Staggered Boards, Managerial Entrenchment, and Dividend Policy

Abstract

Motivated by agency theory, we explore the potential impact of managerial entrenchment through staggered boards on dividend policy. The evidence suggests that firms with staggered boards are more likely to pay dividends. Among firms that pay dividends, those with staggered boards pay larger dividends. We also show that the impact of staggered boards on dividend payouts is substantially stronger (as much as two to three times larger) than the effect of all other corporate governance provisions combined. Overall, the evidence is consistent with the notion that dividends help alleviate agency conflicts. Thus, firms more vulnerable to managerial entrenchment, i.e., firms with staggered boards, rely more on dividends to mitigate agency costs. Aware of potential endogeneity, we demonstrate that staggered boards likely bring about, and are not merely associated with, larger dividend payouts. Our results are important, as they show that certain governance provisions have considerably more influence than others on critical corporate activities such as dividend payout decisions.

1 Introduction

According to agency theory, dividend payouts help alleviate agency costs by reducing the amount of free cash flow available to managers, who 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 mitigate agency conflicts by exposing firms to more frequent monitoring by the primary capital markets because paying dividends makes it more likely that new common stock has to be issued more often.

Staggered or classified boards represent one of the most controversial governance provisions. Staggered boards can insulate inefficient managers from takeover market forces, thereby promoting managerial entrenchment. Two powerful recent studies by Bebchuk and Cohen (2005) and Faleye (2007) find strong evidence indicating that staggered boards allow managerial entrenchment, ultimately resulting in significantly lower firm value.¹

Motivated by agency theory, we explore the effect of potential entrenchment on dividend payouts. As staggered boards can exacerbate agency conflicts and dividend payouts can help alleviate agency costs, we surmise that dividend policy is affected by whether or not the firm has a staggered board. Our findings are consistent with this hypothesis. In particular, the evidence demonstrates that firms with staggered boards are more likely to pay dividends, and firms that pay dividends pay larger dividends than those with unitary boards.

The results are robust even after controlling for a large number of firm-specific characteristics and for an alternate form of payouts, i.e., share repurchases. Furthermore, we examine the reverse causality argument, where dividend policy might lead to the adoption or rescission of a staggered board. Bebchuk et al. (2005) note that very few firms have either adopted or rescinded staggered boards since 1990. As a result, managers likely make dividend decisions while viewing the existence of a staggered board as pre-determined. It is much more likely that causality runs from staggered boards to dividend payouts than vice versa. In any case, we run an analysis based on two-stage least squares (2SLS) estimation and find consistent evidence that staggered boards appear to bring about larger dividend payouts.

Many recent studies find that dividend payouts are inversely related to corporate governance quality, as measured by Gompers et al.’s (2003) Governance Index (Jiraporn and Ning 2006; Pan 2007; John and Knyazeva 2006; Officer 2006; Hu and Kumar 2004). In other words, firms with poor governance quality tend to pay out more dividends. We contribute to this strand of the literature by showing that staggered boards (which constitute one component of the Governance Index) have a palpable impact on dividend policy. In fact, the effect of staggered boards on dividend payouts ranges from two to three times the impact of other governance provisions combined. Our results are remarkably similar to those in Bebchuk and Cohen (2005), who find that the impact of staggered boards on firm value is multiple times the effect of other governance provisions put together.² Our results are vital, as they show that some governance provisions have considerably more influence than others on firm value and on critical corporate activities such as dividend payout decisions.

Furthermore, we investigate the impact of the Sarbanes-Oxley Act (SOX) on the association between staggered boards and dividend payouts. As the Act is intended to hold managers more accountable to shareholders, thereby alleviating agency costs, it may render dividends less necessary as a device for resolving agency costs and therefore may change how dividend payouts are affected by staggered boards. Nevertheless, we do not find empirical evidence consistent with this notion. The impact of staggered boards on dividend payouts exists both before and after the enactment of SOX. The governance reforms introduced by SOX do not appear to materially affect the association between staggered boards and dividend policy.

The results of our study contribute to the literature both in agency theory and in dividend policy. The evidence is in favor of the agency role of dividend payouts and against the signaling hypothesis. We also contribute to the literature in corporate governance by highlighting the role of staggered boards on a critical corporate activity such as dividend policy. Given the recent focus on the impact of staggered boards on firm value and other corporate outcomes (Bebchuk and Cohen 2005; Faleye 2007; Jiraporn and Liu 2008), our results appear to be timely. Finally, our results also contribute to the fierce debate on the costs and benefits of the Sarbanes-Oxley Act.

The rest of this article is organized as follows: Section 2 reviews the literature and develops the hypotheses. Section 3 discusses the sample selection and the relevant data. Section 4 presents and discusses the empirical results. Finally, Section 5 offers the concluding remarks.

2 Background, literature review, and hypothesis development

2.1 The role of staggered boards in entrenching incumbents

In the U.S., boards of directors can be either unitary or staggered. In firms with a unitary board, all directors stand for election each year. In firms with a staggered or classified board, directors are divided into three classes, with one class of directors standing for election at each annual meeting of shareholders. Ordinarily, a classified board has three classes of directors, which in most states of incorporation is the maximum number of classes allowed by state corporate law (Bebchuk and Cohen 2005).

Boards can be removed in one of the following two ways. First, a replacement can occur due to a stand-alone proxy fight brought about by a rival team that attempts to replace the incumbents but continues to run the firm as a stand-alone entity. Second, a board may be replaced as a consequence of a hostile takeover. Either way, the difficulty with which directors can be removed critically depends on whether the firm has a staggered board.

In a stand-alone proxy contest, staggered boards make it considerably more difficult to win control by requiring a rival team to prevail in two elections. In a hostile takeover, staggered boards protect incumbents from removal due to the interaction between incumbents and a board’s power to adopt and maintain a poison pill.³ Before the adoption of the poison pill defense, staggered boards were deemed only a mild defense mechanism, as they did not impede the acquisition of a control block. The acceptance of the poison pill, however, has immensely strengthened the anti-takeover power of staggered boards.

Two powerful recent studies by Bebchuk and Cohen (2005) and Faleye (2007) demonstrate that firms with staggered boards exhibit significantly lower value than those with unitary boards. Thus, the evidence is in accordance with the notion that staggered boards promote managerial entrenchment, exacerbate agency conflicts, and ultimately hurt firm value.

2.2 Prior literature

Existing literature provides evidence consistent with the agency role of dividends in alleviating Jensen’s (1986) free cash flow problem (Easterbrook 1984; Lang and Litzenberger 1989; Smith and Watts 1992; Gaver and Gaver 1993). Agency theory represents a general framework for the role of dividends as a way of reducing the costs of manager-shareholder agency conflict (Easterbrook 1984). Dividends reduce the amount of sub-optimal investment, impose additional monitoring by forcing the manager to address the external financing market, and increase managerial risk-taking (by replacing leverage, dividends lower the expected loss of human capital due to bankruptcy).

Many recent studies document a negative relation between dividend payouts and Gompers et al.’s (2003) Governance Index (Jiraporn and Ning 2006; Pan 2007; John and Knyazeva 2006; and Officer 2006). The Governance Index has a serious weakness in that it assigns equal weights to all the governance provisions included in the construction of the index. Although other governance provisions may also exacerbate managerial entrenchment, there is strong empirical evidence that staggered boards have a far more potent effect than any other governance provision.⁴ Two crucial studies by Bebchuk and Cohen (2005) and Bebchuk and Cohen (2005) show that, even after accounting for the effects of other governance provisions, staggered boards still exhibit a strong negative impact on firm value. In fact, the regression results reveal that the impact of staggered boards on firm value is seven times stronger than the effects of other governance provisions. Bebchuk and Cohen (2005) conclude that “staggered boards play a relatively large role compared to the average role of other provisions included in the GIM Index.”⁵ The effect of staggered boards on firm value is not only statistically significant but also economically significant. Having a staggered board is associated with Tobin’s q that is lower by 17 percentage points (Bebchuk and Cohen 2005).

Additional evidence on the effect of staggered boards is reported in several recent studies. For example, Faleye (2007) reports that staggered boards reduce the probability of forced CEO turnover, are associated with a lower sensitivity of CEO turnover to firm performance, and are correlated with a lower sensitivity of CEO compensation to changes in shareholder wealth. Masulis et al. (2007) demonstrate that announcement period returns are 0.57% to 0.91% lower for bidding firms with staggered boards. They attribute this finding to the self-serving behavior of acquiring firm managers, who themselves are insulated from the market for corporate control.

Jiraporn and Liu (2008) examine how capital structure decisions are influenced by the presence of a staggered board. The evidence reveals that even after controlling for the effects of other governance provisions, firms with staggered boards are significantly less leveraged than those with unitary boards. They argue that staggered boards promote managerial entrenchment, thereby allowing opportunistic managers to eschew the disciplinary mechanisms associated with debt financing. The regression results show that the impact of staggered boards on leverage is six to nine times stronger than the effects of other governance provisions included in Gompers et al.’s (2003) Index.

Furthermore, staggered boards have become a subject of intense investor scrutiny. Institutional Shareholder Services (ISS) recommends in its 2006 proxy voting guidelines that its membership vote against proposals to stagger a board or vote for proposals to repeal staggered board provisions. Additionally, ISS recommends withholding votes for directors who ignore shareholder resolutions to de-stagger a board. ISS also lowers its governance score for firms with staggered boards.⁶ Similarly, CalPERS, the largest public pension fund in the U.S., has targeted firms for shareholder votes to remove staggered boards from their corporate charters. Various mutual fund companies including TIAA-CREF and Fidelity Investments also call for voting against the adoption of and for the removal of staggered board provisions. No other governance provisions have attracted nearly as much controversy from investors as staggered boards, underscoring staggered boards’ dominant role relative to other governance provisions.

Given the above discussion, it is obvious that staggered boards have a serious impact on several critical corporate outcomes, including overall firm value, capital structure, CEO compensation, CEO turnover and takeover gains. It also appears that the effect of staggered boards is large relative to the average effect of other corporate governance provisions. The significance of staggered boards cannot be overemphasized. Consequently, in this study, we narrowly concentrate on the role of staggered boards and investigate their impact on dividend payouts.

2.3 Hypothesis development

Grounded in agency theory, our general hypothesis is that there is a link between staggered boards and dividend payouts, as both are related to agency costs. However, it is unclear what the exact relation should be between staggered boards and dividend policy. On the basis of previous literature in this area, we advance four possible hypotheses.

2.4 The irrelevance hypothesis

This view posits that there is no significant difference in dividend policy between firms with staggered boards and those with unitary boards. Dividends are “sticky.” Once dividends are initiated, managers are extremely unwilling to cut back or terminate dividends (Lintner 1956; Allen and Michaely 2003; Brav et al. 2005), possibly making irrelevant any managerial entrenchment engendered by staggered boards.

2.5 The managerial opportunism hypothesis

This argument is based on the free cash flow hypothesis (Jensen 1986). This view argues that dividend policy is determined by managers who would rather retain cash within the firm for perquisite consumption, for empire building or for investing in projects that enhance their personal prestige but do not necessarily benefit shareholders. As staggered boards can entrench inefficient managers, opportunistic managers may choose to keep more cash within the firm and pay less out as dividends. The empirical prediction of this hypothesis is that firms with staggered boards should pay less dividends than those with unitary boards.

2.6 The agency cost alleviation hypothesis

Payout policy is one mechanism for alleviating the manager-shareholder conflict. However, the efficacy of payout policy in reducing agency costs hinges largely on the degree of restriction on managerial actions. Without pre-commitment, poorly-monitored managers can ex post deviate from the payout policy and use free cash flow to finance inefficient investment. Given the negative market reaction to dividend cuts and infrequent deviations from dividend policy, dividends help constrain the manager through the high cost of deviation and constitute an effective pre-commitment mechanism in the presence of a severe agency conflict (John and Knyazeva 2006).

As shareholders observe that firms with staggered boards may be more prone to managerial entrenchment and rationally anticipate the larger extent of the free cash flow problem, the necessity for dividends should be stronger for firms with staggered boards than for firms with unitary boards.

Dividend payment imposes a tax cost on the payer firm. Moreover, dividend paying firms also incur the cost of forgone positive-NPV projects or the additional cost of raising external financing to fund them when internal cash flow is inadequate. Since dividends are costly, firms that are less vulnerable to managerial entrenchment (i.e., firms without staggered boards) should be less inclined to pay dividends and should pay lower dividends on average.

2.7 The signaling hypothesis

This hypothesis is based on an argument made by La Porta et al. (2000). This view relies critically on the assumption that firms need 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 from shareholders. One way to establish such a reputation is by paying out dividends, which reduces what is left for expropriation.⁷ A reputation for good treatment of shareholders is worth more in firms where opportunistic managers are more likely to be entrenched, i.e., in firms with staggered boards. As a result, the need to establish a reputation is greater for such firms. By contrast, for firms with unitary boards, the need for a reputation mechanism is less necessary, and thus, so is the need to pay dividends. This view, therefore, posits that dividend payouts should be higher in firms with staggered boards.

Because firms with plenty of growth opportunities need more financing and are thus more likely to raise capital in the external markets, the need for signaling should be stronger for these firms. As a result, one crucial implication of this hypothesis is that firms with staggered boards that exhibit stronger growth opportunities should pay more dividends than those that show weaker growth (Pan 2007).

3 Sample formation and data description

3.1 Sample selection

The original sample is compiled from the Investor Responsibility Research Center (IRRC). The IRRC reports data on corporate governance provisions for about 1,500 firms. The sample firms, mainly drawn from the S&P 500 and other large corporations, represent over 90% of total market capitalization on NYSE, AMEX, and NASDAQ. The IRRC collects data on 24 corporate governance provisions, one of which is staggered boards.⁸ The sample is narrowed down further by dropping firms whose financial data do not exist in COMPUSTAT. Financial firms are excluded due to their unique accounting and financial characteristics.

The final sample consists of 9,918 firm-year observations from 1990 to 2004. This sample is the largest and most recent among most studies in this area. The year distribution of the sample is displayed in Table 1. It appears that about 60% of firms in the sample have staggered boards. This proportion is remarkably constant over time, in a narrow range from 59.07% to 63.25%.

Table 1 Sample distribution by year and by board type. The original sample is complied from the Investor Responsibility Research Center (IRRC)

Because the IRRC data include only large firms, it could be argued that our studies are biased towards firms of large size. The IRRC data covered, in 1990, over 93% of the market capitalization of the combined NYSE, AMEX, and NASDAQ markets (Gompers et al. 2003). Like Pan (2007), we propose that this possible large-firm bias should not constitute a serious concern for our study. Given the purpose of this study, we need a sample of firms that are potential candidates to pay dividends and attempt to understand whether their dividend payouts are related to their level of managerial entrenchment. As dividends are paid mainly by large and mature firms, the IRRC sample firms are those that should be most suitable for this study. Additionally, firms can only pay dividends when they are able to generate stable earnings, i.e., when they become mature firms. On the contrary, most fast growing young firms cannot or choose not to pay dividends. Consequently, a study like ours that examines dividend policy and managerial entrenchment probably ought to include large firms in the sample, which is precisely what we do here.

3.2 Descriptive statistics

Table 2 Panel A shows the summary statistics for the sample. Sales and total assets average $3,735 and $3,937 million respectively. Leverage, measured as the ratio of total debt to total assets, averages 24.12%. Profitability, computed as EBITDA divided by total assets, averages 16.19%. We measure growth opportunities by calculating the ratio of capital expenditures to total assets. The average capital expenditures ratio is 6.21%. The average income tax rate is calculated as the ratio of total income taxes to EBIT. The average firm has a tax rate of 29.32%. Finally, the average Governance Index is 9.28.

Table 2 Descriptive statistics. EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers, Ishii, and Metrick (2003)

We also contrast firms with staggered boards with those with unitary boards. The results show that firms with staggered boards are smaller in terms of total assets, are more leveraged, have more growth opportunities and have a higher governance index.

Table 2 Panel B exhibits the descriptive statistics for the dividend payout measures. We investigate three alternative ratios of dividends: dividends divided by total assets, dividends divided by sales, and dividends divided by net income. The averages of the three ratios are shown in the table. When we compare the three ratios between firms with staggered boards and those without, we find that all the dividend ratios are significantly higher for firms with staggered boards than those with unitary boards. Further, the percentage of firms that pay dividends (of any size) is 73.71% for firms with staggered boards and is only 64% for firms with unitary boards. Thus, the preliminary statistics seem to show that firms with staggered boards exhibit a stronger tendency to pay dividends and those that pay dividends pay larger dividends than firms with unitary boards.

Finally, Table 2 Panel C displays the correlation coefficients for staggered boards, the three dividend ratio and the dividend dummy.⁹ Note that all the correlation coefficients are positive and statistically significant at the 1% level. Again, the correlation analysis indicates that firms with staggered boards pay dividends more generously than do firms with unitary boards. Thus, all the descriptive statistics are in agreement with both the signaling hypothesis and the agency cost alleviation hypothesis, which predict a positive relation between staggered boards and dividend payouts. These results should be viewed with caution, nonetheless, because they may merely reflect differences in firm characteristics. As a result, we perform a multivariate regression analysis in the following section.

4 Empirical results

4.1 Staggered boards and dividend payouts

As dividend payouts cannot be below zero, we employ Tobit regressions. A number of control variables are included. We employ the natural logarithm of total assets to proxy for firm size. 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 total debt to total assets.

Profitability has been found to affect dividend policy (DeAngelo and DeAngelo 1990; DeAngelo et al. 1992). Thus, we control for profitability by using the ratio of EBITDA to total assets. Growth has been reported to impact dividend policy as well (Rozeff 1982). We control for firm growth by using capital expenditures scaled down by total assets. Furthermore, with the growing importance and popularity of share repurchases, we include several share repurchase ratios to control for this alternative means of cash distribution.¹⁰ Share repurchases could affect dividend payouts because cash spent on repurchases could otherwise be distributed as dividends (Grullon and Michaely 2002). Fama and French (2001) report that the proportion of firms paying dividends declines substantially over time. As a result, we include year dummies to capture temporal variation in dividend payouts. To account for the tax efficiency of dividends, we also include the ratio of income taxes to EBIT as a control variable (John and Knyazeva 2006).¹¹ Finally, we control for other governance provisions by constructing an index of all other 23 management-favoring provisions reported by the IRRC.¹²

Table 3 shows the results of the regression analysis. In Model 1, we use dividend payouts divided by total assets as the dependent variable. The coefficient of staggered boards is positive and statistically significant at the 1% level. To confirm the result, we use alternative measures of dividend payouts in Model 2 and Model 3, i.e., the ratios of dividend payouts to sales and dividend payouts to net income respectively. Staggered boards exhibit positive and strongly significant coefficients in both Model 2 and Model 3 as well. The results in all three models are consistent and indicate that firms with staggered boards pay larger dividends than do firms with unitary boards.

Table 3 Tobit and logit regression analyses of dividend payouts. EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers et al. (2003). The Index of Other Governance Provisions is from Bebchuk and Cohen (2005). Repurchase activity is measured as described in Dittmar (2000). The Dividend Dummy is set to zero if the firm does not pay dividends and one if it pays a dividend (of any size)

It is worth noting that staggered boards are correlated with the index of other governance provisions (which enters the regressions as an independent variable). Because the correlation between staggered boards and the index of other governance provisions introduces a problem of colinearity, our results may be biased against finding significance for either one of these variables. Thus, even if the inclusion of the index of other governance provisions were to make the coefficient of staggered boards insignificant, it would not eliminate the possibility that staggered boards are correlated with our measures of dividend payouts. We do not have to address this problem, however, because the coefficient of staggered boards is positive and statistically significant despite the stacking of the deck against such a finding.

In addition, it should also be highlighted that the magnitude of the coefficient of staggered boards is large relative to the coefficient of the index of other governance provisions. In Models 1 and 3 (Table 3), the coefficient of staggered boards is twice as large as the coefficient of the index of other governance provisions, whereas in Model 2, the coefficient of staggered boards is three times the coefficient of the index of other governance provisions. This result implies that staggered boards play a relatively large role compared to the average role of other provisions in the Governance Index. Our results remarkably mirror the results in Bebchuk and Cohen (2005), who document that the impact of staggered boards on firm value is considerably larger than the impact of other corporate governance provisions combined.

We also investigate how staggered boards affect the propensity to pay dividends (of any size). Pan (2007) argues that it may be more useful to examine the propensity to pay dividends rather than the dividend payout ratio. An analysis of the payout ratio requires a model to estimate the optimal payout ratio for a particular firm. It does not appear that such a model has been developed in the literature. As a result, examining the propensity to pay dividends may offer cleaner results, not tainted by a bad model problem. Further, the fact that dividend payouts are sticky (Lintner 1956; Brav et al. 2005) makes the decision to pay dividends more crucial than the decision to change the amount of payouts.

In Model 4, we run a logistic regression where the dependent variable is a dichotomous variable equal to one if the company pays dividends and zero if the company does not pay dividends. The coefficient of staggered boards is positive and highly significant, showing that firms with staggered boards exhibit a stronger propensity to pay dividends than those with unitary boards.

Overall, the empirical results based on the tobit and logit analyses seem to be consistent with both the signaling hypothesis and the agency cost alleviation hypothesis, as both hypotheses predict a positive association between staggered boards and dividend payouts.

4.2 Possible simultaneity

The association between staggered boards and dividend payouts identified earlier raises the question of simultaneity. Do staggered boards cause firms to pay higher dividends? Or are firms that pay higher dividends more likely to adopt a staggered board? It is more plausible that the direction of causality runs from staggered boards to dividend payouts rather than vice versa. There are several reasons why this should be the case. First, theoretically, it is unclear why firms that pay higher dividends would be motivated to adopt a staggered board. In fact, agency theory suggests that firms that pay higher dividends should incur lower agency costs-either due to having lower free cash flows (Jensen 1986) or to being exposed to external market monitoring more frequently (Easterbrook 1984)-and would thus be less likely to adopt a governance provision that could entrench inefficient managers.

Second, the adoption of a staggered board is not entirely under managers’ control because it usually requires shareholders’ approval, which takes a great deal of effort and a lengthy period of execution. Dividend payouts, on the other hand, are much more subject to managerial discretion. It is hence much more plausible that managers make decisions on dividend policy while taking board structure as pre-determined. Third, due to the difficulty of adopting or rescinding staggered boards, board structure is very stable. With very few exceptions, the sample firms did not change their board structure over our sample period, i.e., firms that had staggered boards in the beginning of the sample period continue to have staggered boards throughout the sample period. Firms that did not have staggered boards in the beginning of the sample period never adopt staggered boards later. Thus, it is extremely unlikely for causality to run from dividend payouts to staggered boards.¹³

In any event, to further assure that endogeneity is adequately addressed, we execute a two-stage least squares (2SLS) estimation. This approach explicitly takes into account possible endogeneity. This technique requires an instrumental variable that is exogenous. We consult the literature and follow the method employed by John and Knyazeva (2006). Examining the impact of corporate governance on dividend payouts, John and Knyazeva (2006) use the industry medians of their governance variables as instrumental variables. The logic is that industry structure is unique to each industry and therefore is expected to be exogenous. As a result, we use the percentage of firms in each industry in each year that have staggered boards as our instrumental variable.

The 2SLS results are shown in Table 4. The coefficient of predicted staggered boards is positive and statistically significant in all four of the regressions. Thus, even when we employ the 2SLS technique, which explicitly takes into consideration possible endogeneity, we still obtain consistent results. Firms with staggered boards pay higher dividends.

Table 4 2SLS Estimation of staggered boards and dividend payouts. EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers, Ishii, and Metrick (2003). The Index of Other Governance Provisions is from Bebchuk and Cohen (2005). Repurchase activity is measured as described in Dittmar (2000). The Dividend Dummy is set to zero if the firm does not pay dividends and one if it pays a dividend (of any size)

4.3 Distinguishing the hypotheses using firm growth

A positive association between staggered boards and dividend payouts is consistent with both the signaling hypothesis and the agency cost alleviation hypothesis. To distinguish between these two hypotheses, we take advantage of a crucial implication of the signaling hypothesis. The signaling hypothesis suggests that firms with stronger growth are more likely to raise capital from the external markets. Consequently, using dividends as a signaling mechanism is more necessary for these firms. Hence, the signaling hypothesis implies that dividend payouts should be higher for firms with a staggered board that exhibit stronger growth (Pan 2007).

We measure percentage growth based on total assets over the previous 3 year period. Then, we interact this growth variable with staggered boards and include the interaction term in the regression analysis. If the signaling hypothesis is correct, the coefficient of the interaction term is expected to be positive.

Table 5 shows the results. We use three alternative measures of dividend payouts in Models 1–3, as we did previously. Note that the coefficient of the interaction term is not positive in any of the three models, contradictory to the prediction of the signaling hypothesis. In Model 4, we explore the propensity to pay dividends by running a logistic regression. The interaction term in Model 4 does not show a significant coefficient either.¹⁴ We check the robustness of the results by using an alternative measure of growth based on operating profits. The results remain similar. There is thus no empirical support for the signaling hypothesis. It appears that the evidence is more in favor of the agency cost alleviation hypothesis.

Table 5 Tobit regression analysis of dividend payouts with interactions between staggered boards and growth. EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers, Ishii, and Metrick (2003). The Index of Other Governance Provisions is from Bebchuk and Cohen (2005). Repurchase activity is measured as described in Dittmar (2000). The Dividend Dummy is set to zero if the firm does not pay dividends and one if it pays a dividend (of any size). Firm growth is computed as change in total assets in the previous 3-year period

It is also worth noting that staggered boards retain significant coefficients even when the interaction terms are included in Table 5. The coefficients of firm growth are negative and significant in all of the regressions as well, indicating that high-growth firms tend to pay lower dividends. Thus, the results are consistent with prior literature and make intuitive sense.

4.4 Repurchases vs. dividends

Share repurchases are more discretionary cash distributions than dividends. Ex post deviations from the previously announced repurchase policy are more typical and less detrimental to the stock price than dividend cuts (John and Knyazeva 2006). The managerial opportunism hypothesis would predict that entrenched managers attempt to escape the disciplinary role of dividends and exploit the flexibility associated with repurchases. Thus, this view argues that firms with staggered boards should be more inclined to use repurchases than dividends.

By contrast, the agency cost alleviation hypothesis implies that firms with staggered boards rely more on dividends rather than repurchases. The benefit of dividends in resolving agency costs is low when the firm is less vulnerable to managerial entrenchment (i.e., when the firm does not have a staggered board). Hence, staggered boards should be associated with an inclination to use dividends.

We investigate the relative roles of dividends and repurchases in firms with and without staggered boards. It is not advisable to use the share of repurchases as the dependent variable in a regression analysis because it potentially overstates the role of repurchases in the presence of mixed dividend and repurchase policies, since repurchases tend to be larger than dividends (John and Knyazeva 2006). We select to run a logistic regression analysis on firms that pay only dividends, firms that use only repurchases, and firms that both pay dividends and use repurchases.

The regression results are shown in Table 6. In Model 1, the probability of using dividends versus only repurchases is higher in firms with staggered boards. In Model 2, we compare firms that use a mix of dividends and repurchases versus repurchases only. The evidence shows that firms with staggered boards are more likely to use a combination of dividends and repurchases rather than repurchases only. Finally, in Model 3, we compare firms that use only dividends and those that use only repurchases. The coefficient of staggered boards is not statistically significant.

Table 6 Staggered boards and payout choices (dividends vs. repurchases). EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers, Ishii, and Metrick (2003). The Index of Other Governance Provisions is from Bebchuk and Cohen (2005). Repurchase activity is measured as described in Dittmar (2000)

Taken together, the evidence demonstrates that the use of pre-commitment to dividends as the only component or as part of payout policy is more likely when the firm is more subject to managerial entrenchment (i.e., when the firm has a staggered board). This is in agreement with the prediction of the agency cost alleviation hypothesis.¹⁵

4.5 Possible impact of the Sarbanes-Oxley Act

The Sarbanes-Oxley Act (SOX) was enacted on July 30, 2002 as a consequence of Congressional hearings conducted after the first admissions of fraudulent behavior made by Enron. President George W. Bush characterized this Act as one of “the most far-reaching reforms of American business practices since the time of Franklin D. Roosevelt.”¹⁶ The Act introduces new provisions for management, directors, auditors and analysts, and significantly raises criminal penalties for securities fraud and for destroying, altering or fabricating records in federal investigations or any scheme or attempt to defraud shareholders. Evidently, this Act is intended to hold managers more accountable to shareholders. The increased accountability should bring manager and shareholder interests in better alignment, thereby alleviating agency costs.¹⁷ Therefore, the role of dividends in resolving agency costs may have been altered after the passage of SOX.

To ascertain the impact of SOX on the association between staggered boards and dividend payouts, we construct a dichotomous variable equal to one for the post-SOX period and to zero for the pre-SOX period. Because SOX was enacted in 2002, we use 2002 as the dividing point. Then we interact this dichotomous variable with staggered boards. The coefficient of this interaction term should reveal the relative impact of staggered boards on dividend payouts before and after the passage of SOX.

The regression results are reported in Table 7. The coefficient of the interaction variable is not significant in any of the regressions.¹⁸ It appears that the effect of staggered boards on dividend policy did not materially change after SOX was enacted. Finally, post-SOX carries negative and significant coefficients in all of the regressions, suggesting that firms pay lower dividends after the enactment of SOX. A possible interpretation is that the role of dividends in alleviating agency costs is made less necessary because the governance reforms mandated by SOX already mitigate the agency conflict. This result seems to be consistent with the prediction of agency theory.

Table 7 Tobit regression analysis of dividend payouts (before and after enactment of the Sarbanes-Oxley Act). EBITDA is earnings before interest, taxes, depreciation, and amortization. EBIT is earnings before interest and taxes. Leverage is total debt divided by total assets. The Governance Index is from Gompers, Ishii, and Metrick (2003). The Index of Other Governance Provisions is from Bebchuk and Cohen (2005). Repurchase activity is measured as described in Dittmar (2000). The Dividend Dummy is set to zero if the firm does not pay dividends and one if it pays a dividend (of any size)

5 Concluding remarks

Dividends play a role in reducing agency costs (Easterbrook 1984; Jensen 1986), while staggered boards promote managerial entrenchment and exacerbate agency conflicts (Bebchuk and Cohen 2005; Faleye 2007). Grounded in agency theory, this study examines the impact of staggered boards on dividend policy. We document a number of interesting findings. First, we find that firms with staggered boards exhibit a stronger likelihood to pay dividends than firms with unitary boards. Moreover, firms with staggered boards that do pay dividends pay larger dividends. This evidence is in accordance with the notion that the value of dividends as a mechanism for resolving agency conflicts is higher in firms that are more vulnerable to managerial entrenchment (i.e., firms with staggered boards). As a result, these firms are more likely to pay dividends and pay larger dividends. We also show that the impact of staggered boards on dividend payouts is larger, by a magnitude of two to three times, than the effect of all other governance provisions combined. Our results are similar to Bebchuk and Cohen (2005), who document a considerably larger impact of staggered boards on firm value than of all other governance provisions.

Second, we show that staggered boards are not merely associated with larger dividends. There is evidence that staggered boards bring about larger dividends. The direction of causality more likely exists from staggered boards to dividends than vice versa. Finally, we compare the probability of paying dividends relative to the probability of using repurchases and discover that firms with staggered boards are more likely to pay dividends than to use repurchases, consistent with the stronger pre-commitment role of dividends. The results of this study contribute to the literature in agency theory, dividend policy and corporate governance and add interesting empirical evidence to the on-going debate on the benefits and detriments of staggered boards.