Docstoc

On the Uses of Corporate Governance Provisions

Document Sample
On the Uses of Corporate Governance Provisions Powered By Docstoc
					On the Uses of Corporate Governance Provisions

Morris G. Danielson Temple University School of Business Philadelphia, PA USA (215) 204-8453 email: morrisd@surfer.sbm.temple.edu Jonathan M. Karpoff University of Washington School of Business Seattle, WA 98195 USA (206) 685-4954 email: karpoff@u.washington.edu

March 3, 1997 Revised: April 22, 1997 Revised: September 3, 1997 Revised January 7, 1998

-2-

We thank Ted Klastorin, Ken Lehn, Paul Malatesta, Megan Partch, an anonymous referee, and participants at the finance seminar at the University of Notre Dame for helpful comments and suggestions, and Wayne Marr for providing access to the data from Institutional Shareholders Services, Inc.

On the Uses of Corporate Governance Provisions Abstract

We document a large and broad-based increase in the use of corporate governance provisions in the late 1980's. As a result, most large publicly traded firms have complex governance structures. This violates an assumption implicit in many empirical studies that provision use is mutually independent. While overall provision use is not systematically related to industry grouping, the uses of some types of provisions are correlated. Most notably, supermajority vote requirements, classified boards, and shareholder meeting requirements tend to be used in concert. Firms reincorporating to Delaware tend to eliminate cumulative voting, and coverages by certain types of state antitakeover laws are correlated. We also find that firms with poison pills tend to have relatively high institutional ownership and low managerial ownership, but a high proportion of independent directors.

-3-

On the Uses of Corporate Governance Provisions 1. Introduction Corporate managers can propose or adopt a wide variety of provisions that affect their working relationship with the firm's directors, the directors' relationship to stockholders, and the firm's vulnerability to takeover. Examples include poison pills, blank check preferred stock, supermajority vote requirements, dual-class voting rights, fair price charter amendments, and coverage by state antitakeover laws. Numerous studies examine the effects of various types of provisions on stockholder wealth and managerial decisions.1 Our objective is more modest. We examine how frequently, and in what combinations, various types of corporate governance provisions are used, and whether provision use is related to ownership structure and board composition. To do so, we employ data complied through records maintained by Institutional Shareholders Services, Inc. on the corporate governance features of large, primarily Standard and Poor's 500, firms. The database contains information on the use of 20 different provisions, including 15 firm-level provisions and coverage by five types of state antitakeover laws. The provisions are described in Table 1. Our data show that director/officer liability indemnity provisions, blank check preferred stock, coverage by state freeze-out laws, poison pills, and classified boards are deployed by a majority of firms. Fair price charter amendments and shareholder meeting requirements also are widespread, covering approximately 40 percent of our sample firms. Several provisions, including unequal voting rights and charter language that directs directors to consider the interests of non-shareholder groups, are relatively infrequent, covering 10 percent or fewer of the firms. We document several patterns in the uses of corporate governance provisions. Some provisions -- for example, supermajority vote requirements, classified boards, and shareholder meeting requirements -- tend to be used in concert. Provision use is not systematically related to firms' industry groupings, as measured by Compustat SIC codes. Provision use is sensitive
1For

examples, see DeAngelo and Rice (1983), Linn and McConnell (1983), Jarrell and Poulsen (1987), Malatesta and Walkling (1988), Ryngaert (1988), Agrawal and Mandelker (1990), McWilliams (1990), Bhagat and Jefferis (1991), Brook and Rao (1994), Brickley, Coles, and Terry (1994), Karpoff and Malatesta (1995), Kini, Kracaw, and Mian (1995), Comment and Schwert (1995), Hirschey and Jones (1995), and McWilliams and Sen (1998).

-4-

to the legal environment, however, as several provisions became widespread only after important court decisions removed ambiguity about their legal status. The data also support a widely-held view that the use of corporate governance provisions increased markedly during the latter 1980's, immediately before the decline in takeover activity in the early 1990's. These data provide evidence about the presumptions used to justify specific research designs involving corporate governance. Most of the studies cited in footnote 1 examine single provisions in isolation, based on the implicit assumption that the uses of governance provisions are mutually independent. Our findings indicate that this assumption is justified in some applications (e.g., poison pills), but not in others (e.g., most antitakeover charter amendments). As another example, researchers sometimes base their research designs on assertions about the prevalence of certain corporate governance provisions (e.g., see and Garvey and Hanka (1996)). Bhagat and Jefferis (1996), for example, assert that they can ignore state takeover laws before 1988 in certain empirical tests because such laws were rare until 1988. Our data reveal whether such claims are accurate. The paper is organized as follows. Section 2 describes the 20 provisions included in this study. Section 3 reports on the frequencies of the provisions in our sample and the changes in provision use over the 1984 - 1989 period. Section 4 examines firm and industry patterns in provision use, and reports on factor analyses that identify clustering in provision use. In section 5 we discuss the implications of our findings for the design of empirical tests involving corporate governance provisions. Section 6 concludes. 2. Description of the corporate governance provisions Table 1 describes the 20 corporate governance provisions identified in our database. For expositional purposes we group the provisions into four categories. The primary criterion used to classify the provisions is Manne’s (1965) distinction between the external and internal markets for corporate control. The external market for corporate control refers to such actions as tender offers, taken by outside bidders to gain control through the accumulation of a large block of shares. The internal market for corporate control refers to such actions as proxy challenges, taken by current shareholders to

-5-

influence corporate policy by exercising voting rights. Additional categories are used for state antitakeover laws and several miscellaneous provisions.2 1. External control provisions are those that directly impede or may be used to impede hostile acquisitions, and which therefore work primarily through their effects on the external market for control. The provisions in this category are poison pills, blank check preferred stock, and stakeholder charter clauses. Previous studies report that the average stock price change upon the announcement of these provisions is negative.3 2. Internal control provisions are those that increase a large shareholder's cost of exercising control or influencing corporate policies, and which therefore directly affect the internal market for control. This category consists of classified boards, fair price provisions, recent (within three years) reincorporation to Delaware, shareholder meeting requirements, unequal voting rights, supermajority voting requirements, and recent (within three years) elimination of cumulative voting. Unlike those in category 1, these provisions do not directly affect a bidder's cost of accumulating a large block of stock. By imposing costs on large shareholders, however, these mechanisms also can impede external bids for control. The evidence on the valuation effects of these provisions is mixed. DeAngelo and Rice (1983), for example, find that announcements of supermajority voting requirements and classified boards are associated with negative but statistically insignificant abnormal returns. Linn and McConnell (1983) find that abnormal returns upon announcements of classified boards, shareholder meeting requirements, supermajority voting requirements, and fair price provisions are, in general, positive but insignificant. Jarrell and Poulsen (1987) report that fair price provisions are associated with statistically insignificant abnormal returns, but that such other provisions as supermajority voting requirements and classified boards are associated with statistically significant negative abnormal returns.
defensive governance provisions impede unsolicited bids for control, and the effects of any one provision are firm-specific and difficult to unravel. Thus, some readers may dispute our choice of categories and placement of specific provisions within categories. The grouping was not done haphazardly, however. In constructing the categories we have conferred with ISS advisors. ISS is in the business of selling information on governance provisions to institutional investors. Therefore, the principals at ISS have a direct incentive to be well-informed about the effects of the various governance provisions on the market for corporate control. 3For poison pills, see Malatesta and Walkling (1988), Ryngaert (1988), and Brickley, Coles, and Terry (1994); for blank check preferred stock, see Jarrell and Poulsen (1987). We are unaware of an empirical examination of stakeholder charter provisions. However, Alexander, Marr, and Spivey (1993) report that state laws that resemble the language of such provisions are associated with negative and statistically significant stock returns for affected companies.
2All

-6-

Reincorporations to Delaware and recent eliminations of cumulative voting are in fact actions, not provisions. Nonetheless, we treat these items as provisions because ISS does so. We repeated the tests reported below while omitting data on these two provisions. None of the results are qualitatively affected, except for those pertaining particularly to those two provisions. 3. State Takeover Laws included in this category are control share acquisition laws, fair price laws, freeze-out laws, cash-out laws, and poison pill laws. By 1992, a total of 110 of these laws had been adopted by a total of 39 states, covering most of the large corporations in the United States. The laws impose restrictions on both internal and external control mechanisms. For example, poison pill laws provide statutory sanction for poison pills, thereby strengthening the legal position of firms adopting poison pills. Control share acquisition and freeze-out laws do not directly impede the accumulation of a large block of stock, but each can be used to prevent a large outside shareholder from exercising control over corporate resources. Karpoff and Malatesta (1989) report that state takeover laws are associated with small but statistically significant negative abnormal stock returns for companies covered by the laws. Unlike firm-level provisions, takeover laws are imposed by state legislatures and not by directors or stockholder vote. Most laws, however, result from direct lobbying by managers of firms incorporated in the state. In addition, many antitakeover laws allow firms to opt out of coverage. Thus, individual firms have broad discretion over whether to be covered by a law. 4. Other miscellaneous provisions constitute the fourth category, and include antigreenmail, confidential voting, and cumulative voting provisions, provisions that indemnify company directors and officers from personal liability for corporate actions, and an indicator variable for firms that opted out of coverage by a state antitakeover law. The provisions in the first three categories decrease managers’ vulnerability to unsolicited takeover bids. Provisions in this miscellaneous category, however, could increase such vulnerability. For example, ISS classifies cumulative voting, confidential voting, and decisions to opt out of state antitakeover laws as favorable for shareholders. Anti-greenmail and director/officer indemnity provisions have ambiguous effects on a firm's vulnerability to outsider takeover. Prohibiting greenmail, for example, prevents managers from paying a premium in a targeted

-7-

share repurchase to fend off an unwanted takeover bid. But, as Eckbo (1990) observes, it might also deter the unsolicited bid in the first place. Because the provisions in this miscellaneous category appear to differ substantially from the other provisions, we repeat all the descriptive tests reported below while omitting the miscellaneous provisions. None of our primary conclusions, however, are qualitatively affected. 3. Frequencies of the provisions in our sample 3.a. Increased use from 1984 to 1989 Table 2 reports the frequencies of each governance provision in the sample for each year, 1984-89. Data on the provisions in panels A, B, and D are obtained from records made available to us by Institutional Shareholder Services, Inc. The ISS sample consists of 513 firms, including most firms in the S&P 500 Index. The ISS data end in 1989, but crossreferencing many firms' data with more recent data available in Rosenbaum (1995) suggests that the rate of change in firms' governance structures decreased markedly after 1989. The state law data (Panel C) are obtained from Karpoff and Malatesta (1989), updated by data reported in McGurn, Pamepinto, and Spector (1992). In 1984, there was an average 2.07 provisions per firm. By 1989, the average number of provisions per firm increased to 5.93. These data document a widespread assertion (e.g., see Black (1992); Gordon and Pound (1993)) that provision use increased in the late 1980's. Comment and Schwert (1995) claim that the proliferation of defensive governance provisions had little to do with the subsequent decline in takeover activity in the early 1990's. Coffee (1991), Pound (1992), and others argue the opposite, claiming that takeover frequency declined because firm-level and state antitakeover provisions increased the cost to potential bidders of making unsolicited takeover bids. These data do not address this controversy directly, but they indicate that a significant increase in the use of governance provisions did indeed precede the early 1990's decline in takeover activity. Classified boards, fair price charter provisions, shareholder meeting requirements, and supermajority vote requirements all were adopted widely by 1989. There also were large increases in the use of poison pills and director/officer liability indemnity provisions such that, by 1989, each was in place at more than half the sample firms.

-8-

Blank check preferred stock is the only provision widely adopted at the start of our sample period in 1984. The number of firms with this provision increased by 1989. But most such firms also implemented a poison pill, so the number of firms with stand-alone blank check preferred stock declined. Among state laws, freeze-out laws affect by far the greatest number of firms. This is because of the large number of S&P 500 firms incorporated in Delaware, which adopted a freeze-out law in 1988. In contrast, the relative importance of fair price and cash-out laws has declined steadily since 1985. Once among the most prominent of the second-generation type state antitakeover laws, fair price and cash-out laws covered a smaller number of firms by 1989 than freeze-out, poison pill, and control share acquisition antitakeover laws.4 By 1989, stakeholder charter clauses, unequal voting rights, antigreenmail provisions, confidential voting rights, and cumulative voting rights all were used by relatively small numbers of firms. Fewer still had opted out of state laws, or had recently reincorporated to Delaware or eliminated cumulative stock or shareholders’ right to alter the board size. 3.b. The impact of court rulings on state laws and provision use Court rulings correspond to the widespread adoption of several provisions. In 1985 the Delaware Supreme Court upheld the use of a poison pill takeover defense in Moran v. Household International (500 A. 2d 1346, Del. 1985). Ryngaert (1988) reports that the number of firms adopting poison pills increased substantially in the following year. In our sample, the number of firms adopting poison pills increased substantially over the next several years, from 20 in 1985 to 140 in 1986, and to 307 by 1989. In another important 1985 decision, the Delaware Supreme Court greatly expanded the types of corporate decisions for which officers and directors could be held personally liable (Smith v. Van Gorkom, 488 A. 2d 858, Del. 1985). This created what Rao and Brook (1995) and others call a director liability crisis. Our data indicate that many corporations responded by adopting provisions that indemnify directors and officers against claims for actions they take in their roles as officers and directors. The number of firms with such provisions jumped from 20 in 1985 to 349 in 1987 and 441 in 1989.
4Regarding

the Bhagat and Jefferis (1996) assertion cited in the introduction: Table 2 shows that state takeover laws were widespread before 1988. In 1987, 212 of our sample firms (41%) were covered by at least one state takeover law.

-9-

The increase in the number of firms covered by state antitakeover laws corresponds to a 1987 U.S. Supreme Court decision that upheld Indiana’s control share acquisition law (CTS v. Dynamics Corp. of America, 107 S. Ct. 1673, 1987). The decision applied only to Indiana's law. But it created the presumption that other takeover laws also were valid, a presumption that eventually received support in 1989 when the Seventh U.S. Court of Appeals upheld Wisconsin's freeze-out law (Amanda Acquisition Corporation v. Universal Foods Corporation, 877 F.2d 496; 1989 U.S. App.) In the year immediately following the CTS decision at least 15 states, including Delaware, adopted over 20 new antitakeover laws. The number of firms in our sample covered by poison pill state laws increased from zero in 1985 to 156 in 1989. These laws grant firms statutory authority to adopt poison pill takeover defenses, thus decreasing any ambiguity over the pills' legality. One might expect that firms covered by poison pill laws would be most likely to have poison pills. We find statistically significance evidence of such tendency, however, only in 1987. In 1987, 51% of the 35 firms covered by poison pill state laws had poison pills, compared to 34% of firms not so covered. The Chi-squared statistic testing equality of percentages has a p-value of .04. In all other years, however, poison pill use was roughly equally widespread among firms covered and not covered by poison pill state laws.5 4. Patterns of adoption 4.a. Distribution of the number of provisions per firm There is no standard or boilerplate governance structure among the firms in our sample. In 1989, for example, there are 318 different combinations of the 20 governance provisions among the 513 firms in the sample. Nonetheless, we can make several observations about how the provisions are used. Table 3 presents information on the distribution of the number of provisions per firm in each year of our sample. The distributions for 1984 - 1986 exhibit positive skewness, whereas the 1987 - 1989 distributions are roughly symmetrical. The median number of provisions per firm increases from two in 1984 to six in 1989. By 1989, only a single firm (Placer Dome, Inc.) has no governance provisions, and only two firms (Playboy Enterprises
most firms, poison pills can be adopted quickly whenever the managers want. This might explain why poison pills are not more prevalent in states that have poison pill laws -- managers of firms in those states can acquire pills on short notice.
5In

- 10 -

and Seagram Ltd.) have one provision. These data indicate that the secular increase in provision use does not reflect large increases in provision use by a minority of firms, but rather, a broad tendency for most firms to acquire more provisions over time. Some firms nonetheless use more provisions than others. Table 4 reports 1989 Spearman rank-order correlation coefficients for the number of provisions by firm in each of our four provision categories. The 1989 correlations are similar to those from 1984 - 1988 as well. In 1989 (as in the other years), the number of external and internal control provisions per firm is positively and significantly correlated. The correlation coefficient is .282 with a p-value of .001. Thus, there is a tendency for firms using relatively large numbers of external control provisions (poison pills, blank check preferred stock, and stakeholder provisions) also to employ relatively large numbers of internal control provisions. The number of internal control provisions, in turn, is negatively correlated with the number of state laws offering takeover protection. The correlation coefficient is -.113 with a p-value of .010. The negative correlation between state law coverage and the use of internal control provisions suggests that managers consider the two types of provisions to be substitutes. Upon inspection, some state antitakeover laws impose restrictions that are similar to some internal control provisions. Fair price laws, for example, closely mimic the language used in most fair price charter amendments. Control share acquisition laws restrict the voting rights of large acquiring shareholders, imposing restrictions similar to those imposed by unequal voting rights and supermajority vote provisions. Even freeze-out laws can be thought of as combinations of two internal control provisions. The typical freeze-out law imposes requirements similar to fair price charter amendments, only with a forced delay. (In the case of Delaware’s law, the delay period is three years.) Such a delay can be similar to that imposed by classified boards, which require acquiring shareholders to win two or more consecutive board elections to wield board control. 4.b. Distribution of the number of provisions by industry We also examine industry clustering in the use of governance provisions. Firms are grouped by two-digit SIC codes provided in the 1994 Compustat files. Adjacent two-digit industry groupings with fewer than nine firms are combined into a single industry, so that

- 11 -

each industry group has a minimum of nine firms. Using this procedure, the sample is divided into 25 industry groups. Table 5 reports the results of one-way analysis of variance tests using industry as the categorizing variable. As shown in the top row, in no single year can we reject at the 5% level the null hypothesis that the total number of provisions is unrelated to industry group. In one year, 1986, the p-value for the F-statistic used to test the null hypothesis is .065, but in other years the F-statistics are much lower. This result implies that the total number governance provisions employed by a firm is not systematically related to the firm's industry. As shown in Table 5, however, the uses of several individual provisions cluster by industry, at least in some years. These include poison pills, blank check preferred stock, classified boards, control share acquisition state laws, fair price laws, confidential voting, and cumulative voting. The table also identifies the industries in which each of the provisions appear with unusually high or low frequencies. For example, poison pills appear with relatively low frequency among electric, gas, and sanitary utilities (SIC 49), and such utilities tend to have high rates of cumulative voting and coverage by fair price laws. 4.c. How provisions cluster In this section we report the results of factor analyses to identify which of the various provisions tend to be used together. Each factor analysis was performed using data on all 20 governance provisions in each year 1984 through 1989.6 The variable for each provision is assigned a value of one if the firm has the provision, and zero otherwise. Factor analysis identifies linear combinations (i.e., factors) of the provisions that explain sizable portions of the variance in firms' uses of the provisions. We infer how the provisions cluster from the factor loadings. A factor loading can be interpreted as the proportion of the provision's use that is explained by that factor. Two provisions with high loadings on the same factor therefore can be interpreted as clustering together. To infer clustering in our application, we require that the provisions' factors each be 0.50 or greater, and that the provisions' pairwise correlation be statistically significant at the .001 level. When two provisions have factor

a description of factor analysis, see Overall and Klett (1972). Since no firms were covered by freezeout or poison pill laws in 1984, these provisions are omitted in the 1984 tests. Poison pill laws are omitted in 1985 tests

6For

- 12 -

loadings greater than 0.5, opposite signs, and a statistically significant negative pairwise correlation, we infer that the provisions tend to be used separately. To illustrate how we draw inferences, Table 6 reports factor loadings using 1989 data for the three factors with the highest eigenvalues, which together explain 29% of the variance in provision use. The factor loadings are based on a varimax rotation of the component axes to obtain interpretable results. The first factor in Table 6 has the highest loading on classified boards, and also high positive loadings on shareholder meeting requirements, supermajority vote requirements, and fair price charter amendments. We call this the internal governance factor. The factor loadings indicate that these four provisions tend to appear together in firms’ governance structures. As reported in the table, shareholder meeting requirements, supermajority vote requirements, and fair price charter amendments also are positively and significantly correlated with the use of classified boards. Two other provisions are significantly correlated with the use of classified boards, but their factor loadings indicate that less than half of the variance in their use is explained by the internal governance factor. The second factor in Table 6 has a high positive loading on control share acquisition laws and fair price laws, and a negative loading on freeze-out laws. We call this the state law factor. The third factor has high positive loadings on the variables that represent recent reincorporations to Delaware and recent elimination of cumulative voting. We performed factor analyses such as that reported for 1989 for all years in our sample. With minor variations, the internal control, state law, and anti-cumulative voting factors have the highest eigenvalues, although not always in that order. Plots of the factors' eigenvalues indicate that, in most years, the governance provisions are parsimoniously modelled using only three factors. Table 7 reports a summary of the factor analysis results for all sample years. The table reports, in descending order, the provisions that load highest on each of the three key factors. For a provision to be listed in Table 7, the provision's factor loading must be at least 0.5 in absolute value and the provision must be significantly correlated (at the .001 level) with the provision that has the highest loading on the factor. The results indicate that several provisions tend to be observed together:

- 13 -

1. In all six years, classified boards, shareholder meeting requirements, and supermajority voting rules all have high loadings on the internal governance factor. Thus, these three provisions tend to be used together. In two years (1985 and 1988), stakeholder provisions also load heavily on this factor, and in two other years (1984 and 1989), fair price charter amendments have high loadings. These results indicate that a typical firm with any one of these provisions is likely to have one or more of the others. 2. The state law factor indicates that, in 1988 and 1989, firms covered by freeze-out laws tend not to be covered by control share acquisition laws, and, to a lesser extent, fair price laws. This reflects Delaware's 1988 adoption of a freeze-out law, which instantly covered more than half the firms in the sample. Delaware does not also have control share acquisition or fair price laws, so coverage by these types of law are negatively correlated with coverage by a freeze-out law. Before 1988, control share laws tend to appear together with poison pill laws (in 1986 and 1987) and with cumulative voting provisions (in 1984 and 1985). 3. The anti-cumulative voting factor indicates that firms reincorporating to Delaware also tend to eliminate cumulative voting. Many states, including California, require cumulative voting, whereas Delaware does not. Thus, it appears that one motive for reincorporating to Delaware is to eliminate cumulative voting. This result is consistent with Heron and Lewellen’s (1997) finding that many firms reincorporating to Delaware simultaneously eliminate cumulative voting. Several other factors with lower eigenvalues indicate that some other pairs of provisions tend to appear together in individual years. Most notably, fair price charter amendments and anti-greenmail provisions are positively and significantly correlated in 1985 through 1989. Anti-greenmail provisions forbid targeted share repurchases at higher-than-market prices. Fair price provisions forbid many two-tiered tender offers in which the price paid in the second tier is lower than that paid in the first tier. The use of both provisions together reflects a policy of assuring that all shares are treated equally in control or repurchase transactions. Another notable result that is not included in Table 7 is that poison pills are negatively correlated with unequal voting rights in 1988 and 1989. This suggests that managers of firms with unequal voting rights have been relatively slow to implement poison pills, perhaps because unequal voting rights and poison pills are substitute takeover defenses.

- 14 -

4.d. Delaware v. other corporations Research by Dodd and Leftwich (1980) and Herron and Lewellen (1997) suggests that companies incorporated in Delaware have different governance structures than firms incorporated in other states. At the very least, Delaware firms differ from other firms in the types of state antitakeover laws that apply to them. It turns out that Delaware firms also differ from non-Delaware firms in their use of several other provisions. For each provision, we calculated a Chi-squared statistic to test the null hypothesis that the proportions of Delaware firms and non-Delaware firms deploying the provision are equal. The null hypothesis is rejected at the .05 significance level for several provisions. In 1989, for example, 265 of the 513 sample firms were incorporated in Delaware. In this year, the Delaware firms have unequal voting rights, supermajority vote requirements, shareholder meeting requirements, and director/officer liability indemnity provisions more frequently than non-Delaware firms, and have cumulative voting less frequently than other firms. Delaware firms also have a significantly greater average percentage of inside board members than non-Delaware firms. Some of these differences reflect the uniqueness and influence of Delaware corporate law. For example, Delaware permitted firms to adopt director/officer indemnity provisions beginning in June 1986, before most other states did. A disproportionate number of firms adopting such provisions therefore were incorporated in Delaware during our sample period. (Even after 1986, Herron and Lewellen (1997) report that many firms reincorporated to Delaware so that they could adopt such provisions.) Similarly, the fact that some states require cumulative voting, whereas Delaware does not, suggests that cumulative voting would be observed relatively less frequently among Delaware firms. Despite any differences that mechanically may reflect Delaware’s corporate law, the results indicate that Delaware firms deploy a greater number of antitakeover provisions than do non-Delaware firms. Even for those antitakeover provisions for which the test statistic is small, the provisions are relatively frequent among Delaware firms. Sixty-two percent of Delaware firms have poison pills, for example, compared to fifty-seven percent for nonDelaware firms. (The Chi-squared test statistic is 1.34 with a p-value of .25.) Overall, of the 15 provisions in our sample that are not state takeover laws, Delaware firms have an average

- 15 -

4.67 in 1989, compared to 4.06 for non-Delaware firms. This difference is statistically significant at the .001 level (t-statistic = 3.92).

4.e. Provisions’ relations with ownership structure and board composition Recent research suggests that the use and share value effects of certain corporate governance provisions are affected by the firm's ownership structure and board composition. For example, Brickley, Coles, and Terry (1994) and McWilliams and Sen (1997) conclude that the share value effects of announcements that a firm is adopting an antitakeover provision depends on the mix of insiders and independent outsiders on the firm's board. Boyle, Carter, and Stover (1998) report that, for savings and loans associations converting to stock ownership, the use of antitakeover provisions is negatively related to managerial share ownership. To examine whether ownership structure and board composition are related to a firm's tendency to deploy any of the 20 governance provisions, we obtained ISS data on ownership and board composition as of December 1989. Table 8 reports the results of a series of paired sample t-tests for these three provisions: poison pills, blank check preferred stock, and confidential voting. Results for other provisions are statistically insignificant. For each provision, we partition the sample into groups with and without the provision, and test the null hypothesis that the ownership or board composition measure differs between the groups. We use within-sample comparisons because comparisons with other samples of firms (e.g., as reported by Mikkelson and Partch (1989)) are complicated by slightly different definitions of managerial and outside blockholder ownership. In our sample, managerial ownership is defined as the fraction of shares held by officers and directors. Outside blockholder ownership refers to the fraction of shares held by shareholders who are not managers, institutions, or employees and who own at least 10,000 shares. The first row of Table 8 reports that, among firms with poison pills, the mean institutional share ownership is 55.3%. For firms without poison pills, the mean is 48.0%. The t-statistic for the difference in means is 5.35, indicating that, on average, firms with poison pills have relatively high institutional ownership. Consistent with the findings of Malatesta and Walkling (1988), the evidence in Table 8 also indicates that firms with poison pills have relatively low managerial ownership (4.8%, compared to 9.5% for non-poison pill firms).

- 16 -

High institutional share ownership and low managerial ownership both facilitate outside takeover, so these results suggest that poison pills tend to be used by firms that are relatively vulnerable to outside takeover. This result is consistent with our earlier finding that the use of poison pills is negatively correlated with the existence of unequal voting rights. As reported in Table 8, firms with poison pills have relatively low ownership by outside blockholders. This observation is consistent with conjectures that large blockholders tend to prevent managers from adopting poison pills, or that the presence of a large blockholder decreases a poison pill's net benefit to managers. Table 8 also shows that firms with poison pills tend to have a slightly smaller percentage of inside board members (23.9%, compared to 26.4% for non-poison pill firms), and a higher percentage of independent board members (62.1% versus 57.6%). These results are inconsistent with the view that poison pills tend to be implemented by management-controlled boards. Firms with blank check preferred stock, like those with poison pills, tend to have low managerial ownership, low insider representation on the board, and a high proportion of independent directors. Firms with confidential voting have low mean managerial and outside blockholder shareholdings. For all other provisions, however, there is little evidence of a systematic relation between provision use, ownership structure, and board composition. 5. Issues raised by the patterns of adoption 5.a. Why do we observe these patterns? Our results raise two questions about the reasons firms employ governance provisions. First, why do firms adopt governance provisions? One conjecture is that firms adopt provisions incrementally and in response to specific perceived needs. For example, managers fearing a mid-year proxy fight might seek a prohibition on the right to act by written consent (which we categorize as a shareholder meeting requirement), but not a fair price charter amendment. Managers of a firm facing a potential two-tiered tender offer, in contrast, might seek a fair price charter amendment but not a prohibition on the right to act by written consent. This conjecture implies that the increasing complexity of firms' governance structures reflects unique changes in the firms' business conditions and risks of outside takeover.

- 17 -

An alternative conjecture is that firms acquire restrictive governance provisions without regard to the provisions' specific attributes. According to this view, managers seeking takeover protection adopt many possibly redundant defensive provisions: "If a little takeover protection is good, more is even better." It is possible that firms mimic each other when adopting provisions. According to this view, the increase in provision use reflects herding behavior, as managers copy the provisions used at other firms. These conjectures are not mutually exclusive. It is possible that some firms adopt governance provisions en masse, copying provisions that they or their legal counsel observe at other firms. It also is possible that firms adopt some provisions for specific purposes. These purposes could include attempts to defend takeover bids, responses to shareholder pressure, or in the case of director and officer liability indemnity, a desire to attract qualified independent directors. The second question raised by our results concerns the reasons certain provisions are used together. One possibility is that classified boards, supermajority provisions, fair price provisions, and shareholder meeting requirements tend to appear together because managers view their effects as complementary. An alternative explanation, however, is that these provisions are jointly determined by some underlying factors that we do not measure. Yet a third conjecture is that there are economies to the adoption of certain governance provisions. Shareholder approval for one defensive charter amendment may increase the probability that managers can successfully propose additional antitakeover charter amendments. If so, one would expect such charter amendments as classified boards, supermajority provisions, fair price provisions, and shareholder meeting requirements to appear together. 5.b. Implications for empirical research Many investigations of individual governance provisions rely implicitly on the assumption that the provisions are used independently of each other. Our data show that this assumption sometimes is justified, and other times is not. The use of poison pills, for example, appears not to cluster with most other provisions, except perhaps for a negative correlation with unequal voting rights. Other provisions -- especially classified boards, fair price charter amendments, shareholder meeting requirements, and supermajority voting rules -- cluster together, indicating that the independence assumption does not hold for these provisions.

- 18 -

If any one provision's valuation effects depend on the presence or absence of other provisions, then a firm's governance structure can provide important conditioning information when investigating a single provision's valuation effects. This is illustrated by previous research that shows how information about a firm's governance structure can help explain the cross-section of abnormal returns in event studies. For example, Karpoff and Malatesta (1989, 1995) show that the valuation effect of a state antitakeover law depends on whether a firm has other takeover defenses, suggesting that the marginal change in takeover deterrence depends on the firm's pre-existing governance provisions. As another example, Brickley, Coles, and Terry (1994) find that the average abnormal stock market reaction to poison pill adoption announcements is positive for firms with a majority of independent board members and negative for other firms. In both of these examples, a provision's conditional valuation effect differs from its unconditional effect. Studies that ignore interactions with other provisions can misestimate the unconditional importance of certain governance provisions. The relations between individual governance provisions and measures of firm performance also could be complicated by the existence of other provisions. Gordon and Pound (1992), Hirschey and Jones (1995), and Stangeland (1995) examine the correlations between financial performance and several of our 20 governance provisions, each considered separately. Hirschey and Jones (1995), for example, find that the adoptions of classified board, fair price provisions, and supermajority vote requirements all are associated with high industry-adjusted operating performance. Since the uses of these three provisions are highly correlated, the empirical results may not be independent of each other. 6. Conclusions Researchers have examined the effects of many corporate governance provisions on firm value and operations, typically each in isolation. Firms, however, use these provisions in groups. From 1984 through 1989, the average firm in our sample increased the number of provisions in its governance structure from 2.07 to 5.93. This increase immediately preceded the decline in takeover frequency in the early 1990's, consistent with the argument that a rapid proliferation of governance provisions contributed to the takeover decline. The increased use of several key provisions -- including poison pills, director/officer liability

- 19 -

indemnity, and state antitakeover laws -- followed important court decisions that increased the provisions' net benefits. The proliferation of governance provisions reflects their broad-based adoption by many firms, not the accumulation of many provisions by a small number of firms. Firms tend not to specialize heavily in the categories of governance provisions they employ, as the numbers of external control and internal control provisions per firm are positively correlated. We also find that the intensity of provision use, as measured by the number of provisions employed by each firm, is not significantly related to industry grouping. Several individual provisions, however, cluster by industry. These include poison pills, blank check preferred stock, and cumulative voting. Factor analyses identify several subsets of provisions that tend to be used together. Most prominently, supermajority vote requirements, classified boards, and shareholder meeting requirements tend to be used together, sometimes in concert with fair price charter amendments or stakeholder clauses. Firms reincorporating to Delaware tend to eliminate cumulative voting, revealing one motive for the reincorporation. Certain state antitakeover laws also form patterns: in 1986 and 1987, firms covered by control share acquisition laws also tended to be covered by poison pill laws, whereas in 1988 and 1989 firms with control share acquisition laws tended not to be covered by freeze-out laws. We also observe slight tendencies for anti-greenmail and fair price provisions to be used in concert, and for firms with unequal voting rights to avoid poison pills. The broad increase in the number of governance actions and provisions used by corporations indicates that during the late 1980’s firms’ governance structures grew increasingly complex. Our results indicate that the uses of some provisions are not independent, implying that a more complete description of a firm’s governance structure can help explain the incremental valuation and operating consequences of some individual provisions.

- 20 -

References Agrawal, Anup and Gershon N. Mandelker, 1990. Large shareholders and the monitoring of managers: The case of antitakeover charter amendments, Journal of Financial and Quantitative Analysis 25, 143-161. Alexander, John C., M. Wayne Marr, and Michael F. Spivey, 1993. State constituency statute and firm value: An empirical examination, Clemson University working paper. Bhagat, Sanjai and Richard H. Jefferis, 1991. Voting power in the proxy process: The case of antitakeover charter amendments, Journal of Financial Economics 30, 193-226. Bhagat, Sanjai and Richard H. Jefferis, 1996. Corporate performance, governance, and discipline: The impact of defensive activity on takeovers and managerial turnover, University of Colorado working paper. Black, Bernard S., 1992. Institutional investors and corporate governance: The case for institutional voice, Journal of Applied Corporate Finance, Fall, 19-32. Boyle, Glenn W., Richard B. Carter, and Roger D. Stover, 1998. Extraordinary antitakeover provisions and insider ownership structure: The case of converting savings and loans, Journal of Financial and Quantitative Analysis, forthcoming. Brickley, James A., Jeffrey L. Coles, and Rory L. Terry, 1994. The board of directors and the enactment of poison pills, Journal of Financial Economics 35, 371-390. Brook, Yaron, and Ramesh K. S. Rao, 1994. Shareholder wealth effects of directors’ liability-limitation provisions, Journal of Financial and Quantitative Analysis 29, 481-497. Bruner, Robert F., 1991. The poison pill anti-takeover defense: The price of strategic deterrence, The Research Foundation of The Institute of Chartered Financial Analysts, Charlottesville, Virginia. Coffee, Jr., J.C., 1991. Liquidity versus control: The institutional investor voice, Columbia Law Review 91, 1336-1338. Comment, Robert and William G. Schwert, 1995. Poison or Placebo? Evidence on the deterrence and wealth effects of modern antitakeover measures, Journal of Financial Economics 39, 3-43. DeAngelo, Harry and Edward M. Rice. 1983, Antitakeover charter amendments and stockholder wealth, Journal of Financial Economics 11, 329-360. Dodd, Peter and Richard Leftwich, 1980. The market for corporate charters: “Unhealthy Competition” versus federal regulation, Journal of Business 53, 259-282. Eckbo, B. Espen, 1990. Valuation effects of greenmail prohibitions, Journal of Financial and Quantitative Analysis 25, 491-505. Garvey, Gerald T. and Gordon Hanka, 1996. The management of corporate capital structure: Theory and evidence, University of British Columbia and Pennsylvania State University working paper. Gordon, Lilli and John Pound, 1992. Governance matters: An empirical study of the relationship between corporate governance and corporate performance, Harvard University working paper. Heron, Randall A. and Wilbur G. Lewellen, 1997. An analysis of the reincorporation decision: The evidence since 1980, University of Notre Dame working paper, May 1997.

- 21 -

Hirschey, Mark, and Elaine Jones, 1995. Long-term performance and financial policies of shark repellentadopting firms, Advances in Financial Economics 1, 155-181. Jarrell, Gregg A. and Michael Bradley, 1980. The economic effects of federal and state regulations of cash tender offers, Journal of Law and Economics 23, 371-407. Jarrell, Gregg A. and Annette B. Poulsen, 1987. Shark repellents and stock prices: The effects of antitakeover amendments since 1980, Journal of Financial Economics 19, 127-168. Karpoff, Jonathan M. and Paul H. Malatesta, 1989. The wealth effects of second-generation state takeover legislation, Journal of Financial Economics 25, 291-322. Karpoff, Jonathan M. and Paul H. Malatesta, 1995. State takeover legislation and share values: The wealth effects of Pennsylvania's Act 36, Journal of Corporate Finance 1, 367-382. Kini, Omesh, William Kracaw, and Shehzad Mian, 1995. Corporate takeovers, firm performance, and board composition, Journal of Corporate Finance 1, 383-412. Linn, Scott C. and John J. McConnell, 1983. An empirical examination of antitakeover amendments on common stock prices, Journal of Financial Economics 11, 361-399. Malatesta, Paul H. and Ralph A. Walkling, 1988. Poison pills securities: Stockholder wealth, profitability, and ownership structure, Journal of Financial Economics 20, 347-376. Manne, Henry G., 1965. Mergers and the market for corporate control, Journal of Political Economy, 73, 110120. McWilliams, Victoria B., 1990. Managerial share ownership and the stock price effects of antitakeover amendment proposals, Journal of Finance 45, 1627-1640. McWilliams, Victoria B. and Nilanjan Sen, 1997. Board monitoring and antitakeover amendments, Journal of Financial and Quantitative Analysis 32, 491-505. McGurn, Patrick S., Sharon Pamepinto, and Adam B. Spector, 1992. State Takeover Laws, Washington, D.C.: Investor Responsibility Research Center. Mikkelson, Wayne and M. Megan Partch, 1989. Managers' voting rights and corporate control, Journal of Financial Economics 25, 263-290. Overall, John E. and C. James Klett, 1972. Applied multivariate analysis, New York, NY: McGraw-Hill. Pound, J., 1992. Raiders, targets, and politics: The history and future of American corporate control, Journal of Applied Corporate Finance 5, 6 - 18. Partch, M. Megan, 1987. The creation of a class of limited voting common stock and shareholder wealth, Journal of Financial Economics 18, 313-339. Rosenbaum, Virgina K., 1995. Corporate Takeover Defenses. Washington, D.C.: Investor Responsibility Research Center. Ryngaert, Michael, 1988. The effect of poison pill securities on shareholder wealth, Journal of Financial Economics 20, 377-417.

- 22 -

Table 1 Descriptions of the 20 corporate governance provisions examined in this study External control provisions 1. Poison pills - Securities that entitle their holders to special rights if the issuing firm becomes the subject of a takeover bid. In a typical case, the poison pill is established by issuing a special dividend in the form of a right to purchase additional shares of the issuing firm’s common stock. The rights trade with the common shares until a triggering event such as an unsolicited takeover bid. When such an event occurs, the rights detach and may be exercised at a low price by shareholders other than the bidder. In most poison pills, rights-holders can purchase shares in the bidding company at low prices if the bidder acquires control of the issuing company. Poison pills are also called “shareholders rights plans.” 2. Blank check preferred stock - Authorized preferred stock for which the board of directors has broad discretion to establish voting, dividend, conversion, and other rights. Blank check preferred stock can be issued to parties friendly to management to block unwanted hostile bids. A primary use, however, is as a vehicle to implement a poison pill. For firms already with poison pills, the blank check preferred stock is a largely redundant takeover defense. We therefore provide information in Table 2 on the number of firms with “stand alone blank check preferred stock,” i.e., for firms that do not also have a poison pill. 3. Stakeholder clause - Charter language that permits directors to consider the effects of their decisions on constituencies other than stockholders. For example, directors may, or may be required to, evaluate the impact of a proposed change in control on employees, host communities, suppliers and others. The clause provides a target company’s board members with an explicit legal basis to reject takeover bids that is attractive to shareholders. Internal control provisions 4. Classified (or staggered) board - Board in which directors are divided into separate classes and elected to overlapping terms. For example, a firm with a three-class classified board would hold elections each year for one-third of the board seats, each with a three-year term. Because only one-third of the board can be replaced each year, a hostile bidder with voting control of the corporation may be unable to control the board for up to two years. Classifying the board also can deter proxy contests for control, because only one-third of the directors stand for elections in any one year. 5. Fair price charter amendment - Requires a large shareholder to pay a price set by formula for all shares acquired in the back end of a two-tier acquisition. Typically the price to be paid is the highest price the shareholder paid for any shares acquired during the first stage of the acquisition. Most fair price provisions do not apply if the large shareholder’s offer is approved by a target company’s board or if the bidder obtains a specified supermajority level of approval from the target's shareholders. 6. Recent (within three years) reincorporation to Delaware - Corporations can change their state of incorporation to strengthen their antitakeover defenses. The majority of firms in our sample that reincorporate in Delaware moved from California, which requires cumulative voting and prohibits board classification. Delaware corporate law, in contrast, permits companies to choose whether to classify the board or to make cumulative voting available. 7. Supermajority vote requirement - Establishes a level of approval for specified actions that is higher than the minimum set by state law. Such provisions often establish approval levels of 75 or 85 percent for actions

- 23 that otherwise would require simple majority approval. These requirements often exceed the level of shareholder participation at a meeting, making action that requires supermajority approval very costly. Supermajority provisions can apply to mergers, sale of assets, or other specified transactions. Those that apply to the amendment of bylaws or corporate charters often are referred to as lock-in provisions. 8. Unequal voting rights - Occurs when common shares do not all have the same voting rights. Under one typical plan called a dual class capitalization, two classes of stock exist, one with voting rights superior to the other. A second type of plan grants long-term stockholders super voting rights. 9. Shareholder meeting requirements - Refers to restrictions on either of two mechanisms that otherwise can be used to circumvent the normal corporate decision-making process. The first restriction is on the right to call special shareholder meetings. The second restriction is on the right to act by written consent, which enables shareholders with sufficient votes to take actions that otherwise would have to await a special or annual shareholder meeting. Both restrictions override default provisions in most states’ incorporation statutes that permit special meetings or action by written consent. 10. Firm eliminated cumulative voting or right to alter board size - Indicates that a firm recently (within three years) has eliminated cumulative voting and replaced it with straight voting, or has adopted a charter provision requiring a supermajority vote among directors, shareholders, or both to alter the board size. In the absence of a board size restriction, large outside shareholders could avoid other antitakeover restrictions by increasing the size of the board and packing it with supporters. State takeover laws 11. Freeze-out law - Prohibits a large shareholder from engaging in any business combination with the covered firm for a specified number of years unless approval is obtained from the target firm's directors before the bidder acquires more than a specified fraction of target shares. Even after the mandatory waiting period, most freeze-out laws allow the business combination to proceed only if the transaction satisfies fair price provisions. Thus, the typical freeze-out law is like a fair price law with a forced delay. The Delaware freeze-out law, adopted in 1988, covers a large number of corporations. It requires a three-year waiting period, although it permits business combinations during the freeze-out period that are approved by twothirds of the voting stock not held by the interested shareholder. It also permits some business combinations with third parties, such as a sale of assets, during the freeze-out period. 12. Control share acquisition law - Requires shareholder approval before a large shareholder may vote shares obtained in a control share acquisition. A control share acquisition refers to an accumulation of shares to above a threshold level, for example, to one-fifth the outstanding shares of a covered corporation. 13. Fair price law - Similar to fair price charter amendments adopted by many firms. These laws regulate the back-end price in a two-tiered takeover bid or other significant business combination involving a large shareholder. The typical fair price law prohibits business combinations between the firm and a large stockholder unless one of two conditions is met. Either (1) prior approval is granted by a supermajority (e.g., 80%) of all outstanding voting stock and by a supermajority (e.g., two-thirds) of the outstanding stock not held by the interested stockholder; or (2) stockholders receive a stipulated price for the stock acquired by the large stockholder as part of the business combination. The stipulated price is set by a formula that guarantees the price paid likely will be very high. 14. Cash-out law - Requires any person who acquires a large stake (e.g., 20%) in a firm to notify all other shareholders of the acquisition. All other shareholders are then entitled to sell their shares to the acquirer at a price at least as high as the highest price the acquirer paid in the period over which the large shareholder acquired its shares. Adopted by only three states and rescinded by one state.

- 24 -

15. Poison pill law - Grants firms the right to adopt poison pill takeover defenses. These poison pill laws can be important because the right to use poison pill defenses is presumably more secure when explicitly authorized by statute and is thus less likely to be limited by the courts. Other miscellaneous provisions 16. Anti-greenmail provision - Greenmail refers to an arrangement in which a company repurchases the stock held by a large shareholder, usually at an above-market price, in exchange for the shareholder’s agreement not to launch a contest for control of the company for a specified time period. An anti-greenmail provision prohibits a company’s managers from entering such an agreement unless the repurchase offer is made to all shareholders on a pro-rata basis or shareholders approve of the transaction. 17. Confidential voting - Establishes a procedure in which all proxies, ballots, and voting tabulations that identify individual shareholders are kept confidential. Only vote tabulators and election inspectors may examine individual proxies and ballots. Managers and shareholders are told only of vote totals. 18. Cumulative voting - Permits shareholders to distribute their total votes in any fashion they desire among the nominees to a company's board. Each shareholder’s total votes is equal to the number of shares held times the number of directors to be elected. By concentrating votes on selected candidates, minority shareholders can elect a small number of representatives to the board. 19. Director/officer liability indemnity - Such indemnity occurs when the corporation adopts a provision in which it promises to reimburse its directors and/or top officers for legal expenses, damages, and judgements incurred as a result of any lawsuit relating to the directors’ and officers’ corporate actions. In virtually all cases, firms that adopt such a provision purchase indemnity insurance to cover its risk. 20. Firm opted out of state takeover law - Some state takeover laws contain language that allows affected companies not to be covered by part or all of the law’s provisions. Firms that have taken steps not to be covered by a state takeover law are identified in the sample.

- 25 Table 2 Frequencies of 20 corporate governance provisions in the sample of 513 (primarily S&P 500) companies tracked by Institutional Shareholders Services, Inc. on behalf of large investor clients. Descriptions of the provisions are provided in Table 1. Corporate governance provision Panel A: External control provisions Poison pill 10 Blank check preferred stock -- Stand alone blank check preferred stocka Stakeholder clause Panel B: Internal control provisions Classified board Fair price charter amendment Firm reincorporated to Delaware Supermajority vote requirement Unequal voting rights Shareholder meeting requirements Firm eliminated cumulative voting or right to alter board size Panel C: State takeover laws Freeze-out law Control share acquisition law Fair price law Cash-out law Poison pill law Panel D: Other miscellaneous provisions Anti-greenmail provision Confidential voting Cumulative voting Director/officer liability indemnity Firm opted out of state takeover law Mean number of provisions per firm Median number of provisions per firm
a

1984

1985

1986

1987

1988

1989

20 382 373 13

140 399 381 17

182 413 287 24

253 424 257 26

307 426 195 26

431 156 28

166 96 4 79 39 81 4

210 153 7 103 41 123 10

244 186 11 119 44 162 19

255 199 9 124 49 182 22

265 204 6 131 51 195 22

277 206 3 134 52 203 16

0 34 22 15 0

43 34 53 17 0

77 43 53 17 33

90 81 65 17 35

398 93 65 17 93

427 116 65 17 156

4 17 76 17 3 2.07 2.00

36 17 80 20 3 2.70 2.00

40 17 81 56 4 3.48 3.00

44 18 81 349 5 4.40 4.00

46 23 82 430 6 5.52 6.00

47 26 82 441 6 5.93 6.00

Blank check preferred stock when the firm does not also have a poison pill. Stand alone blank check preferred stock is not used in calculating the mean or median number of provisions per firm.

- 26 -

Table 3 Distributions of the number of provisions per firm in each year, 1984 - 1989, for 513 sample firms. Data on 20 types of provisions are provided by Institutional Shareholders Services, Inc. (Graphics may not translate.)

Mean = 2.07 Median = 2 Standard deviation = 1.39 Skewness = 0.66

Mean = 2.70 Median = 2 Standard deviation = 1.63 Skewness = 0.41

Mean = 3.48 Median = 3 Standard deviation = 1.88 Skewness = 0.28

Mean = 4.40 Median = 4 Standard deviation = 1.94 Skewness = 0.13

Mean = 5.52 Median = 6 Standard deviation = 1.95 Skewness = -0.09

Mean = 5.93 Median = 6 Standard deviation = 1.91 Skewness = -0.17

- 27 -

Table 4 Spearman correlation coefficients (with significance levels in parentheses) using 1989 data for the number of provisions firms have in each of four categories. The categories and provisions are described in Table 1. Data are from 513 (primarily S&P 500) companies tracked by Institutional Shareholders Services, Inc. on behalf of large investor clients.

Number of external control provisions

Number of internal control provisions

Number of state antitakeover laws

Number of internal control provisions Number of state antitakeover laws Number of other miscellaneous provisions

.282 (.000) -.012 (.779) .003 (.941) -.113 (.010) .077 (.080) -.025 (.572)

Table 5 F-statistics from one-way analysis of variance tests designed to test whether provision use is related to industry group. Significance levels are in parentheses. The 513 sample firms are partitioned into 25 industry groups based on 1994 Compustat SIC numbers. Results are reported for the total number of provisions per firm, and for the individual provisions for which evidence of industry clustering is apparent in more than two individual years. For each provision not included in the table, we cannot reject the null hypothesis in most years that the provision’s use is unrelated to industry grouping. Descriptions of the provisions are provided in Table 1. Corporate governance provision represented) Total provisions 0.74 Individual provisions: Poison pill 0.77 Blank check preferred stock Classified board Control share acquisition law Fair price law 1.11 (.777) 1.46 (.074) 1.16 (.277) 1.45 (.077) 1.33 (.135) 1.58 (.042) 2.58 (.000) 1.59 (.329) 0.70 (.854) 1.52 (.056) 1.45 (.077) 1.32 (.143) 1.58 (.042) 2.37 (.000) 2.04 (.039) 1.40 (.100) 1.65 (.028) 1.51 (.059) 1.32 (.143) 1.58 (.042) 2.35 (.000) 1.87 (.003) 1.57 (.044) 1.78 (.013) 1.48 (.068) 1.54 (.049) 1.83 (.010) 2.35 (.000) 2.11(SIC 49: Utilities) (.008) (.002) 1.46 (.073) 1.63 (.030) 1.45 (.077) 1.54 (.049) 1.52 (.055) 2.43 (.000) 1.51 (.057) 1.67 (.026) 1.11 (.322) 1.54 (.049) 1.37 (.114) 2.43 (.000) (SIC 27: Printing, publishing) (SIC 36: Electronics) SIC 30-31: Rubber, leather products SIC 54: Food stores SIC 49: Utilities SIC 63-69: Insurance and holding companies SIC 29: Petroleum refining SIC 60: Commercial banks SIC 49: Utilities 1.15 (.270) 1.49 (.288) 1.15 (.065) 1.05 (.285) 0.98 (.401) (.490) 1984 1985 1986 1987 1988 1989 Industries heavily represented (or, in parentheses, under-

Confidential voting Cumulative voting

Table 6 Loadings for the first three factors from a principal components factor analysis using 1989 data from 513 firms. The factor matrix is rotated using the varimax method. The variable for each provision is assigned a value of one if the firm has the provision, and zero otherwise. The factors are linear combinations of the provisions that explain sizable portions of the variance in firms' use of the corporate governance provisions. We infer how provisions cluster from the factor loadings. Boldface type denotes provisions with factor loadings of at least 0.5. Descriptions of the provisions are provided in Table 1. Factor 1 "Internal control factor" Factor 2 "State law factor" Factor 3 "Anti-cumulative voting factor"

Corporate governance provision External control provisions Poison pills Blank check preferred stock Stakeholder clause Internal control provisions Classified board Fair price provision Firm reincorporated to Delaware Supermajority vote requirement Unequal voting rights Shareholder meeting requirements Firm eliminated cumulative voting or right to alter board size State takeover laws Freeze-out law Control share acquisition law Fair price law Cash-out law Poison pill law Other provisions Anti-greenmail provision Confidential voting Cumulative voting Director/officer liability indemnity Firm opted out of state takeover law Percent of variance explained by factor Factor Eigenvalue
a b

.298b .275 .226 .674a .579b -.051 .595b .019 .610b -.019

-.045 .114 .039 .012 -.056 -.083 -.049 -.156 -.182b .091

.022 -.027 -.064 -.128 .035 .796b -.092 -.027 .103 .852a

-.052 -.100 -.174 -.061 -.024 .308b -.267 .140 .007 -.022 11.9 2.39

-.713b .770a .605b -.277 .263b .083 -.083 .362b -.243b .014 9.4 1.89

-.112 .023 -.155 -.065 -.006 .108 -.111 -.096 .087 .031 7.5 1.51

Identifies the provision with the highest loading for the factor. The pairwise Spearman rank-order correlation coefficient between the provision and the provision with the highest loading for the factor is significant at the .001 level.

Table 7 Summary of principal components factor analyses using data each year separately, 1984 - 1989. These three factors explain the largest portions of the total variance in firms’ use of 20 corporate governance provisions. The provision listed first for a given factor and year has the highest loading for that factor and year. Other provisions in each cell are listed in descending order of the absolute value of the factor loading. To be included in the table, each provision’s factor loading must be at least 0.5 in absolute value, and be significantly correlated at the .001 level with the provision listed first in the cell. Provisions in parentheses have negative factor loadings. Provisions that load heavily on an identical single factor are interpreted as being correlated with each other. Descriptions of the provisions are provided in Table 1. Factor 1 "Internal control factor" Shareholder meeting requirements Supermajority vote requirement Classified board Fair price charter amendment Shareholder meeting requirements Supermajority vote requirement Classified board Stakeholder clause Shareholder meeting requirements Supermajority vote requirement Classified board Shareholder meeting requirements Supermajority vote requirement Classified board Supermajority vote requirement Classified board Shareholder meeting requirements Stakeholder clause Classified board Factor 2 "State law factor" Control share acquisition law Cumulative voting Factor 3 "Anti-cumulative voting factor" Reincorporated to Delaware Eliminated cumulative voting

Year 1984

1985

Control share acquisition law Cumulative voting

Reincorporated to Delaware Eliminated cumulative voting

1986

Control share acquisition law Poison pill law Poison pill law Control share acquisition law Freeze-out law (Control share acquisition law) (Cumulative voting) Control share acquisition law

Reincorporated to Delaware Eliminated cumulative voting Eliminated cumulative voting Reincorporated to Delaware Eliminated cumulative voting Reincorporated to Delaware

1987

1988

1989

Eliminated cumulative voting

- 31 Shareholder meeting requirements Supermajority vote requirement Fair price charter amendment (Freeze-out law) Fair price law Reincorporated to Delaware

- 32 -

Table 8 Tests of differences in ownership structure and board composition between firms with and without three selected provisions, using 1989 data on 513 firms provided by Institutional Shareholders Services, Inc. For omitted provisions, the differences between firms with and without the provisions are generally statistically insignificant. Descriptions of the provisions are provided in Table 1. Percent of firm's common stock owned by: Corporate governance provision Institutions Managers Blockholders Employees Percent of firm's board consisting of: Insiders Affiliated outsiders Independent outsiders

Firms with poison pills Firms without poison pills t-statistic for difference in means Firms with blank check preferred stock Firms without blank check preferred stock t-statistic for difference in means Firms with confidential voting Firms without confidential voting t-statistic for difference in means

55.3 48.0 5.35 52.9 49.7 1.68 52.3 55.3 -1.00

4.8 9.5 -4.25 6.1 9.8 -2.24 3.8 6.8 -2.12

3.8 8.1 -2.23 5.1 6.9 -0.66 0.6 5.6 -4.76

9.4 10.1 -0.50 9.5 10.0 -0.35 10.6 9.5 0.35

23.9 26.4 -2.02 24.3 28.5 -2.36 21.7 25.1 -1.80

14.0 16.0 -1.79 14.5 16.2 -1.08 13.7 14.8 -0.48

62.1 57.6 2.80 61.2 55.4 2.58 64.6 60.1 1.39


				
DOCUMENT INFO