay dividends. This view, therefore, posits that, all else equal, dividends payout should be higher in firms with weaker shareholder rights. In other words, an inverse relationship should be observed. Following Gompers, Ishii, and Metrick (2003), we employ the Governance Index to gauge the strength of shareholder rights. The Governance Index measures the strength of shareholder rights by determining how many corporate governance provisions exist that restrict shareholder rights. The empirical results in our study indicate that the strength of shareholder rights does affect dividend payouts. Specifically, the relationship

1

La Porta et al. (2000) examine dividend policies of firms in 33 countries and provide strong empirical support for agency theory as an explanation of dividend payouts.

4

is negative, i.e. firms where shareholder rights are more restricted pay higher dividends. This evidence is in favor of the substitution hypothesis. It is worth noting that our evidence is different from La Porta et al. (2000), who do not find support for the substitution hypothesis. Perhaps, the difference stems from the fact that La Porta et al. (2000) examine dividend policies across disparate legal systems whereas our study investigates variation in dividend payouts within the U.S. legal system. Our studies, however, are similar in spirit for we both concentrate on the link between dividend policy and shareholder protection. Finally, we investigate how regulation may influence the relationship between dividend policy and shareholder rights. Regulation is likely to affect agency costs. Because regulators already provide a certain degree of monitoring, managers of regulated firms should be less able to reap private benefits at the expense of shareholders (Booth, Cornett, and Tehranian, 2002; Kole and Lehn, 1997). This potential reduction in agency costs may have implications for the association between dividend payouts and shareholder rights. Accordingly, we explicitly distinguish between regulated2 and unregulated firms and find that regulation does have an impact. I. Sample Selection and Data A. Sample Selection The original sample in this study is compiled from the Investor Responsibility Research Center (IRRC) corporate governance database. The IRRC collects data on corporate governance provisions, which we use as the measure of the strength of shareholder rights, for about 1500 firms from various sources, such as annual reports, proxy statement, and SEC 10-Q and 10-K documents. The sample of IRRC firms, mainly 2

In this article, we employ utility firms to represent regulated companies.

5

drawn from the S&P 500 and other large corporations, represents over 90% of total market capitalization on NYSE, AMEX, and NASDAQ. However, the IRRC expanded the sample by adding several hundreds of smaller firms with high-institutional-ownership firms in 1998.3 The IRRC collects data only periodically and our sample is, therefore, restricted to the years in which the IRRC has data on corporate governance. For this study, we use the data from 1993, 1995, 1998, 2000, and 2002.4 We then eliminate firms whose accounting data are not available in COMPUSTAT. Financial firms, whose SIC codes fall between 6000 and 6999, are excluded because they have different accounting and financial characteristics. Table 1 displays the year distribution of the final sample. The final sample consists of 3,732 firm-year observations. -----Insert Table 1 about here-----

B. The Governance Index (GINDEX) To measure the strength of shareholder rights, we employ the Governance Index (GINDEX) developed by Gompers, Ishii and Metrick (2003), henceforth GIM. They use data from the Investor Responsibility Research Center (IRRC), which publishes detailed listings of corporate governance provisions for individual firms in Corporate Takeover Defenses (Rosenbaum 1993, 1995, and 1998, 2000, 2002).5 The data on governance

3

Since the IRRC database only included large corporations before 1998, and added several hundred firms in 1998, the empirical results using the full sample may be biased towards large corporations. As a robustness check , we employ only the data from 1998 and later years and repeat our analysis. The findings remain similar. 4 The data for 1990 are available. However, some variable definitions were changed between 1990 and 1993 by the IRRC, causing possible inconsistency. To avoid any confusion, we do not include the data from 1990. 5 This index has been widely employed in a large number of recent studies. For instance, the Governance Index has been related to capital structure (Jiraporn and Gleason, 2006), to the cost of debt financing (Klock, Mansi, and Maxwell, 2004), to the cost of bank loans (Chava, Dierker, and Livdan, 2005), to the

6

provisions are derived from various sources, such as corporate bylaws, charters, proxy statements, annual reports, as well as 10-K and 10-Q documents filed with the Security and Exchange Commission (SEC). Table 2 displays the individual governance provisions included in the construction of the Governance Index. The detailed explanation for each governance provision is available in the appendix of GIM. They classify provisions into 5 categories: tactics for delaying hostile bidders (Delay); voting rights (Voting); director/officer protection (Protection); other takeover defenses (Other); and state laws (State). -----Insert Table 2 about here----The Governance Index is constructed as follows; for every firm, GIM add one point for every provision that restricts shareholder rights (increase managerial power). While this index may not accurately reflect the relative impacts of the various provisions, it has the advantage of being transparent and easily reproducible. The index does not require any judgments about the efficacy or wealth effects of any of these provisions; GIM only considers the impact on the balance of power. To clarify the logic behind the construction of the Governance Index, GIM use the following example in their paper: For example, consider classified boards, a provision that staggers the terms and elections of directors and, thus, can be employed to slow down a hostile takeover. If management uses this power judiciously, it could possibly lead to an increase in overall shareholder wealth; if management, however, uses this power to maintain private benefits of control, then this provision would diminish shareholder wealth. In either case, it is apparent that classified boards enhance the power of managers and weaken the control rights of large shareholders, which is all that matters for constructing the index. (p. 114) cost of equity (Huang, 2005), to corporate diversification (Jiraporn, Kim, Davidson, and Singh, 2006), to CEO compensation (Jiraporn, Kim, and Davidson, 2005), and auditor selection (Jiraporn, 2006).

7

Most provisions other than classified boards can be viewed by the same logic. Almost every provision enables management to resist different types of shareholder activism, such as calling special meetings, changing the firm’s charter or bylaws, suing directors, or replacing them all at once. GIM note, however, that there are two exceptions, secret ballots (confidential voting) and cumulative voting. A secret ballot or confidential voting designates a third party to count proxy votes and, therefore, prevents management from observing how specific shareholders vote. Cumulative voting enables shareholders to concentrate their directors’ votes so that a large minority shareholder can ensure some board representation. These two provisions are usually proposed by shareholders and opposed by management because they enhance shareholder rights and diminish the power of management. GIM add one point to the Governance Index when firms do not have each of these provisions. For all other provisions, GIM add one point when firms do have each of them. The Governance Index is the sum of one point for the presence (or absence) of each provision. II. Empirical Results A. Descriptive Statistics Table 3 shows the descriptive statistics for the Governance Index, which is utilized to gauge the strength of shareholder rights. A high value for the index indicates that a firm has weak shareholder rights, and a low value indicates that the firm has strong shareholder rights. The Governance Index averages 9.36 (9.00 median) in our sample, suggesting that, on average, the sample firms impose 9 governance provisions that restrict shareholder rights. The median firm in GIM has 9 governance provisions as well. Thus, it

8

seems that our sample is reasonably representative of the slightly larger sample used by GIM.6 -----Insert Table 3 about here----We employ three alternative measures of dividends. We use the dividend/earnings ratio, the dividend/sales ratio, and the dividend yield7. Table 3 Panel B shows the descriptive statistics for the dividend payout ratios. The dividend/earnings ratio averages 0.2966 (0.1263 median) whereas the dividend/sales ratio averages 0.0179 (0.0086 median). The average dividend yield is 1.7% (1.0% median). The percentage of firms paying dividends in our sample is 69.02%. Table 3 Panel C shows the summary statistics for the sample of firms. The average firm in the sample has 4,168 million dollars in sales (1,292 median) and 4,520 million dollars (1,330 median) in total assets, suggesting that our sample firms are large. The long-term debt to total assets ratio averages 23.65% (22.66% median), indicating many firms in our sample have significant long-term debt in their capital structure. On average, the sample firms are profitable as evidenced by the average EBIT ratio of 12.48% (13.60% median)8. Tobin’s Q9, which proxies for growth opportunities, averages 1.43 (1.06 median). Finally, Table 3 Panel D shows the descriptive statistics for the board of directors, the average (median) percentage of independent directors is 64.22% (66.67%) while the average board size is 9.33(9.00). The board size in our sample is 6

The Governance Index ranges from 2 to 18. Gompers et al (2003) classify a firm whose Governance Index is less than 5 as a democracy whereas and one whose Index is greater than 14 is regarded as a dictatorship. Hence, it appears that our sample encompasses some “democratic” and “dictatorial” firms. 7 The dividend yield is the cash dividend divided by the stock price (end-of-year price). 8 Another measure of a firm’s profitability is return on assets (ROA). The mean (median) ROA for our sample firms is 3.21% (4.28%). We use ROA instead of EBIT ratio to repeat all regressions in Table 5 and obtain very similar findings. 9 Tobin’s Q is calculated based on Chung and Pruitt (1994). Tobin’s Q is the sum of the market value of equity and the book value of debt divided by the book value of assets. The book value of debt is the difference between the book value of assets and the book value of equity.

9

similar to that in Denis and Sarin (1999), who employ a sample of randomly-selected CRSP-listed firms, but the proportion of independent directors is higher than their randomly-chosen sample. B. Univaritate analysis A univariate analysis is conducted and the results are displayed in Table 4. To determine the association between dividends and shareholder rights, we create two extreme portfolios based on the Governance Index as in GIM. Firms where the Governance Index is equal to or higher than 14 are included in the “dictatorship” portfolio for shareholder rights are severely restricted in these firms. On the contrary, firms where the Governance Index is lower than or equal to 5 are placed in the “democracy” portfolio as shareholders enjoy strong shareholder rights in these firms. -----Insert Table 4 about here----Then, we compare dividend payouts between the dictatorship portfolio with 232 firms and the democracy portfolio with 330 firms. The results in Table 4 indicate that the dividend/earnings ratio averages 0.8295 (0.2693 median) for the dictatorship portfolio and 0.0428 (0.0000 median) for the democracy portfolio. The difference is statistically significant at the 5% level. The results also reveal that the average dividend/sales ratio is 0.0176 (0.0152 median) for the dictatorship portfolio and 0.0136 (0.0017 median) for the democracy portfolio. The difference is statistically significant at the 10% level. Finally, the dividend yield averages 0.0209 (0.0197 median) for the dictatorship portfolio and 0.0116 (0.00 median) for the democracy portfolio. The difference is significant at the 0.1% level. We also find that the percentage of dividend-paying firms in the dictatorship portfolio (83.19%) is much higher than that in the democracy portfolio (53.64%).

10

Thus, it appears that firms with more restrictive governance are more likely to pay dividends, and dividend-paying firms pay more dividends than do those with more liberal governance. This is evidence in favor of the substitution hypothesis. Firms with suppressive governance (and, hence, weak shareholder rights) need to establish a reputation for not exploiting shareholders. One way to do so is by distributing cash through dividend payouts. Hence dividends serve as a substitute for shareholder rights. These univariate results, however, may merely reflect differences in firm-specific attributes. Therefore, we further test the hypothesis using a regression analysis in the next section. C. Regression Analysis Table 5 shows the results of the regression analysis. A number of control variables are included. We employ the natural logarithm of total assets to proxy for firm size. Profitability has been found to affect dividend policy (DeAngelo and DeAngelo, 1990; DeAngelo, DeAngelo, and Skinner, 1992). Thus, we control for profitability by using the EBIT/sales ratio. Growth has been reported to impact dividend policy as well (Rozeff, 1982). We control for firm growth by using Tobin’s Q. Leverage also influences dividend policy both because of its role in mitigating agency costs and because of debt covenants on dividends imposed by debtholders. Our proxy for leverage is the ratio of long-term debt to total assets. Finally, due to the growing importance and popularity of share repurchases, we include several share repurchase10 ratios to control for this alternative means of cash distribution. Share repurchases could affect dividend payouts

10

Repurchase activity is measured as in Dittmar (2000) using COMPUSTAT item 115 adjusted for change in preferred stock.

11

because cash spent on repurchases could otherwise be distributed as dividends (Grullon and Michaely, 2002).11 -----Insert Table 5 about here----In Model 1 where the dividend/earnings ratio is the dependent variable, the Governance Index produces a positive and significant coefficient (t = 3.33, significant at 0.1% confidence level). Likewise, in Model 2 where the dependent variable is the dividend/sales ratio, the coefficient of the Governance Index is positive and significant (t = 3.97, significant at 0.1% level). Similarly, the Governance Index shows a positive and significant coefficient (t = 1.83, significant at 10% level) in Model 3 when we employ the dividend yield as the dependent variable. The results indicate a positive association between dividends and the Governance Index. This implies that firms where shareholder rights are weak (high Governance Index), pay higher dividends. This is in accordance with the substitution hypothesis. Firms with weak shareholder rights pay out higher dividends to alleviate the perception that shareholders are exploited. In other words, high dividends compensate for weak shareholder rights in these firms. As a robustness check, we run alternative regressions allowing individual firm fixed-effects. Fich and Shivdasani (2005) champion the use of fixed-effects regressions because this approach is robust to the presence of omitted firm-specific variables. The results remain qualitatively similar and, therefore, appear to be robust to the estimation method.12

11

Since cross-sectional dividend payout models are sensitive to the selection of control variables, we conduct a sensitivity analysis by using alternative control variables. For example, we use sales instead of total assets to measure firm size, total debt ratio in lieu of financial leverage, market-to-book ratio for growth opportunities, and return on assets (ROA) rather than EBIT ratio to proxy for firm profitability. The alternative specifications yield similar results. 12 The fixed-effects results are not shown but available upon request.

12

Finally, in Model 4, we explore whether shareholder rights affect the decision to pay a dividend of any size or not pay anything at all. The dependent variable is a dichotomous variable equal to 1 if the firm pays a dividend of any size and 0 if the firm does not pay a dividend at all. The results of the logistic regression reveal that the Governance Index has a positive and significant coefficient. This implies that firms with more restrictive governance (hence, weaker shareholder rights) are more likely to pay dividends, which further confirms our findings from the univariate analysis. Again, this evidence is in agreement with the prediction of the substitution hypothesis. Incidentally, it is worth pointing out that two of the three repurchase ratios (repurchase/earnings and repurchase/sales) and the repurchase dummy in Table 5 show positive and significant coefficients. This suggests that firms that pay more dividends tend to spend more on repurchases as well. Dividends and repurchases are not found to substitute for each other, at least, in our sample. The board of directors is a crucial governance mechanism. Hence, the extent of agency conflicts may be affected by the structure of the board. For example, firms with weak shareholder rights may not suffer severe agency costs if they have strong boards. Since our hypothesis on dividend policy is predicated on agency theory, it would be appropriate to control for board structure. Two features of the board that have garnered enormous attention in the literature are board composition and board size. 13 Therefore, we collect data on the percentage of independent directors and on board size from the

13

For board composition, see Hermalin and Weisbach (1991), Cotter, Shivdasani, and Zenner (1997), Mayers, Shivdasani, and Smith (1997), Bhagat and Black (2001), among others. For board size, see Lipton and Lorsch (1992), Jensen (1993), Yermack (1996), Denis and Sarin (1999), and Eisenberg, Sundgren, and Wells (1998), among others.

13

IRRC director database. The data on board structure are available for 2,283 observations (out of 3,723). In Table 6, we replicate the regressions in Table 5 but include board composition and board size as control variables. As shown in Table 6, the coefficients of the Governance Index in all models are positive and significant. Thus, even when board structure is taken into account, the inverse relationship between shareholder rights and dividend payouts still remains. Finally, we capture the temporal variation in dividend policy by including the year dummies. Fama and French (2001) document that dividend payouts decline over time. In additional regressions (results not shown), we include the year dummies and obtain qualitatively similar results for the Governance Index. Therefore, it does not appear that the declining dividend payouts documented in Fama and French (2001) materially affect our results.14 -----Insert Table 6 about here----D. Possible Endogeneity between Shareholder Rights and Dividend Policy We have found empirical evidence, supporting the substitution hypothesis that firms with weak shareholder rights are more likely to pay out dividends more generously. We draw the conclusion assuming that the firm’s dividend policy is caused by the strength of shareholder rights. However, numerous prior studies have documented the potential effects of dividend payments on agency conflicts between shareholders and managers (Grossman and Hart, 1980; Easterbrook, 1984; Jensen, 1986). Therefore, it is conceivable that a firm’s dividend policy may affect its shareholder rights as measured by

14

Furthermore, Fama and French (2001) report that firms that pay less dividends tend to be small firms with less profitability and strong growth opportunities. As a result, we control for firm size, profitability, and growth opportunities in this study.

14

the Governance Index. In light of this argument, we further examine a possible endogenous relation between dividend policy and shareholder rights. Shareholder rights and dividend policy may have a bi-directional causality. One way to test the endogenous relation between shareholder rights and dividend policy is to construct a simultaneous equation model that can be estimated by 2-stage or 3-stage least square regressions. However, it is difficult to find truly exogenous instruments variables for the model (Himmelberg, Hubbard, and Palia, 1999; Gompers, Ishii, and Metrick, 2003). As a consequence, we follow Chi (2005) in exploring the correlations between the changes in Governance Index and the changes in dividend payments. If the changes in the Governance Index are correlated with future changes in dividend payments, then it is more likely that shareholder rights influence dividend policy. On the contrary, if the changes in Governance Index are correlated with past changes in dividend payments, then it is more likely that dividend policy affects the strength of shareholder rights. We conduct the correlations between the changes of GINDEX and the three dividend variables respectively15. We find some evidence that the changes in the Governance Index are positively correlated with the future changes in the dividend/sales ratio (p = 0.012) and the dividend yield (p = 0.020) from Year +1 to Year + 2, indicating that it is more likely that the strength of shareholder rights affects dividend policy. On the other hand, we do not find any significant evidence that dividend policy affects the Governance Index.

15

The correlation analysis between the changes in shareholder rights and dividend policies are not shown in tables but available upon request.

15

We further explore the causal relationship between dividend policy and shareholder rights using a Granger causality test (Granger, 1969)16. The Granger causality analysis tests the condition that changes in the cause variables should precede changes in the effect variables. So we consider the following models: DIV t = α1 + δ1 GI t-1 + δ2 DIV t-1 + Other control variables + µt GI t = β1 + γ1 GI t-1 + γ2 DIV t-1 + Other control variables + εt In the models, µt and εt are uncorrelated error terms. If δ1 ≠ 0 and γ2 ≠ 0, we infer an endogenous bi-directional relationship between shareholder rights and dividend policy. If δ1 ≠ 0 but γ2 = 0, it is more likely that the Governance Index affects dividend policy. If δ1 = 0 but γ2 ≠ 0, it is more likely that dividend policy affects shareholder rights. If both δ1 = 0 and γ2 = 0, dividend policy has no relation with shareholder rights. ----- Insert Table 7 about here ----First, we use the model to test the causal relation between the Governance Index and the ratio of cash dividend to earnings. We find that δ1 = 0.055, which is significantly different from zero (t = 2.69, significant at the 1% level). But γ2 is not different from zero statistically (t = -0.79, insignificant). Hence, we infer unidirectional and positive Granger causality from shareholder rights to the ratio of cash dividend to earnings. Similarly, we test the causal relations between the Governance Index and the cash dividend/sales, as well as between the Governance Index and the dividend yield. We also document similar uni-directional Granger causality from the Governance Index to the cash dividend/sales

16

A typical caveat is that Granger-causality can establish that one variable helps forecast or predict the other variable. While this may be a necessary condition for causality, it is certainly not a sufficient condition since causality should be grounded a priori on theory and should not be inferred from empirical data. Hence, we do recognize that true causality may not be determined simply by using this methodology. we only make a modest claim that Granger-causality can increase the probability that the alleged causal relationship does in fact exist.

16

ratio (δ1 ≠ 0, t = 2.67; γ2 = 0, t = 0.24), and from the Governance Index to the dividend yield (δ1 ≠ 0, t = 5.84; γ2 = 0, t = 0.24). The Governance Index has been found to have a positive impact on both the dividend/sales ratio and the dividend yield. Based on the correlation analysis and the Granger causality test shown in Table 7 and Table 8, we conclude that there is no endogenous relation between a firm’s shareholder rights and its dividend policy. While a firm’s Governance Index positively affect dividend policies, supporting the substitute hypothesis, the firm’s dividend policy has no observable effects on the strength of shareholder rights. E. Regulated Firms As discussed earlier, the association between dividends and shareholder rights may be impacted by regulation. Regulation serves as an external mechanism that helps control agency costs. Hence, regulated firms may rely, to a lesser extent, on dividends to mitigate agency conflicts. To test the impact of regulation, we segregate the full sample into industrial (unregulated) firms and regulated firms and re-estimate the regressions. The regulated subsample includes 367 utility firms whose SIC codes fall between 4900 and 4999. The rest of the firms in the full sample are classified as industrial unregulated firms 17. -----Insert Table 8 about here----The results using the industrial (unregulated) subsample are qualitatively similar to those using the full sample (results not shown). However, the results using the regulated utility subsample are different. Table 8 displays the results of the regressions using only the regulated subsample. In Model 1, the Governance Index does not show a

17

Financial firms (SIC codes 6000-6999) are not included due to their different accounting and financial characteristics.

17

significant coefficient (t = 1.64). In Model 2, the coefficient of the Governance Index is significant but negative (t = -1.76, significant at 10% confidence level). Finally, in Model 3, the Governance Index produces an insignificant coefficient (t = -0.56) when we use the dividend yield as the dependent variable. Hence, none of the results here suggest a positive relationship. This is in contrast to the results obtained earlier from the full sample and from the industrial sample. It seems that regulation does affect the association between dividends and shareholder rights. The positive relationship reported earlier does not appear to exist in regulated firms. Finally, we estimate a logistic regression in Model 4 where the dependent variable is a dummy variable equal to 1 if the firm pays a dividend of any size and 0 if it does not pay anything at all. The coefficient of the Governance Index is positive and significant, implying that regulated firms with weaker shareholder rights are more likely to pay dividends (as opposed to not paying anything at all). However, for firms that do pay dividends of any size, the amount of the dividend is not related to the strength of shareholder rights. Thus, there is partial but not complete support for the substitution hypothesis here. F. Robustness Checks A number of robustness checks are conducted to confirm the results. First, out of concern that the results may be driven by industry effects, we industry-adjust all the variables and re-estimate the regressions. The industry adjustment for each variable is accomplished by subtracting the industry median value from each given variable. We use the first 2 digits of the SIC to identify the industry. The results remain qualitatively the same even after the adjustment and, therefore, do not seem to be susceptible to the

18

industry effects. Second, for fear that outliers may have unduly affected the results, we exclude the extreme 5% of the observations and re-estimate the regressions. Again, the results remain similar, suggesting that outliers do not pose a problem in our dataset. Furthermore, to control for variation across time, we create the year dummies and include them in the regressions. The results still remain similar after the inclusion of the year dummies. Third, since preferred stocks are hybrid securities and have some features of liabilities (e.g., fixed amount of preferred dividend payments), one argument is that we should exclude preferred dividends in the dividend measures. To address this concern, we first conduct a Pearson correlation test between common dividends and total cash dividends. The Pearson correlations coefficient for the two dividend measures is as high as 99.9%. In addition, we calculate the ratios of common dividends to earnings, common dividend to sales, and common dividend yield18 and, then, repeat the previous univariate analysis and multiple regressions. We obtain similar results as before. Finally, it can be argued that dividend payouts are a censored variable as it cannot be below zero. As a result, we replicate all the regressions using the Tobit analysis and obtain similar results. Since the results have been subject to so many robustness checks and still do not change qualitatively, they seem to be robust. III. Concluding Remarks The objective of this study is to examine the impact of shareholder rights on dividend payouts. The empirical evidence demonstrates that dividend payouts are inversely related to the strength of shareholder rights. Firms where shareholder rights are weak pay out higher dividends. This is in agreement with the substitution hypothesis, 18

In other words, we exclude preferred dividends.

19

which posits that firms where shareholder rights are weaker try to establish a good reputation for not taking advantage of shareholders by paying more out as dividends. Hence, dividends compensate for the weak shareholder rights. Further analysis also reveals that regulation does influence the association between dividends and shareholder rights. This study is conducted in the same spirit as La Porta et al. (2000) for both of our studies relate dividend payouts to agency problems. The empirical results, however, are dissimilar. La Porta et al. (2000) do not find support for the substitution hypothesis while we do in this study. We conjecture that the dissimilarity of our results, perhaps, lie in the fact that the external capital markets in the U.S. are highly developed and, thus, provide better monitoring. Therefore, the need to establish a favorable reputation in order to raise capital on attractive terms is stronger here than elsewhere in the world. As a result, the substitution hypothesis is supported when we look at the variation across firms but within the U.S. legal system as opposed to La Porta et al. (2000) who examine dividend policies across disparate legal systems. In any case, both of our studies agree that the agency approach is highly relevant to our understanding of corporate dividend policy. This study contributes to the literature both in dividend policy and in agency theory. As far as dividend policy, we show that shareholder rights are a significant determinant of dividend policy. For the literature in agency theory, this study shows that dividends are utilized as a tool to establish a reputation for favorable treatment of shareholders for firms where shareholder rights are weak (hence, agency costs are likely severe).

20

References Allen, F. and R. Michaely, 2002. Payout policy. Working paper, University of Pennsylvania. Bhagat, S., 1986. The effect of management’s choice between negotiated and competitive offerings on shareholder wealth. Journal of Financial and Quantitative Analysis 2, 181-195. Bhagat, S. and R. Jefferies, 2002. The Econometrics of Corporate Governance Studies. MIT press, Boston. Bhagat, S. and B. Black, 2001. The non-correlation between board independence and long term performance. Journal of Corporation Law 27, 231-274. Berle, A.A. and G.C. Means, 1932. The Modern Corporation and Private Property. Macmillan, New York, NY. Bhattacharya, S., 1979. Imperfect information, dividend policy, and ‘the bid in the hand’ fallacy. Bell Journal of Economics 10 (1), 259-270. Booth, J. R., M.M. Cornett,, and H. Tehranian, 2002. Boards of directors, ownership, and regulation. Journal of Banking and Finance 26, 1973-1996. Chava, S., M. Dierker, and D. Livdan, 2005, Do shareholder rights affect the cost of bank loans? Working paper, University of Houston. Chi, J., 2005. Understanding the endogeneity between firm value and shareholder rights. Financial Management, forthcoming. Chung K., and S. Pruitt, 1994. A simple approximation of Tobin’s Q. Financial Management 23, 70-75.

21

Cotter, J., A. Shivdasani, and M. Zenner, 1997.Do independent directors enhance target shareholder wealth during tender offers? Journal of Financial Economics 43, 195218. DeAngelo, H. and L. DeAngelo, 1990. Dividend policy and financial distress: An empirical investigation of troubled NYSE firms. Journal of Finance 45, 1415-1431. DeAngelo, H., L. DeAngelo, and D. Skinner, 1992. Dividends and losses. Journal of Finance 47, 1837-1864. Del Guercio, 1996. The distorting effect of the prudent-man laws on institutional equity investments. Journal of Financial Economics 40, 31-62. Denis, D., and A. Sarin, 1999. Ownership and board structures in publicly traded corporations. Journal of Financial Economics 52, 187-224. Dittmar, A., 2000. Why do firms repurchase stock? Journal of Business 73, 331-355. Easterbrook, F. H., 1984. Two agency-cost explanations of dividends. American Economic Review 74 (4), 650-659. Eisenberg, T., S. Sundgren, S, and M. Wells, 1998, Larger board size and decreasing firm value in small firms. Journal of Financial Economics 48, 35-54. Fama, Eugene and French Kenneth, 2001, Disappearing dividends: changing firm characteristics or lower propensity to pay? Journal of Financial Economics 60, 3-43. Fich, E. and A. Shivdasani, 2005. Are busy boards effective monitors. Journal of Finance, forthcoming. Gompers, P., J. Ishii, and A. Metrick, 2003. Corporate governance and equity prices. Quarterly Journal of Economics 118, 107-155.

22

Granger, C.W.J., 1969. Investigating casual relations by econometric models and crossspectral methods. Econometrica 37(3), 424-438. Grossman, S. J. and O.D. Hart, 1980. Takeover bids, the free-rider problem, and the theory of the corporation. Bell Journal of Economics 11, 42-54. Grullon, G. and R. Michaely, 2002. Dividends, share repurchases, and the substitution hypothesis. Working paper, Rice University. Hansen, R. S. and P. Torregrosa, 1992. Underwriter compensation and corporate monitoring. Journal of Finance 47, 1537-1556. Hermalin, B.E. and M.S. Weisbach, 1991. The effects of board composition and direct incentives on firm value. Financial Management 20, 101-112. Himmelberg, C.P., R.G. Hubbard, and D. Palia,1999. Understanding the determinants of managerial ownership and the link between ownership structure and performance. Journal of Financial Economics 53, 353-384. Huang, H., 2004, Shareholder rights and the costs of capital. Working paper, University of Houston. Jain B.and O. Kini, 1999. On investment banker monitoring in the new issues market. Journal of Banking and Finance 23, 49-84. Jensen, M. C., 1986. Agency costs of free cash flow, corporate finance and takeovers. American Economics Review 76, 323-339. Jensen, M., 1993. The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance 48 (3), 831-880. Jensen, M. C. and W. Meckling, 1976. Theory of the firm: managerial behavior, agency costs, and capital structure. Journal of Financial Economics 3, 305-360.

23

Jiraporn, P., 2006, Corporate governance, shareholder rights, and Arthur Andersen, Journal of Applied Finance, forthcoming. Jiraporn, P. and K. Gleason, 2006. Capital structure, shareholder rights, and corporate governance. Journal of Financial Research, forthcoming. Jiraporn, P., Y.S. Kim, W.N. Davidson, 2005, CEO compensation, shareholder rights, and corporate governance: An empirical investigation, Journal of Economics and Finance 29, 242-258. Jiraporn, P., Y.S. Kim, W.N. Davidson, and M. Singh, 2005, Corporate governance, shareholder rights, and firm diversification: An empirical analysis. Journal of Banking and Finance 30 (30), 947-963. Kalay, A., 1982. The ex-dividend day behavior of stock prices: A re-examination of the clientele effect. Journal of Finance 37, 1059-1070. Klock, M., S. Mansi, and W. Maxwell, 2005, Does corporate governance matter to bondholders? Journal of Financial and Quantitative Analysis, forthcoming. Kole, S., and K. Lehn, 1997. Deregulation, the evolution of corporate governance structure and survival. American Economic Review 87, 421-425. La Porta, R., F. Lopez-De Salinas, A. Shleifer, and R. Vishny, 2000. Agency problems and dividend policy around the world. Journal of Finance 55, 1-33. Lang, L. H. P., and R.H. Litzenberger, 1989. Dividend announcement: Cash flow signaling vs. Free cash flow Hypothesis. Journal of Financial Economics 24 (1), 181192. Lipton, M. and J. Lorsch, 1992, A modest proposal for improved corporate governance. Business Lawyer 48, 59-77.

24

Litzenberger, R. and K. Ramaswamy, 1979. The effects of personal taxes and dividends on capital asset prices: Theory and empirical evidence. Journal of Financial Economics 7, 163-195. Mayers, D., A. Shivdasani, and C. Smith, 1997. Board composition and corporate control, evidence from the insurance industry. Journal of Business 70, 33-62. Miller, M. and F. Modigliani, 1961. Dividend policy, growth and the valuation of shares. Journal of Business 34, 411-433. Miller, M. and K. Rock, 1985. Dividend policy under asymmetric information. Journal of Finance 40 (4), 1031-1051. Miller, M., and M. Scholes, 1978. Dividends and taxes. Journal of Financial Economics 6, 333-264. Miller, M., and M. Scholes, 1982. Dividends and taxes: empirical evidence. Journal of Political Economy 90, 1118-1141. Rosenbaum, V., 1993, 1995, 1998, 2000, 2002. Corporate Takeover Defenses. Investor Responsibility Research Center Inc., Washington D.C. Rozeff, M., 1982. Growth, beta, and agency costs as determinants of dividend payout ratios. Journal of Financial Research 5, 249-259. Smith, Jr., C. W., 1977. Alternative methods for raising capital: Rights versus Underwritten offerings. Journal of Financial Economics 5, 273-307. Yermack, D., 1996, Higher market valuation of companies with a smaller board of directors. Journal of Financial Economics 40, 185-211. Yoon, P. and L. Starks, 1995. Signaling, investment opportunities, and dividend announcements. Review of Financial Studies 8 (4), 995-1018.

25

Table 1: Sample Distribution by Year The sample is derived from the Investor Responsibility Research Center (IRRC). Firms are excluded whose financial data are not available on COMPUSTAT. IRRC collects data on corporate governance provisions only periodically. Hence, our sample is limited to the years 1993, 1995, 1998, 2000, and 2002. Year

N

Percent

1993

598

16.02%

1995

650

17.42%

1998

843

22.59%

2000

804

21.54%

2002

837

22.43%

Total

3,732

100.00%

26

Table 2: Individual Governance Provisions Employed in the Construction of the Governance Index The detailed explanation for each governance provision is available in the Appendix of Gompers, Ishii, and Metrick (2003)

Delay Blank Check Classified Board Special Meeting Written Consent Protection Compensation Plans Contracts Golden Parachutes Indemnification Liability Severance Voting Bylaws Charter Cumulative Voting Secret Ballot (Confidential Voting) Supermajority Unequal Voting

Other Anti-greenmail Directors' duties Fair Price Pension Parachutes Poison Pill Silver Parachutes State Anti-greenmail Law Business Combination Law Cash-out Law Directors' Duties Law Fair Price Law Control Share Acquisition Law

Note: Cumulative Voting and Secret Ballot (Confidential Voting) are usually proposed by shareholders and opposed by management because they enhance shareholder rights and diminish management powers. GIM add one point to the Governance Index when firms do not have each of these provisions. For all other provisions, GIM add one point when firms do have each of them. The Governance Index is the sum of one point for the presence (or absence) of each provision.

27

Table 3: Descriptive Statistics Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Mean

Median

S.D.

Governance Index

9.36

9.00

2.73

N

3,732

3,732

3,732

Cash Dividends

90.67

9.20

305.92

Cash Dividends / Earnings

0.2966

0.1263

2.8973

Cash Dividends / Sales

0.0179

0.0086

0.0293

Dividend Yield

0.0170

0.0100

0.0577

% of Dividend-Paying Firms

69.02%

-

-

Sales

4,168

1,292

9,842

Total Assets

4,520

1,330

10,033

Leverage

23.65%

22.66%

15.65%

EBIT Ratio

12.48%

13.60%

86.97%

Tobin’s Q

1.43

1.06

1.23

64.22%

66.67%

18.42%

9.33

9.00

2.62

Panel A: Governance

Panel B: Dividends

Panel C: Firm Characteristics

Panel C: Board Characteristics % of Independent Directors Board size

28

Table 4: Univariate Analysis of the Dictatorship and Democracy Portfolios The dictatorship portfolio represents firms where the Governance Index is equal to or greater than 14 while the democracy portfolio includes firms whose Governance Index is equal to or less than 5.Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Means are shown in the table, medians in parentheses.

Dictatorship

Democracy

Difference (t-statistics)

Cash Dividends/Earnings

0.8295 (0.2693)

0.0428 (0.0000)

-2.01**

Cash Dividends/Sales

0.0176 (0.0152)

0.0136 (0.0017)

-1.87*

Dividend Yield

0.0209 (0.0197)

0.0116 (0.0000)

-3.83***

% of Dividend-paying Firms

0.8319 (-)

0.5364 (-)

-

Total Assets

3,382 (2,033)

3,964 (756)

0.89

Leverage

26.17% (24.97%)

23.70% (21.73%)

-1.74*

EBIT Ratio

12.40% (12.78%)

12.79% (14.48%)

0.14

Tobin’s Q

1.26 (1.06)

1.55 (1.11)

3.16***

69.74% (72.73%)

52.62% (50.00%)

6.77***

10.28 (10.00)

8.16 (8.00)

-8.40***

232

330

% of Independent Directors19

Board Size

N

19

The data on the percentage of outside directors and board size are available for only 137 and 183 observations in the dictatorship and democracy portfolio respectively.

29

Table 5: Regressions of dividend payouts on the Governance Index and Controls Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Repurchase activity is measured as in Dittmar (2000) using COMPUSTAT item 115 adjusted for change in preferred stock. The repurchase dummy is equal to one if the firm has repurchase activity of any amount, 0 otherwise.

Model 1

Model 2

Model 3

Model 4

(t-statistics)

(t-statistics)

(t-statistics)

(Wald-statistics)

Div./Earnings

Div./Sales

Dividend Yield

Div. Dummy

Intercept

-0.602** (-2.31)

-0.027*** (-9.83)

-0.000 (-0.05)

-3.572*** (216.41)

Governance Index

0.053*** (3.33)

0.001*** (3.97)

0.001* (1.83)

0.155*** (114.09)

Ln (Total Assets)

0.050* (1.67)

0.005*** (15.06)

0.002*** (3.65)

0.423*** (195.67)

Leverage

-0.147 (-0.52)

0.001 (0.47)

-0.001 (-0.17)

-0.727*** (8.83)

Profitability

-0.001 (-0.02)

0.002*** (3.26)

0.002* (1.90)

1.048*** (23.27)

Growth Opportunities

-0.027 (-0.74)

0.002*** (3.88)

-0.004*** (-5.17)

-0.166*** (27.34)

Repurchase/Earnings

0.362*** (27.32)

-

-

-

Repurchase/Sales

-

0.026*** (2.92)

-

-

Repurchase/MV of Equity

-

-

-0.025 (-1.23)

-

Repurchase Dummy

-

-

-

0.377*** (23.37)

3,732

3,732

3,731

3,732

F-statistics

126.96***

55.13***

8.39***

-

Adjusted R2

16.8%

8.0%

1.2%

-

-

-

-

19.2%

Independent Variable

N

Pseudo R2

*, **, *** statistically significant at the 10%, 5%, and 1% levels respectively.

30

Table 6: Regressions of dividend payouts on the Governance Index and Controls with Board Variables Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Repurchase activity is measured as in Dittmar (2000) using COMPUSTAT item 115 adjusted for change in preferred stock. The repurchase dummy is equal to one if the firm has repurchase activity of any amount, 0 otherwise. Board data are not available until 1998.

Model 1

Model 2

Model 3

Model 4

(t-statistics)

(t-statistics)

(t-statistics)

(Wald-statistics)

Div./Earnings

Div./Sales

Dividend Yield

Div. Dummy

Intercept

-1.069** (-2.03)

-0.057*** (-16.61)

-0.033*** (-2.98)

-7.670*** (278.63)

Governance Index

0.072*** (2.81)

0.0003* (1.77)

0.000* (1.83)

0.128*** (40.58)

Log of Board Size

0.216 (0.81)

0.015*** (8.57)

0.021*** (3.75)

2.173*** (104.32)

% of Independent Directors

-0.385 (-1.05)

0.018*** (7.28)

0.019** (2.49)

0.645** (5.40)

Ln (Total Assets)

0.046 (0.84)

0.002*** (6.88)

-0.001 (-0.85)

0.224*** (26.04)

Leverage

-0.067 (-0.16)

0.012*** (4.11)

0.007 (0.75)

0.774*** (5.40)

Profitability

-0.059 (-0.31)

0.008*** (6.56)

0.015*** (3.68)

0.773*** (8.39)

Growth Opportunities

-0.010 (-0.21)

0.002*** (5.69)

-0.004*** (-3.47)

-0.107*** (7.61)

Repurchase/Earnings

0.299*** (13.39)

-

-

-

Repurchase/Sales

-

0.024*** (3.28)

-

-

Repurchase/MV of Equity

-

-

-0.023 (-0.87)

-

Repurchase Dummy

-

-

-

0.385***

Independent Variable

31

(13.82) N

2,283

2,283

2,283

2,283

24.06***

71.51***

7.13***

-

Adjusted R

7.5%

19.8%

2.1%

-

Pseudo R2

-

-

-

26.9%

F-statistics 2

*, **, *** statistically significant at the 10%, 5%, and 1% levels respectively.

32

Table 7: Granger Causality Test of Shareholder Rights and Dividend Policies Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Repurchase activity is measured as in Dittmar (2000) using COMPUSTAT item 115 adjusted for change in preferred stock. The repurchase dummy is equal to one if the firm has repurchase activity of any amount, 0 otherwise. Div./ Earnings(t)

GINDEX(t)

Div./ Sales(t)

GINDEX(t)

Dividend Yield(t)

GINDEX(t)

Constant

-0.703** (-2.01)

1.38*** (11.0)

-0.012*** (-5.58)

1.382*** (10.93)

0.009*** (-3.66)

1.393*** (11.07)

Governance Index (t-1)

0.055*** (2.69)

0.895*** (122.7)

0.0003*** (2.67)

0.894*** (122.6)

0.001*** (5.84)

0.894*** (122.8)

-0.004 (-0.21)

-0.006 (-0.79)

--

--

--

--

Div./Sales(t-1)

--

--

0.515*** (45.89)

0.154 (0.24)

--

--

Dividend Yield(t-1)

--

--

--

--

0.046*** (8.70)

-0.260 (-0.93)

Ln (Total Assets)

0.056 (1.40)

-0.019 (-1.36)

0.002*** (7.30)

-0.020 (-1.37)

0.003*** (9.32)

-0.020 (-1.38)

Leverage

0.162 (0.42)

0.156 (1.13)

0.002 (0.63)

0.159 (1.15)

0.009*** (3.35)

0.163 (1.18)

Profitability

0.003 (0.06)

0.010 (0.53)

0.001 (1.63)

0.010 (0.53)

0.001 (1.66)*

0.010 (0.55)

Growth Opportunities

-0.027 (-0.58)

0.001 (0.08)

0.002*** (5.76)

0.003 (0.17)

0.003*** (-9.43)

0.000 (-0.1)

Repurchase/Earnings

0.311*** (15.60)

-0.002 (-0.29)

--

--

--

--

Repurchase/Sales

--

--

-0.001 (-0.09)

-0.168 (-0.42)

--

--

Repurchase/MV of Equity

--

--

--

--

-0.007 (-0.88)

-0.353 (-0.90)

N

2375

2375

2375

2375

2375

2375

Adjusted R2

9.4%

86.5%

52.5%

86.5%

12.4%

86.5%

36.22 ***

2178.5***

376.5***

2178.0***

48.8***

2179.5

Div./Earnings(t-1)

F

***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

33

Table 8: Regressions of dividend payouts on the Governance Index and controls for regulated sample Leverage is defined as long-term debt divided by total assets. The EBIT ratio is earnings before interest and taxes divided by total assets. Tobin’s Q proxies for growth opportunities and is calculated based on Chung and Pruitt (1994). Cash dividends are defined as total cash dividends paid to common and preferred stockholders. Repurchase activity is measured as in Dittmar (2000) using COMPUSTAT item 115 adjusted for change in preferred stock. The repurchase dummy is equal to one if the firm has repurchase activity of any amount, 0 otherwise. Model 1

Model 2

Model 3

Model 4

(t-statistics)

(t-statistics)

(t-statistics)

(Wald-statistics)

Div./Earnings

Div./Sales

Dividend Yield

Div. Dummy

Intercept

1.024 (0.45)

0.040*** (2.70)

0.009*** (7.88)

-2.867 (0.64)

Governance Index

0.151 (1.64)

-0.001* (-1.76)

0.000 (--0.56)

0.544*** (8.18)

Ln (Total Assets)

-0.133 (-0.78)

0.001 (0.75)

-0.001 (-0.94)

0.697** (6.54)

Leverage

-1.900 (-0.71)

-0.065*** (-3.77)

-0.008 (-0.63)

-14.064*** (14.78)

Profitability

0.948 (0.58)

0.086*** (8.10)

-0.005 (-0.64)

2.632 (2.44)

Growth Opportunities

-0.766 (-0.66)

0.018** (2.41)

-0.034*** (-5.75)

0.368 (0.03)

Repurchase/Earnings

0.451 (2.57)

-

-

-

Repurchase/Sales

-

0.040 (1.29)

-

-

Repurchase/MV of Equity

-

-

-0.041 (-1.32)

-

Repurchase Dummy

-

-

-

1.389 (2.51)

367

367

367

367

2.03*

18.63***

6.96***

-

1.7%

22.4%

8.9%

-

-

-

-

44.3%

Independent Variable

N F-statistics 2

Adjusted R 2

Pseudo R

*, **, *** statistically significant at the 10%, 5%, and 1% levels respectively.

34