Examining Incentive-Based Executive Compensation in West Virginia: A Low Ranking State, Exams of Art

Download Examining Incentive-Based Executive Compensation in West Virginia: A Low Ranking State and more Exams Art in PDF only on Docsity! The West Virginia Public Affairs Reporter Open for Business? An Examination of Incentive-Based Executive Compensation in West Virginia Brandon N. Cline and Nima Mohebbi Volume XXV, No. II Institute for Public Affairs West Virginia University Morgantown, West Virginia October 2008 Institute for Public Affairs October 2008 The West Virginia Public Affairs Reporter Table of Contents: Volume XXV, No. II Open for Business? An Examination of Incentive-Based Executive Compensation in West Virginia Brandon N. Cline and Nima Mohebbi........ 2 About the Cover Robert Bridges..................................... 15 The West Virginia Public Affairs Reporter is a Publication of: The Institute for Public Affairs Eberly College of Arts and Sciences West Virginia University PO Box 6317 Morgantown, WV 26506-6317 Kevin M. Leyden, Ph. D., Director/Editor Thomas K. Bias, Assistant Editor The West Virginia Public Affairs Reporter is a refereed journal of the Institute for Public Af- fairs. Each issue addresses topics of concern to West Virginia’s citizens and state and local offi- cials. The Reporter’s content is written by pub- lic officials, faculty, research associates, and others familiar with West Virginia affairs. All published articles have gone through a blind re- view process. In all cases, the views of the au- thors are not necessarily those of the Institute for Public Affairs nor of West Virginia University. Journal Production by Sarah M. Bias Cover William Robinson Leigh (American 1866-1955) Three Figures in an Office (illustration for The Golden Flood) 1904 charcoal and gouache on paper 25 x 19 inches West Virginia University Art Collection Gift of Harvey D. Peyton, Nitro, West Virginia Permission of use from Robert Bridges, Curator of the West Virginia University Art Collection 2 Open for Business? An Examination of Incentive-Based Executive Compensation in West Virginia 1. Introduction: The primary goal of incentive-based executive compensation is to align the interests of managers with those of shareholders, namely to increase firm value. Traditional agency theory (Jensen and Meckling 1976; Holmstrom 1979; Grossman and Hart 1983; Murphy 1985; Holmstrom and Milgrom 1987; Abowd 1990; Jensen and Murphy 1990a; and Murphy 1998) suggests that managers should be given incentives to maximize shareholder wealth (stock price) when effort is non-observable or monitoring costs are high. Jensen and Meckling (1976) along with Jensen and Murphy (1990a) show that by properly structuring an incentive contract, firms can motivate corporate insiders to enhance corporate risk taking while developing growth opportunities. In order to more closely align the interests of corporate insiders and shareholders, and thus reduce the potential for agency conflict, executives are often provided with an equity interest in firms. Typically, firms allow discounted stock purchase programs or grant stock options in an attempt to reduce agency conflict. West Virginia currently lags in most fundamental economic indicators, particularly ranking 49th in output per capita (Dubay 2006). This is a telling indication of poor productivity and a deficient business climate. After he was inaugurated, one of the primary focuses of Governor Joe Manchin’s administration was a reformation of this dismal climate in West Virginia, a task ambitiously titled: West Virginia, Open for Business. Undoubtedly, many significant structural impediments remain in the way of this reformation. Indeed, studies addressing these issues have been confined primarily to analyses of the regulatory compliance and administrative costs of doing business in West Virginia. We, however, illuminate a problem with executive compensation in firms in West Virginia, an issue that up to this point has not been examined. It is no secret that West Virginia has struggled economically. Its dismal economic condition can be attributed, in part, to a poor business climate that has hindered entrepreneurial activity and forced major firms to locate outside of the West Virginia border (Sobel 2007). In 2005, West Virginia ranked 48th in income per capita with an average income of $26,029; only hurricane-plagued Louisiana and Mississippi ranking lower (Sobel 2007). When firms invest in capital and perform well, marginal productivity tends to increase, raising income levels and standards of living for all employees. We analyze compensation from a finance-centered perspective, keeping these economic assumptions in mind and focusing on the ratio of total compensation to incentive-based compensation. In other words, we do not question whether firms in West Virginia are paying Institute for Public Affairs October 2008 5 2.3. Restricted Stock Tax Federal Treatment Restricted stock is another form of incentive-based pay. Restricted stock represents equity ownership in the form of shares of stock that can be sold after some pre-specified vesting date. One way the government has enticed firms to pay more performance-based pay is by adding an additional feature to the tax code in 1993. This feature prohibited any corporate deduction for executive pay above $1 million that was not performance-based. Restricted stock is not considered to be performance-based pay by the IRS, and is therefore, subject to this rule, making it taxed at the usual income rate. 2.4. Bonus Bonuses, if adequately structured, can provide managers with incentives to maximize shareholder wealth. Bonus compensation, however, is a slightly different type of incentive-based pay. Whereas executive stock options and restricted stock are always linked to overall firm performance, a bonus can be linked to individual, group, or to the firm’s performance. Balsam (2002) reports that for most executives a bonus is based on a large group’s performance. Balsam also notes that depending on the firm a bonus can be entirely subjective or based on objective criteria such as sales revenue, new clients, or production. Bonuses are treated as ordinary income for executives and in most cases deducted under Section 162 of the Internal Revenue Code as an ordinary business expense for most firms. However, there are some limitations in Section 162(m) depending on whether the bonus was paid with cash or stock. 3. Data and Methodology In this section we identify the data set and describe the methodology used for our study. 3.1. Data Data on executive compensation is contained in firm proxy statements filed with the Securities Exchange Commission (SEC). These statements are maintained and then made publically available by the SEC.3 Data venders such as Standard and Poor’s ExecuComp compile these proxy statements into usable databases. The ExecuComp database itself contains over 100 compensation and financial variables on top executives for over 2,700 companies that are currently included or were once included in the S&P 1500. The primary dataset used in this study consists of data reported in Standard and Poor’s ExecuComp for the period 1992 through 2006. After the removal of multiple stock and option grants and missing data, 131,890 firm-year compensation observations are retained for 29,179 executives from 2,779 firms. Of the total 131,890 observations, 21,965 are reported 3 Proxy statements can be found and viewed publically on the Securities Exchange Commission’s website at www.sec.gov. on 4,856 CEOs from 2,777 firms. We augment the ExecuComp database with 10 publicly traded firms headquartered in the state of West Virginia that were not contained in the ExecuComp database.4 3.2. Methodology Executive compensation has recently become a heavily debated topic academically, politically and in the popular media.5 Many of the critics such as Rose (1998) and Crystal (1991) argue that executives are paid too much. Others such at Kay (1998) argue that firms need to offer compensation packages at these levels to attract, retain, and motivate quality employees. Yet there is a third contingent (e.g., Jensen and Murphy 1990b) that claim it is not how much you pay executives, but how you pay them that really matters. Our hypotheses and methodology align with this third contingent. Specifically, we argue, based on extant theoretical and empirical literature, that firms offering a large proportion of total pay in the form of incentive-based compensation have executives that are highly motivated to increase firm value. This increased incentive results in higher performance and leads to overall economic growth. Hence, it is not necessarily the increased compensation that contributes to this higher level of productivity; rather, the percentage of incentive-based pay offered to executives. As such, our analysis is conducted on the percentage of incentive- based pay to total compensation. States are ranked according to the mean percentage of performance-based compensation, where performance-based compensation is defined as the Black-Scholes-Merton option value6 granted plus the value of restricted stock granted. The percentage of performance compensation is then measured as the relative value of percentage compensation to total executive compensation, where total executive compensation is defined as the sum of salary, bonus, Black-Scholes-Merton option value, and restricted stock value. 4. Results This section is divided into three sub-sections according to the level and rank of the corporate executive. In full, three distinctions are made: (1) all top executives included in the Standard and Poor’s ExecuComp database, (2) only the CEO, and (3) only the top four ranking executives excluding the CEO. 4 All tests were also conducted excluding these 10 firms and the results were consis- tent with those reported in this study. 5 The Wall Street Journal, Business Week, Forbs, and Fortune all frequently publish annual surveys of executive pay. 6 The Black-Scholes-Merton option value is calculated using the Black-Scholes Option Pricing Model. The Black-Scholes Option Pricing Model is a widely used option valuation formula first published in 1973 by Fisher Black and Myron Scholes. This model assumes the underlying asset follows a lognormal distribution and that the volatility of the underlying is constant. Institute for Public Affairs October 2008 6 4.1. Results According to All Executives As previously noted, the primary hypothesis of this study is that improperly motivated executives are one contributing factor to the state of West Virginia’s lackluster economic performance. Specifically, we hypothesize that West Virginia not only ranks low with respect to economic performance but also ranks low compared to other states’ level of incentive-based pay. In this sub-section we test our hypothesis using all executives reported in the Standard and Poor’s ExecuComp for the period 1992 through 2006. After the removal of multiple stock and option grants and missing data, 131,890 compensation observations are retained for 29,179 executives from 2,779 firms. We calculate and present several compensation statistics but focus on the proportion of total pay in the form of incentive- based compensation as our primary measure. These results are reported in Table 1. The sample including all executives from all states is reported in Panel A. Panel B provides the comparison across states. Panel A indicates that on average top executives are paid approximately $1.8 million per year. Of this $1.8 million a mean 35.47% (34.67% median) is provided in the form of incentive-based pay, where performance-based compensation is defined as the Black-Scholes-Merton option value granted plus the value of restricted stock granted. Most studies suggest that because managers are risk averse and shareholder are risk neutral, options are the best form of incentives (Inderst and Muller 2004). Thus, of particular interest is the proportion paid in the form of stock options. We report for all the states contained in our sample a mean of 29.68% (median 25.52%) in total compensation is paid using executive stock options. Panel B ranks each state according to the proportion of incentive based pay. Lending further support to our prediction, we find that West Virginia ranks 48th out of the fifty states. West Virginia executives earn an average $334,593 annually. Of the $334,593 a mean of only 16.22% and a median 8.02% is paid in the form of incentive-based compensation. This differs significantly compared to the 35.47% paid on average (median 34.76%) by all states. A mean of only 14.96% (median 7.14%) is paid in the form of executive stock options. This likewise differs significantly from the mean 29.68% and median 25.52% paid firms of all states. 4.2. Results According to CEO Several studies suggest that the compensation packages given to CEOs matter most. We now partition all executives reported in the Standard and Poor’s ExecuComp for the period 1992 through 2006 according CEOs. The CEO results are reported in Table 2. The sample including all CEOs from all states is reported in Panel A. Panel B provides a comparison of CEO compensation across states. We find that of the 131,890 firm-year all executive observations reported in Table 1, 21,965 are reported on 4,856 CEOs from 2,777 firms. We illustrate in Panel A that CEOs on average earn approximately $4 million per year. Panel A further indicates that of this total $4 million paid to CEOs a mean 41.02% (43.04% median) is awarded using incentive-based pay. A mean 34.30% and a median 32.22% is provided using executive stock options. Thus, on average close to half of all CEOs compensation is awarded in the form of incentive-based compensation. Panel B of Table 2 ranks each state according to the proportion of incentive-based pay provided to the CEOs of the firms headquartered in the state. The results of panel B support our hypothesis. West Virginia ranks 49th out of the fifty states reported in our study. West Virginia CEOs earn an average $544,976 per year. A mean 18.09% (median 6.20%) of this total compensation is performance-based. This stands in stark comparison to the 41.02% (median 43.04%) paid to CEOs by firms headquartered in all states. A mean of only 17.01% and median 0.00% is paid in the form of executive stock options compared to 34.30% and 32.22% for all states in the sample. Both of these comparisons are significantly different. 4.2. Results According to Top Executives Excluding the CEO For completeness, we finally calculate our compensation statistics on the sample including only the top four ranking executives, excluding the CEO. CEO pay is often very large compared to the pay of other executives employed by a firm, thus there is potential for the compensation provided to the CEO to bias our all executive statistics. This sample is therefore constructed as the compliment of the CEO sample. We report that 109,925 firm-year observations are comprised of top level executives other than the CEO. The 109,925 observations include 28,629 different executives from 2,777 different firms. Results reported in Table 3 are consistent with the previous two samples. We illustrate for top executives excluding the CEO that firms headquartered in the state of West Virginia still rank low with respect to incentive-based compensation when compared to other states. Specifically, we show that West Virginia once again ranks 49th when ranked according to the mean percentage of total compensation provided in the form of incentive-based compensation. Firms headquartered in the state of West Virginia offer top executives excluding the CEO a mean 15.62% (8.05% median) of total compensation as incentive-based. This is significantly lower than the 34.36% (33.18% median) paid by firms headquartered in all states. West Virginia firms also only pay a mean 14.31% and a median 7.35% in the form of executive stock options. 4.3. Multivariate Analysis Preliminary support for hypotheses are supported by Institute for Public Affairs October 2008 7 the univariate analysis in Tables 1 through 3. However, the relationship might well be affected by other firm specific, industry specific or even geographic specific variables. For example, it is highly likely that some states have a concentration of firms within one particular industry that may offer more incentive pay. Perhaps there could be concentrations of small firms in other states that consider incentive-based pay too complicated for their business structure. We test the relation between incentive-based pay and states taking into account these other characteristics that explain executive compensation.7 These tests are critical since any policy recommendation should be made only in the context of the environment in which the firm operates. Formally, our regression model is: %PerfComp = a0 + B1 Size + B2 Book-to- Market + B3 Dividend Yield + B4 Prior Performance + B5 CEO + B6 Rural + B7 Urban + BI Industry + BJ Year + ε Our dependent variable, %PerfComp, is measured as the relative value of percentage compensation to total executive compensation. Size is the natural logarithm of total market capitalization at fiscal year end. Book-to-Market is measured as the total accounting book value at fiscal year-end divided by the total year end market capitalization. Dividend Yield is the annualized dividend yield. Prior Performance is the five-year dividend-reinvested return to firm shareholders. CEO is a dummy that equals 1 if the executive is the firm’s CEO. Rural and Urban are indicator variables based on whether the companies’ headquarters is located in or around a major metropolitan area.8 Industry is a vector of indicator variables representing the firms’ two-digit SIC code. There are a total of sixty-five industries represented in our sample. Finally, Year is a vector of indicator variables that control for temporal differences in %PerfComp.9 Since the error term in the regression is the portion of performance-based compensation unexplained by the model, we rank states according to the mean residuals recorded from the model and demonstrate that the ranking results are relatively unchanged. Thus, our inference from the univariate tests is not driven by other firm-specific characteristics. Table 4 presents the regression results. All signs of coefficients are 7 Tests were conducted using various other models. Particularly since our depen- dent variable is considered both bounded and limited we conduct a log transformation and also run a fractional logit regression model. Results were consistent with those reported in Table 4 and Table 5. West Virginia ranked 48th and 49th for each model re- spectively. Since results were unchanged we report the OLS results for interpretation. 8 To determine if a firm is rural or urban we use the 2000 census results from the United States Census Bureau. Rural firms are identified as firms headquarters more than 100 miles from one of the 50 largest metro areas named in the 2000 census. A firm is considered urban if it is headquartered in one of the 10 largest metro areas named in the 2000 census. 9 Year 1992 and two-digit SIC code 01 serve as the base year and industry, respectively. consistent with our expectation. Firms that are larger in size and have benefited from positive prior performance offer a higher proportion of incentive-based pay, while firms that pay earnings out in the form of dividends offer less incentive pay. Consistent with our univariate statistics we find that the CEO is likely to receive significantly more compensation in the form of performance-based pay. A very interesting result is the dummy coefficients for rural and urban firms. The coefficient for rural is negative and significant, indicating that firms in rural locations offer significantly lower percentages of incentive pay, while firms located close to major metropolitan areas offer significantly more incentive-based compensation. This result is important since it may indicate that perhaps the difference we previously identified has less to do with actual state policy and more to do with the number of large cities located within the state. Overall these regression results suggest that our univariate results should be suspect and that these other characteristics should be controlled for in the multivariate framework. Table 5 reports the mean residual state ranking. Consistent with our univariate tests, West Virginia still ranks 49th out of the fifty states reported in our study.10 Specifically, we find that on average West Virginia firms pay 8.32% in less performance-based pay than predicted. The fact that many states do alter rank considerably suggests that much of their difference in incentive pay can be explained by one or more of the control variables in the regression model. What is important, however, is that West Virginia still ranks the same even after controlling for other compensation determinants. Thus, the difference we observe is not due to a concentration of industry, small firm effects, or even geography; it is in fact state related. This is important since it suggests that there may be policies that can be implemented to increase incentive-based compensation and stimulate the state’s economic productivity. In the next section we discuss some of these policies. 5. Policy Recommendations Our study suggests that the discrepancy in incentive- based compensation in West Virginia compared to the rest of the nation is serious enough to warrant an investigation. As previously noted, we believe that this has economic implications for the state. This problem, however, can be mitigated, leading to better firm performance. Both the univariate and multivariate results in the previous section suggest that not only are these results robust but also the differences are state-specific, therefore implying that state policy is of extreme importance. Making a sound policy recommendation is, however, a daunting task. A sound recommendation must be technically feasible and persuasive enough to compel implementation. With regard to the findings of 10 Other regressions were run that included leverage variables such as Total Debt and Total Debt standardized by Total Market Capitalization. The results were largely unchanged. While a few of the states rank were slightly altered, West Virginia’s rank was still 49th in each of these models. Institute for Public Affairs October 2008 10 Table 2 Executive compensation for all CEOs reported in Standard and Poor’s ExecuComp database for the period 1992 through 2006. Panel A reports statistics for all CEOs of all states in the database. Panel B reports statistics according to state. States are ranked according to the mean percentage of performance based compensation. Performance based compensation is defined as the Black-Scholes-Merton option value granted plus the value of restricted stock granted. “% performance comp” is then measured as the relative value of percentage compensation to total executive compensation. “Total executive comp” is defined as the sum of salary, bonus, Black-Scholes-Merton option value, and restricted stock. “% option based comp” is the percent of total compensation represented by the Black-Scholes-Merton option value granted. State % Performance Compensation % Option Based Compensation Total Executive Compensation Number of ObsMean Median Mean Median Mean Median Panel A: CEO Compensation Summary Statistics for All States All States 41.02 (43.04) 34.30 (32.22) 3,990,240 (1,811,090) 21,965 Panel B: CEO Compensation Summary Statistics by State MA 49.95 (57.32) 41.77 (43.03) 3,659,780 (2,138,840) 663 CA 47.53 (52.72) 43.27 (45.95) 5,714,820 (2,125,070) 2,376 AZ 46.37 (51.27) 39.95 (44.17) 2,669,130 (1,661,340) 172 CT 43.99 (47.87) 35.82 (37.11) 5,053,000 (2,439,880) 414 CO 43.98 (52.27) 36.75 (32.58) 4,413,790 (1,895,550) 210 MT 43.80 (48.97) 41.32 (48.97) 1,285,270 (1,021,250) 14 DE 43.64 (46.56) 33.78 (34.02) 3,599,220 (2,249,460) 71 IL 43.53 (45.54) 35.09 (34.95) 3,515,710 (2,160,400) 980 TX 43.38 (47.47) 34.84 (33.61) 3,902,870 (2,026,770) 1,477 NJ 43.06 (49.78) 34.82 (34.82) 4,071,180 (2,299,000) 561 ME 42.96 (50.18) 36.97 (40.17) 3,378,250 (1,830,890) 49 NY 42.11 (45.24) 33.77 (31.33) 7,853,980 (3,059,290) 1,366 TN 41.86 (45.83) 34.54 (35.26) 2,753,160 (1,904,520) 288 MD 41.57 (45.03) 38.17 (39.76) 5,083,690 (2,698,760) 190 PA 41.09 (44.33) 33.70 (33.68) 3,509,860 (2,114,490) 747 MN 40.85 (41.37) 34.07 (34.29) 3,442,940 (1,753,880) 587 LA 40.67 (44.57) 31.48 (25.07) 3,442,350 (1,781,290) 118 UT 40.33 (41.99) 36.01 (39.16) 1,485,210 (960,372) 69 VA 39.97 (41.74) 31.33 (28.97) 4,547,090 (1,866,130) 378 KS 39.86 (47.52) 29.43 (24.62) 2,852,360 (1,949,220) 53 KY 39.80 (36.24) 32.29 (29.54) 4,542,490 (2,746,490) 98 OR 39.18 (39.70) 36.36 (38.10) 2,360,390 (1,652,080) 177 MS 39.18 (45.80) 24.67 (15.85) 1,138,090 (814,538) 28 AL 38.92 (42.27) 28.95 (24.64) 3,136,000 (1,819,830) 138 MI 38.86 (38.95) 31.23 (28.94) 3,403,060 (1,972,070) 430 ID 38.47 (35.62) 32.33 (29.84) 2,267,530 (1,207,380) 61 FL 38.41 (41.07) 30.44 (28.94) 3,389,330 (1,961,120) 469 RI 38.32 (35.10) 28.80 (25.40) 4,324,400 (2,574,960) 57 OH 38.31 (40.78) 30.33 (30.62) 3,378,100 (1,966,340) 894 WI 38.11 (41.57) 32.41 (32.65) 2,411,350 (1,693,590) 361 GA 37.37 (36.75) 27.71 (19.55) 3,565,800 (1,900,380) 444 AR 36.41 (38.02) 30.82 (28.67) 3,317,140 (1,370,580) 127 NC 36.28 (39.39) 26.14 (17.92) 3,841,500 (1,681,660) 358 NM 36.16 (38.66) 31.51 (31.45) 1,142,390 (933,162) 12 AK 36.06 (25.54) 35.34 (22.92) 878,419 (465,153) 10 MO 34.81 (34.45) 27.53 (18.53) 2,876,570 (1,438,670) 385 OK 34.60 (38.53) 26.05 (26.17) 2,024,850 (1,357,980) 107 NE 33.56 (33.02) 21.88 (0.00) 3,189,410 (1,431,990) 91 IN 33.20 (31.15) 25.88 (17.31) 3,747,110 (1,384,200) 225 HI 32.53 (30.19) 24.65 (21.21) 622,238 (1,500,000) 41 NV 32.36 (30.69) 27.27 (18.75) 2,985,790 (1,283,050) 109 WA 32.11 (30.79) 28.30 (23.49) 2,126,790 (1,091,760) 274 IA 30.86 (31.12) 22.87 (17.12) 2,258,810 (1,139,410) 104 NH 29.30 (24.25) 19.99 (13.30) 1,457,600 (895,645) 42 SC 28.20 (27.38) 25.18 (23.44) 1,707,350 (1,360,570) 110 SD 27.79 (23.68) 15.49 (16.29) 1,098,340 (911,195) 21 DC 25.75 (21.43) 19.74 (0.00) 2,982,040 (1,426,920) 79 VT 18.63 (17.05) 14.35 (9.64) 591,761 (391,094) 33 WV 18.09 (6.20) 17.01 (0.00) 544,976 (418,001) 97 ND 13.17 (0.00) 7.10 (0.00) 1,578,380 (1,796,310) 7 Institute for Public Affairs October 2008 11 Table 3 Executive compensation for non-CEO executives reported in Standard and Poor’s ExecuComp database for the period 1992 through 2006. Panel A reports statistics for all CEOs of all states in the database. Panel B reports statistics according to state. States are ranked according to the mean percentage of performance based compensation. Performance based compensation is defined as the Black-Scholes-Merton option value granted plus the value of restricted stock granted. “% performance comp” is then measured as the relative value of percentage compensation to total executive compensation. “Total executive comp” is defined as the sum of salary, bonus, Black-Scholes-Merton option value, and restricted stock. “% option based comp” is the percent of total compensation represented by the Black-Scholes-Merton option value granted. State % Performance Compensation % Option Based Compensation Total Executive Compensation Number of Obs Mean Median Mean Median Mean Median Panel A: Non-CEO Compensation Summary Statistics for All States All States 34.36 (33.18) 28.76 (24.41) 1,374,310 (668,086) 109,925 Panel B: Non-CEO Compensation Summary Statistics by State MA 43.51 (46.68) 36.48 (35.68) 1,340,820 (730,150) 3,253 CA 41.17 (42.26) 37.49 (35.86) 1,767,760 (804,609) 11,902 NJ 38.58 (40.99) 31.25 (28.41) 1,462,710 (854,463) 2,687 CO 38.55 (39.43) 31.20 (26.03) 1,446,150 (709,501) 1,064 AZ 38.44 (40.49) 34.08 (32.00) 911,133 (625,995) 878 TX 37.90 (39.04) 30.65 (28.25) 1,342,480 (747,989) 7,126 MD 37.49 (38.37) 33.13 (31.72) 1,869,420 (1,110,580) 977 NY 35.82 (35.95) 28.50 (25.58) 3,096,110 (1,095,610) 6,602 RI 35.82 (35.14) 28.66 (27.63) 1,528,380 (1,118,370) 251 DE 35.68 (33.34) 28.89 (25.51) 1,344,120 (818,206) 319 MT 35.38 (36.20) 33.21 (33.21) 381,032 (323,275) 57 OR 35.32 (33.30) 32.41 (30.47) 907,939 (597,884) 892 KY 35.18 (31.63) 28.10 (22.53) 1,433,070 (1,005,110) 472 MN 34.99 (34.15) 29.49 (26.60) 1,107,050 (658,578) 2,796 TN 34.87 (36.56) 28.64 (25.64) 958,109 (676,526) 1,423 CT 34.81 (33.12) 28.58 (26.14) 1,538,530 (760,164) 1,945 ME 34.44 (40.08) 31.53 (35.23) 1,016,040 (617,452) 231 PA 34.01 (34.11) 28.00 (26.30) 1,232,680 (710,772) 3,638 IL 33.85 (33.10) 26.52 (22.85) 1,232,600 (785,458) 4,852 FL 33.38 (32.23) 26.99 (22.21) 1,232,390 (752,129) 2,349 MI 32.92 (32.26) 25.24 (21.84) 1,322,510 (780,013) 2,006 VA 32.72 (30.51) 25.87 (21.21) 1,436,120 (750,023) 1,884 GA 32.47 (28.89) 23.71 (15.94) 1,253,770 (719,239) 2,229 OH 31.68 (31.69) 24.88 (22.81) 1,084,420 (738,731) 4,217 AR 31.50 (32.60) 26.51 (24.00) 1,403,500 (636,494) 598 WI 31.22 (32.65) 26.59 (25.72) 872,406 (598,285) 1,672 NV 30.27 (30.24) 24.84 (14.02) 1,275,390 (539,123) 519 AL 30.19 (31.06) 22.64 (20.75) 817,771 (585,208) 703 KS 30.17 (25.77) 19.06 (0.00) 1,026,200 (585,367) 298 UT 30.10 (31.58) 27.19 (27.79) 580,283 (426,428) 350 LA 30.07 (28.62) 24.42 (15.33) 1,060,800 (594,250) 578 NC 29.52 (28.67) 21.40 (12.66) 1,409,090 (649,177) 1,799 OK 29.16 (29.40) 22.25 (17.77) 844,191 (447,140) 524 ID 28.09 (24.85) 23.62 (18.14) 1,044,170 (388,222) 326 MO 27.80 (22.21) 22.59 (12.71) 966,306 (551,275) 1,930 NE 27.66 (24.22) 20.62 (8.96) 1,069,500 (567,013) 418 IN 27.56 (23.33) 20.61 (13.24) 1,222,480 (549,161) 1,158 WA 27.50 (21.76) 24.48 (19.55) 1,692,820 (506,000) 1,391 MS 26.47 (19.45) 15.67 (0.00) 482,065 (322,500) 133 DC 26.40 (19.26) 21.72 (10.67) 1,342,180 (623,003) 393 HI 25.94 (21.81) 18.86 (15.42) 688,030 (559,897) 204 NH 24.12 (20.78) 17.46 (8.83) 628,070 (364,890) 216 IA 23.67 (22.53) 18.73 (15.62) 684,435 (455,775) 485 SC 23.63 (23.03) 20.85 (17.61) 628,826 (485,701) 535 NM 22.90 (26.59) 20.38 (23.04) 346,901 (327,484) 60 AK 21.52 (0.00) 21.52 (0.00) 441,426 (264,947) 54 SD 16.34 (13.80) 8.84 (0.00) 471,995 (396,038) 109 ND 15.71 (0.00) 9.25 (0.00) 541,374 (488,364) 51 WV 15.62 (8.05) 14.31 (7.35) 269,394 (219,525) 313 VT 10.80 (6.43) 8.38 (1.70) 258,672 (205,403) 168 Institute for Public Affairs October 2008 12 Table 4 This table provides the cross-sectional regression with percentage of performance-based compensation as the dependent variable. OLS regression residuals are used to determine the state performance- based rank in the multivariate framework. Size is measured as the log of total market capitalization at fiscal year-end. Book-to-market is the ratio of book value at fiscal year-end to total market capitalization. Dividend Yield is annualized dividend yield. Prior Performance is the five-year dividend-reinvested return to firm shareholders. CEO is a dummy variable equal to one if any executive is the CEO of the firm. Rural and Urban are dummy variables indicating whether the firm is headquartered in a rural or urban area, respectively. Industry is a vector of indicator variables representing the firms’ two-digit SIC code. Finally, Year is a vector of indicator variables that control for temporal differences in the percentage of performance-based compensation. Year 1992 and two-digit SIC code 01 serve as the base year and industry, respectively. Absolute values of t statistics are in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%. Variable Anticipated Sign Parameter Estimates t Value Constant -0.223*** 11.28 Size + 0.053*** 86.64 Book-to- Market +/- -0.009 0.83 Dividend Yield - -0.003*** 11.85 Prior Performance + 0.001*** 8.55 CEO + 0.054*** 23.69 Rural - -0.02*** 4.20 Urban + 0.014*** 7.31 Industry Yes Year Yes Observations 82,463 Adjusted R-squared 0.21 Table 5 Executive compensation for all executives reported in Standard and Poor’s ExecuComp database for the period 1992 through 2006. Panel A reports statistics for all executives of all states in the database. Panel B reports statistics according to state. States are ranked according to the mean residual of regression (1). The residual represents the portion of performance-based compensation not explained by the model. Performance based compensation is defined as the Black-Scholes-Merton option value granted plus the value of restricted stock granted. “% performance comp” is then measured as the relative value of percentage compensation to total executive compensation. “Total executive comp” is defined as the sum of salary, bonus, Black-Scholes-Merton option value, and restricted stock. “% option based comp” is the percent of total compensation represented by the Black-Scholes-Merton option value granted. State Mean Regression Residual Median Regression Residual Number of Observations Panel A: Regression Residuals for All States All States 0.00% 0.27% 82,463 Panel B: Regression Residuals by State NM 10.72% 14.25% 72 SD 6.57% 4.43% 130 AZ 5.24% 9.93% 864 MA 5.16% 9.33% 3,232 ME 4.87% 5.30% 226 OR 4.10% 4.32% 982 NV 3.72% 7.19% 525 HI 3.45% 2.84% 244 CA 3.45% 6.40% 11,738 NH 3.25% -1.37% 248 KS 2.92% -5.48% 286 ID 2.19% 1.55% 322 CO 2.16% 4.51% 1,020 UT 2.16% 1.74% 367 MS 2.11% 4.30% 121 TN 2.08% 2.51% 1,493 MT 1.81% -13.65% 52 CT 1.45% 1.37% 1,927 RI 1.09% -0.31% 288 MN 0.8% 0.95% 3,042 NJ 0.68% 2.18% 2,890 TX 0.49% 1.44% 7,434 VT 0.41% 2.17% 116 MI 0.19% -0.24% 1,961 LA -0.02% -1.43% 617 OK -0.14% 0.73% 608 MD -0.19% -0.53% 1,001 AL -0.37% 0.53% 702 PA -0.58% 0.14% 4,159 DE -1.25% -4.72% 383 IA -1.38% -4.24% 531 NC -1.53% -2.31% 1,927 FL -1.58% -1.52% 2,254 IL -1.67% -0.99% 5,115 NY -1.68% -1.72% 7,142 OH -1.79% -2.04% 4,680 WI -1.94% -0.88% 1,923 SC -1.98% -3.95% 600 VA -2.06% -0.50% 1,911 AR -2.11% -0.33% 662 AK -2.47% -19.85% 64 KY -2.63 -7.85 431 GA -3.10% -5.34% 2,212 NE -3.34% -6.41% 472 MO -4.00% -5.34% 2,014 IN -4.63% -6.79% 1,226 WA -6.32 -7.10% 1,515 ND -7.02% -21.31% 58 WV -8.32% -8.28% 51 DC -11.38% -12.29% 462 Institute for Public Affairs October 2008 15 ABOUT THE COVER by Robert Bridges William Robinson Leigh a native of Berkeley County West Virginia, was educated at the Maryland Institute of Art and spent twelve years studying at the Royal Academy in Munich, Germany. Today he is primarily known as a painter of the American West and along with Charles Russell and Frederick Remington was one of the best of that genre. In his early years Leigh made his career as an illustrator in New York City. The cover image is one of several original drawings used to illustrate the 1905 novel The Golden Flood, by Edwin Lefevre. Lefevre produced a series of works set in Wall Street. The Golden Flood is about a man that appears to have developed a method for manufacturing gold. The influential Wall Street heads fear this may wreak havoc with the financial world. Robert Bridges is an Assistant Professor of Art at West Virginia University and Curator of the West Virginia University Art Collec- tion, which houses more than 2,500 art objects. In addition, he serves as Curator of the Mesaros Galleries. Local Government Leadership Academy Details Coming Soon! Please check http://www.polsci.wvu.edu/IPA/academy.html for current information. Nonprofit Org. U.S. Postage PAID Morgantown, WV Permit No. 34 The West Virginia Public Affairs Reporter Institute for Public Affairs Eberly College of Arts and Sciences West Virginia University PO Box 6317 Morgantown, WV 26506-6317 Change Service Requested Open for Business? An Examination of Incentive-Based Executive Compensation in West Virginia Agency theory suggests that tying an executive’s compensation to the performance of the firm will align the interests of managers and shareholders. Extant literature shows both theoretically and empirically that agency problems are mitigated and performance is higher for firms that offer incentive-based compensation packages. West Virginia currently lags in most fundamental economic indicators. This is a telling indication of poor productivity and a deficient business climate. After Governor Joe Manchin was inaugurated, one of the primary focuses of his administration was a reformation of the dismal economic climate in West Virginia, a task ambitiously titled: West Virginia, Open for Business. Undoubtedly, many significant structural impediments remain in the way of this reformation. Studies addressing these issues have been confined primarily to analyses of the regulatory compliance and administrative costs of doing business in the state. We, however, illuminate a problem with executive compensation in West Virginia, an issue that has not yet been examined. Specifically, we find that firms headquartered in the state fail to compensate their executives with incentive-based pay comparable to firms in other states. We argue that these firms should consider offering more incentive-based compensation packages to their executives and other employees. There are two suggested courses of action that might be pursued by West Virginia policy makers. First, an education initiative should be pursued. In addition, the state government should consider implementing tax benefits to entice firms to offer more incentive-based pay to their executives, a system that the federal government and many other states already maintain. While we do not claim that increasing incentive-based pay will be a panacea for all the economic challenges facing the state, the results of this study suggest that these changes may very well lead to marginal increases in economic conditions. In this Issue of: The West Virginia Public Affairs Reporter