WP/JCH Testimony-fns 5 & 7 2009 TX Rate Case Page 16 of 25

482

Journal of Financial and Quantitative Analysis

adjust their beliefs accordingly. In equilibrium, while the manager overinvests in the long term, she does not gain from it.24 From equation ( A-1) in the Appendix, it can be seen that the equilibrium investment is determined by the ratio of the previously acquired shares to the number of shares received at date 1.25 This gives a measure of the inefficiency that is likely to occur in practice if the form of compensation is limited to stock. Lambert and Larcker ( 1987) finds that the median ratio of the value of stock ownership to annual cash compensation is 4.4 for Forbes 500 chief executive officers. They cite research that indicates that the cash compensation is 80-90 percent of total compensation. In this case, assume that the median ratio of the value of stock ownership to annual total compensation is approximately 4. This implies that the first-order condition is µ*lf'(ds) < 0.8, which means that the manager will accept projects that lose 20 cents for every dollar invested.

C.

Cash and Stock Contracts

The above proposition suggests that a compensation contract that is part cash and part stock might result in efficient investment. To show this, I use the following procedure. Suppose investors believe that the manager is using the firstbest decision rule, that is, b = d*. I consider the compensation contract where wt = aµd* (d), t = 1, 2, which obviously satisfies the wage constraint (7), and then show that there exists a sequence of {A,} such that the incentive compatibility constraint (8) is satisfied. The following proposition characterizes the optimal compensation contract with a mixture of cash and stock. Proposition 3. There exists a compensation package {w*, \,*} such that the manager's equilibrium decision rule is d*. The contract is defined by i) The market value of the compensation is given by w, = aµd+(d), t = 1, 2. ii) All the date 2 compensation is in stock, that is, A2 = 1. iii) The fraction of the compensation in stock at date 1 is given by B hK (1 - No) (12)

Ai (Yi)

=

B

1

h1K +2K-a+f(d* (yi)) iv) The form of compensation at date 0 is irrelevant. For a given a, the expression for a* can be interpreted as follows. Note that B/ht is the change in the date 1 perceived ability for a marginal change in the level 24A similar result is obtained by Bebchuk and Stole (1993) who show that, if the compensation contract is a linear function of both short-term and long-term stock prices, the manager has an incentive to overinvest in the long-term project if the productivity is unobservable to investors. In their model, if the unobservable variable is investment (instead of productivity), there is underinvestment in the longterm project. By contrast, in my model whether there is under- or overinvestment in the long-term project depends on whether the compensation is cash or stock. 25Equation (A-1) in the Appendix shows that the equilibrium investment decision under all-stock compensation also depends on a factor that is a function of a, K, 0, and hi. This factor signifies that any reduction in the manager's current wage due to overinvestment increases the value of her previously acquired shares. This represents an additional benefit to overinvestment. This benefit is a function of the misperception of the manager's ability (measured by K, 0, and hl) and the wage factor a.

2009 ETI Rate Case

WP-1-71

4951

WP/JCH Testimony-fns 5 & 7 2009 TX Rate Case

Page 17 of 25

Narayanan

483

of date 1 investment d (equation (5)). Therefore, B/hl is the change in the cash compensation for a marginal change in the level of investment. The second term in the denominator of equation ( 12) is the number of date I shares received by the manager. In equilibrium, the change in the value of the firm for a marginal change in date 1 investment is 1. Therefore, the second term in the denominator of equation ( 12) represents the change in the stock compensation for a marginal change in the investment. Thus, a* is the ratio of the change in cash compensation to the sum of the changes in cash and stock compensation for a marginal change in the level of date 1 investment. There are several points worth noting here. First, 0 1.

It can be verified that the second-order condition is satisfied. From the concavity of f(d), it follows that d, is less than d*. Therefore, there is underinvestment in the long-term project. o Pro of of Proposition 2. As in the proof of the previous proposition, let 7r, ( {wr,1 }, b, d) denote the manager's total expected payoff at date 1 under a stock-only contract if her decision is d and investors' belief is b. Then 7r, ({wr, 1} b, d)

2009 ETI Rate Case

=

2 E, E Ur Qwr, 1} , y`, b, d) , r=0

WP_1-76

4956

WP/JCH Testimony-fns 5 & 7 2009 TX Rate Case Page 22 of 25

Journal of Financial and Quantitative Analysis

488

where u, is the true value of the date t package at date 1. Note that uo is the value of the stock portfolio that the manager received as date 0 compensation and its value depends on her current decision. The true value of the date 2 compensation package is given by u2 ({w2, 1},y2 ,

b,d)

_

a bd Vb

sb

=

aµz,

using equations (10a) and ( 11 a). The true value of the date 1 compensation package is given by ^ u, ({wj,l},y ,b,d)

=

2K-a+f(d) *

cxµb(d) ^(d) vd(d,l)

=

a2K-a+f(b)µ^'

using equations ( 10b) and ( l lc). Since the manager already owns No shares from her date 0 compensation, she receives a dividend at date 1 equal to No(y', -d - w]), where y' is the gross output at date 1, which the manager has already observed. In addition, these No shares have a true value of Novb(d, 1). Since w, = aµb(d ) from the wage constraint, the true value of the date 0 compensation package is given by uo

=

No [yi - d - aµb(d ) + vb(d,1)] .

Note that uo and ul are deterministic since the manager already knows the gross output. uZ is a random variable that does not depend on d. Therefore, one can effectively ignore the expectation operator in calculating the first-order condition. Equations (5), (lOb), and (llc) show that 8µb (d)

0

=

hjK'

ad 8vb(d, 1) ad

a^(d) ad

and

{2K - a +f(b)} 0 hjK

=

Therefore, it follows that ad - No (-l + he + µi.f'(d))

8ul 8d

=

, a d ); 2K - a+ f(b) µ^*f (

19U2 and, ad = 0.

The manager chooses d to satisfy the first-order condition 87r, /8d = 0. Let ds be the solution to the above equation. Then, a7ry

=

ad

aµjf'(d) No I -1 + ae +µi.f'(d)^ + 2K - a+f(b) h, \\\

= 0.

In equilibrium, b = d = d, and the first-order condition reduces to No^1-h^K/

]Vo( l-h^K^ (A-1)

2009 ETI Rate Case

µif' (ds)

=

a No+2K-a+f(ds)

=

\No+1Vj


d*. 0 Proof of Proposition 3. Consider a compensation contract {wj, At}, where wt _ aµd* (d) , t= 1, 2. This clearly satisfies constraint (7). For this contract to induce the first-best decision in equilibrium, it needs to be shown that there exist X, such that 0 < X< < 1, t = 0, 1, 2, and the contract satisfies the incentive compatibility condition ( 8) when investors believe that the manager has chosen decision d*. As before, let ir, ( {w,, a,}, b, d) represent the manager's total expected payoff at date I by choosing decision d when investors' belief is b, given the compensation contract {wt, A,}. As in the proof of Proposition 2, z aI({wi,,\t},b,d)

= EiY:ui({wt,Ai},yt,b,d),

where

t=O uo = No [yi - d- aµb(d) + vb (d, az), , U1

b = A I aµi(d)vb(d,1\x)+(1-ai)aµi(d) 4(d) b

+(1-ai) µb(d) , a [2KVl ad+f(b) b b b u2 = A2aFlx(d)vz(d)+(1 - az)aµi(d) sz(d)

using equation (10b),

= a[µZ+(1-az)h SbKd+yz(d)1K+f(b)

K+f(d)/^J

using equations (6), ( 10a), and (I la). Therefore,

Eiuz = a[µ^ +(1 -.^z) hz S b Kd + K f(b)(f(d) -f(b))

IJ •

Note that only the value of the stock component of the date 0 wage is relevant to the manager's decision at date 1. That is why the cash component is omitted in defining uo. For ul and u2, both components are important. The expression for El u2 may be interpreted as follows. The first term is the symmetric information expected wage of the manager and the second term is the distortion caused by the cash portion of the date 2 wage due to misvaluation of ability. The stock portion of the date 2 wage does not cause any distortion since the misvaluation of ability is completely offset by the misvaluation of the stock and since date 3 cash flows are independent of the manager's decision. Such distortion in date 2 compensation due to misvaluation of ability can be eliminated by setting a2 = 1, i.e., by paying all the compensation in stock. This, in turn, eliminates distortions to v6(d, A2), the date 1 true value of the firm (see equations (11b) and (l lc)).

2009 ETI Rate Case

wP_1-78

4958

WP/JCH Testimony-fns 5 & 7 2009 TX Rate Case Page 24 of 25

Journal of Financial and Quantitative Analysis

490

Set A2 = I and differentiate uo, ut, and Et u2 with respect to d to obtain (using the first derivatives calculated for the proof of Proposition 2),

ad aul 7d 8Et u2

ad Therefore,

d

iK h +µif^(d)) - No(-1+ _

a r Alµif'(d) 2K

h1 K

_ 0 -

No (-1 + aO +

+a

Atµift (d) -(1 -al)h1K B . [2K-a+f(b)

To derive the optimal contract, calculate 87rt /8d at b = d = d* and set it equal to zero ( manager's first-order condition). It can be verified that the second-order condition is satisfied. Recall that µ*1f'(d*) = 1,

ad I e=n=d•

=

-(1-No)hK+At IhtK+2K-a+f(d*), = 0. L

Solving the above equation for a* results in (1 -No)h1K

(A-2)

e

q

h1K+2K-a+f(d*) Proofs of Propositions 4, 5, 6, and 7. The proofs follow directly from equation (A-2). q

References Aisenbrey, E. "Long-Term Incentive for Management, Part 3: Restricted Stock" Compensation and Benefits Review, 21 (Issue 6, 1989), 34-46.

Arreglado, E. "Top Executive Compensation: 1992 Edition." Report Number 1016, Conference Board (1992),. Bebchuk, L., and L. Stole. "Do Short-Term Objectives Lead to Under- or Overinvestment in Long-Term Projects?" Journal of Finance, 48 (1993), 719-730. Bhagat, S.; Brickley, J.; and R. Lease. "Me Impact of Long-Range Managerial Compensation Plans on Shareholder Wealth." Journal of Accounting and Economics, 7 (1985), 115-129. Bizjak, J.; J. Brickley; and J. Coles. "Stock-Based Incentive Compensation and Investment Behavior." Journal of Accounting and Economics, 16 (1993), 349-372. Bushman, R., and R. Indjejikian. "Accounting Income, Stock Price, and Managerial Compensation." Journal of Accounting and Economics, 16 (1993), 3-23.

Crystal, G. "Incentive Pay that Does Not Work:" Fortune, 120 (Aug. 28, 1989), 101-104. Dechow, P., and R. Sloan. "Executive Incentives and the Horizon Problem." Journal of Accounting and Economics, 14 (1991), 51-89.

DeFusco, R.; R. Johnson; and T. Zorn. "The Effect of Executive Stock Option Plans on Stockholders and Bondholders." Journal of Finance, 45 (1990), 617-627. DeGroot, M. Optimal Statistical Decisions. New York, NY: McGraw-Hill (1970).

2009 ETI Rate Case

WP_1-79

4959

WP/JCH Testimony-fns 5 & 7 2009 TX Rate Case Page 25 of 25

Narayanan

491

Gaver, J., and K. Gaver. "Investment Opportunity Set and Corporate Policies " Journal of Accounting and Economics, 16 (1993), 125-160. Gibbons, R.; and K. Murphy. "Optimal Incentive Contracts in the Presence of Career Concerns: Theory and Evidence." Journal of Political Economy, 100 (1992), 468-505. Hagerty, K.; A. Ofer; and D. Siegel. "Managerial Compensation and Incentives to Engage in FarSighted Behavior." Working Paper, Northwestern Univ. (1993). Healy, P. "The Effect of Bonus Schemes on Accounting Decisions." Journal of Accounting and Economics, 7 (1985), 85-107.

Hirshleifer, D., and T. Chordia. "Resolution Preference and Project Choice." Working Paper, Univ. of California, Los Angeles (1991). Holthausen, R.; D. Larcker; and R. Sloan. "Annual Bonus Schemes and Manipulation of Earnings." Journal of Accounting and Economics, 19 (1995), 29-74. Jacobs, M. Short-Term America: The Causesand Cures ofOur Business Myopia. Boston, MA: Harvard Business School Press (1991).

Jensen, M., and K. Murphy. "Performance Pay and Top Management Incentives" Journal of Political Economy, 98 (1990), 225-264 Kim, 0., and Y. Suh. "Incentive Efficiency of Compensation Based on Accounting and Market Performance" Journal of Accounting and Economics, 16 (1993), 25-53. Lambert, R., and D. Larcker. "An Analysis of the Use of Accounting and Market Measures of Performance in Executive Compensation Contracts." Journal of Accounting Research, 25 Supplement (1987), 85-125. "Executive Compensation, Corporate Decision Making and Shareholder Wealth: A Review of the Evidence." In Executive Compensation, G. Crystal, ed. Boston, MA: Harvard Business School Press (1991). Lewellen, W.; C. Loderer; and K. Martin. "Executive Compensation and Executive Incentive Problems: An Empirical Analysis." Journal of Accounting and Economics, 9 (1987), 287-3 10.

Narayanan, M. P. "Observability and the Payback Criterion." Journal ofBusiness, 58 (1985a), 309-323. "Managerial Incentives for Short-Term Results." Journal of Finance, 40 (1985b), 1469-1484. Paul, J. "On the Efficiency of Stock-Based Compensation." Review of Financial Studies, 5 (1992), 471-502. ."Managerial Myopia and Observability of Future Cash Flows." Working Paper, Univ. of Michigan (1993). Sloan, R. "Accounting Earnings and Top Executive Compensation." Journal of Accounting and Economics, 16 (1993), 55-100. Smith, C., and R. Watts. "The Investment Opportunity Set and Corporate Financing, Dividend, and Compensation Policies." Journal of Financial Economics, 32 (1992), 263-292. Stein, J. "Takeover Threats and Managerial Myopia" Journal of Political Economy, 96 (1988), 61-80. ."Efficient Capital Markets, Inefficient Firms: A Model of Myopic Corporate Behavior." Quarterly Journal of Economics, 104 (1989), 655-670.

Tehranian, H., and J. Waegelein. "Short-Term Plans and Stock Prices." Journal of Accounting and Economics, 7 (1985), 131-144.

Thakor, A. "Investment Myopia and the Internal Organization of Capital Allocation Decisions." Journal of Law, Economics, and Organization, 61 (1990), 129-154.

2009 ETI Rate Case

WP-1-80

4960

This page has been intentionally left blank.

2009 ETI Rate Case

wP_1-81

4961

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 1 of 17

Performance Pay and Productivity By EDWARD P. LAMAR* Much of the theory in personnel economics relates to effects of monetary incentives on output, but the theory was untested because appropriate data were unavailable. A new data set for the Syftltte Glass Corporation tests the predictions that averAge productivity will rise, the firm will attract a more able work,brce, and variance in output across individuals at the firm will rise when it shifts to piece rates. In ^''efite, productivity effects amount to a 44-percent increase in output per worker, This firm apparently had selected a suboptimal compensation system, as profits also increased with the cllut€,qe. (JEL 7OU, J22, 13)

A cornerstone of the theory in personnel economics is that workers respond to incentives. Specifically, it is a given that paying on the basis of output will induce workers to supply more output. Many sophisticated models have been offered, but they have gm largely untested because of a la& of data Of 0, and al", X)/87C2 > 0 all hold.

II. Data Safelite Glass Corporation is located in Columbus, Ohio, and is the country's largest installer of automobile glass. In 1994, 4afelite, under the direction of CEO Garen Sta& and President John Badow, implemented a new compensation scheme for the auto glass installers- Until January 1994, glass installers were paid an hourly wage ratt3, which did not vary in any direct way with the number of windows to were instaEied. During 1994 and 1995, installers were shifted from an hourly wage schedule to performance pay--wcifically, to a piece-rate schedttle. Rather than being paid for the number of hours that they worked, installers were paid for the number of glass units that they installed. The rates varied somewhat. On average installers were paid about $20 per unit installed. At the time that the piece rates were instituted, the workers were also given a guarantee of approximately $11 per hour. If their weekly pay came out to less than the guarants:e, they would be paid the guaranteed amount. Many workers ended up in the guarantee range. Staglin and Barlow changed the compensation scheme because they felt that productivity was below where it should have been. Productivity could have been raised by requiring a higher minimum level of output under a timerate system. If all workers bad identical preferenees, this would have worked well. Given differences in work preferences, a uniform in'o The condition that some workers continue to opt for the guaranteed wage is not superfluous. If all workers opt for the piece rate, then it is possible that even very low ability workers who did not work before now work for the firm. Their addition could actually result in a lowering of average ability.

4966 !

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 6 of 17 VOL. 90 NO. S

1351

L4ZEAR: PERFORMANCE PAY AND PRODUCTIUN

TAKE 1--DATA DMCUMON

PPP dummy

Base pay

Units-per-worker-per-dsy Regular hours

MM

Definition

Variable

Standard dovislift

A dummy variable equal to 1 if the worker Is on PPP during that mouth Hourly wage

0.53

$11.48

$2.94

Average number of units of glass installed by the given worker

2.98

1.53

153

41

19 52,254

19 $882

$107

$36

$40

$62

4.62

0.29

during the month in question Regular hours worked during the

given month Overtime hours Pay

Overtime hours worked that mouth pay actually received in a given

pay-per-day

Actual pay per eight hours worked; this differs from PPP pay in that the wage guarantee and other

cost-per-unit

Actual pay fof a given w41litOC,

month

payments are included in the total

Log of pay-per-day Separation dummy

divided by the number of urtits installed by that worker in a given mouth l.og of actual pay per eight hours worked A4mmy equal to l if the

0;047

employee quit during this month Noteea: There were 2.755 individuals who worked as installers over the 19-mnnOt period covered by the data. The unit of analysis is a person-month.There; . are 29;837 person-months of good ciata. Pay-pardajr is calculated only for workers whose. total hours in a month exceeded 10 and cost-per-unit only for workers whose nnawhly units installed exceeded 3.

crease in required output, coupled with a wage increase, would not be received in the same way by all workers. In pAr"ar, the iower-output workers would find this more burdensome dm the higher-output worlcers. In order to avoid massive turnover, the firm adopted a piece-rate schedule, which allowed those who wanted to work more to earn more, but also allowed those who would accept lower pay to put forth less effort. Safelite has a very sophisticated computerized information system, which keeps track of how many units of each kind each installer in the company installs in a given week. Safelite provided monthly data. Since PPP (Performance Pay Plan) was phased in over a 19-month period, many workers were employed under both regimes. Thus, data on individual output are available for most installers both during the hourly wage period and during the PPP period. This before-and-after comparison with personspecific data provides a very clean body of information on which to base an analysis of performance pay incentives.

Some basic characteristics of the sample are reported in Table 1. The data we organized as follows. Each month provides an independent unit of observation. Them are 38,764 personmonths of data covering a 19-month period. Over the 19-month period, there was a total of 3,707 different individuals who worked for Safelite as installers. The number of "good" observations is 29,837 when partial months and observations with incomplete data are dropped from the data set. There we a number of possible productivity measures. The one that most Safs}ite managers look to is units-per-wotker-per-day. This is the total number of glass units per eight-hour day that are installed by a given worker. The unitsper-worker-per-day number for each individual observation relates to a given worker in a given month. Thus, units-per-worker-per-day is the average number of units per eight -hour period installed by the given worker during the given month. The average number of glass units installed per day over the entire period is 2.98, with a

4967 1

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 7 of 17

1352

DECEMBER 20M

THE AMERICAN ECONOMIC REVfE1i' TAHtE 2r--MBA'N ANU SrA=AAt) DEVIATIONS OF KEY VARIABLES BY PAY S17tucruxH

Number of observations Variable Units-. per-worker-per-day Actual pay PPP pay Cost-per-unit

Hourly wages

piece rates

13,106

15.246

Mean

Standard deviation

Mean

Hourly wars

2.70 $2,228 $1,587 $44.43

1.42 $794 S923 Z7535

3.24 $2,283 $1,852 $35.24

1.59 $950 $997 S49.00 '

Note. 1.485 observations were dropped because the itnlividual spent part of the month on PPP and pan on hourly wages.

standard deviation of 1.53. The average actual pay was $2,254, which is above the amount that would be paid had the worker received exactly the amount to which he was entitled based on a straight piece raw. The ctifferenoe reflects vacation, holiday, and sick pay, as well as two other factors. FM not all workers are on PPP during the period. When on hourly wages, some received higher compensation than they would have had they been on PPP, given the number of units installed. Of course, when a given worker switches to PPP, incentives change and his output may go up =ou* to cover 04 deficit. Second, even when workers are on PPP, a substantial fraction of person-weAs calculated an the basis of the PPP formula comes in below the guaranteed weekly campensadon. The guarantee binds for those worker weeks, and actual pay then exceeds PPP pay. In all months after the introduction of PPP, at lout some workers received the guaranteed pay and some earned more than the guarantee. Thus, die sufficient conditions for Propositions 2 and 3 are met throughout the period. Means for actual and PPP pay reveal almost nothing about the effects of PPP on performance and sorting. A more direct approach is needed. Table 2 presents some meaus of the key variables and breaks them down by the PPP dummy, which is set equal to one if the worker in question is on PPP during the given month. 11 The story that will be told in more detail below shows up in the simple means. The average level of units-per-worker-perrday is about

Ii Only observations where workers were on one pay regime or the other for the tiill mouth are ttsed. partial month observations we deleted.

0.54 units, or 20 percent higher in the piece-rate regime than in. the hourly wage regime. Also, the variance in output goes up when switching from hourly wages to piece rates, as cn be seen by compoft the standard deviations of 1,59 to 1.42.12 1-nus, Ptopositim 1, 2, and 3, which stow that both mean and variance in output rise when switching from hourly wages to Few rates, am borne out by the simple statistics. FFtirther, note that there is good indication that profitability want up significantly with the switch. The cost per unit is considerably lower in the piects-rate r4me than it is with. hourly wW:a3 The simple statistics do not take other factors into amoumt in particular, am glass demand is closely related to miles dciveiu which varies with weather. M4jor storms, especially had, also cause glass damage. Month effects and year effeets matter. Perhaps more impoftnt, the management change that took place before PPP was instituted had other direct effects on the company that may have changed output during the sample period, irtespeetive of the switch to PPP. To deal with these factors, month and year dummies are included. The simplest specification in the first row of Table 3 yields a coefficient on the PPP dummy of 0.368. Evaluated at " x7hia number includes within-worker components as well as between-worker componentg. The latter is of interest and is investigated. In more detail below. " The fact, that actual pay has only risen slightly after the switch to PPPilin before reflects the phase-in pattecn of the PPP program. l.owr.r wagc areas were brought into the program first, which means that dte PPP = I data are dominated by hower-wage markets. This pattern also affects the differences between piece-rate and hourly wage output if early switchers to PPP have different average output levels than late switchers to E'P'P.

4968

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 8 of 17 VOL 90 NO. 5

1353

LAZFAR: PERFORM"CE PAY AND PRODUCTIVITY TABU 3-Rsa"M" Rg sucIs

uomary for Regression number

PPP personMonth observation

Tenure

rum since PPP

New regime

je

Description

1

0.368 (0.013)

0.04

Dummies for montli and yew included

2

0.197 (0.009)

0.73

Dummies for mondi and year, worker. specific dumnim included (2,75S

individual Waskft) 0.313 (0.014)

0.343 (0.017)

0.107 (0.024)

0.05

Domadce for month and year included

0:202 (0:009)

0.224 (0.038)

0.273 (0.018)

0.76

buminics fas month and yew, wadecgpecifir dummies included t2,795 individual, workesas)

0.309 (0.014)

0.424 (0.019)

0.130 (0.024)

0.06

Dummies; for month and year included

0.243 (0.023)

Notes: Standard errors are reported in parentheses below die coeffficiants. Dependent variable: In o>sput•per-worker-pex-iay. Number of observations: 29.837.

the mean of the log of uniis-per-wo[ker-perday, this coefficient implies that there is a 44percent gain in productivity with a move to PPP. There are three possible interpretations of this extremely large and statistically precise offect. First, the gain in productivity may result from incentive effects associated with the program. Second. the gain may teak froaat sorting. A different group of workers may be pment after the switch to piece rates. IbK the, pattern of impiemeatatufn may cum a sporious positive effect. Suppose that $a€alite picked its best workers to put on pim rates first. The PPP dummy coefficient would pick up an ability effect because high-ability workers would have more PPP months than low-ability workers. Unless ability is correlated with region in a particular way, the third explanation can be ruled out because Safelite switched its stores to PPP on a regional basis, starting with Columbus, Ohio, where the headquarters is located, and moving out. The other two effects can all be identified by using the data in a variety of ways. When worker dummies are included in the regression, the coefficient drops to 0.1.97 from 0.368.11he 0.197 is the pure incentive effect that results from switching from. hourly wages to piece rates. Evaluated at the means, it implies that a given worker installs 22 percent more units after the switch to PPP than he did before

the switch to PPP. This estimate controls for month and year effects. Individtial ability is held cpnstant as is shop location by including the person dummies. AppmKimately half of die 44-percent difteroce in productivity attributed to the PPP program reflects an incentive effect. Nor does this gain appear to be aBawttom efiOd.''' Tfiiscan ks6en by examining 3 in Table 3. Thrregression incll>das a vtniigble1ur team and also me for time that the worker has been on the PPP peoS= It is zero, for altmastft before the individual is on piece rates. Tt is ft number of years that the individual has- been on piece rates in the current per5an-moaah observadon. For ex,ample, a worker who started 1994 on hourly wages and was switched to PPP on July 1, 1994 would have time since PPP equal to zero for the June 1994 observation, to 0.5 for the January 1995 observation, and to 1.0 for the June 1995 observation. Consider the estimates with fixed effects in regression 4. The coefficient of 0.273 on time since tenure coupled with a PPP dummy coefficient of 0.202, means that the initial effect of switcJsing from hourly wage to piece rate is to increase log productivity by 0.202. After one year

u'1ha Hawthorne effe,ct, named after the Hawffiorot Western Electric Plant in Itliopois, alleges that any change is likely to bring about dion-terin gains in productivity.

4969

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 9 of 17 THE AMERICAN ECONOMIC REVIEW

1354

on the program, the increase in log productivity has grown to 0.475. The Hawthorne effect would imply a negative coefficient on time since PPP. If the Hawthorne effect held, then the longer the worker were on the program, the smaller would be the effect of piece raw on productivity. The reverse happens here. After workers are switched to Piece rates, they seem to learn ways to wo4, faster or harder as time progresses.

M. Sorting Tenure effects are large and significant. Using regression 3 of Table 3, it is estimated that one year of tenure raises log productivity by about 0.34. As is true of all tenure estimates, there are two interpretations. The first is learning. Turnover rates are over 41/2percent per month, and the mean level of tenure is only about two-thirds of a year. It would not be surprising to see a worker increase his windshield installation rate dramatically during the first few months on the job. The second interpretation is one of sorting. Those who are not making it get fired or quit early. Regression 4 of the table assists in interpretation. Regression 4 reports the estimates of the regression in regression 3, inclufdittg fixed effects for individuals. Thus, the tenure coefficient reflects the effect of tenure for a given worker, averaged across individuals. The estimate of 0.20 on log productivity can be interpreted as the average effect of learning within the satnple.t' Thus, the effect of learning appears substantial. The theory stated in Propositions 2 and 3 suggests that the optimal piece rate is implemented such that both mean and range of worker ability should rise after the switch to piece rates. The theory implies specifically that there should be no change in the number of low-ability workers who are willing to work with the firm, but that piece rates would allow high-ability workers to use their talents more lucratively. Thus, the top tail of the distribution should thicken. 's The term "average" is used cautiously. The sample contains different numbers of observations at each tenure level so that the average picks up not only nonlinearities. but different tenure effects for different types of individuals that may be more or less heavily weighted in the sample.

DECEMBER 2000

Underlying ability is difficult to measure, but actual output can be observed. The fifth regression of Table 3 provides evidence on this point. "New regime" is a dummy set equal to one if the individual was hired after January 1, 1995, by which point almost the entire firm had switched to piecework. The theory predicts that workers hired under the new regime should produce more output than the previously hired employees '6 Indeed, workers hired under the new regime have log productivity that is 0.24 greater than those hired under the old regime, given tenure. Separations can also be examined. Suppose that workers must try the job for a while to discover their ability levels. Workers who find the job unsuitable leave, Ilea, looking at the relation of ability to separation rates (quits plus layoffs) before and after the switch to piece rates will provide evidence on the validity of Propositions 2 and 3. A separation is defined as an observation in which the worker in question did not work during the subsequent month. Thus, a dummy is set equal to one in the last month of employmerit. Those workers who work through July 1995 (the last month for which data are available) have this dummy set equal to zero for every month in which they worked. A worker who was employed, say from January 1994 through February 1995, would have the dummy equal to zero in every month of employment, except for February 1995, when it would equal one. Table 4 reports a breakdown of separation rates by PPP regime and by worker output deciles where output is defined as units-perworker-per-day during the previous month.'"' First note that simple effect of a move to PPP increases turnover from 3.3 percent per month to 3.6 percent per month, but the difference is not statistically significant.ts The direction of

"Taken literally, the theory implies that none of the low-output incumbents should leave since the guarantee makes them no worse off than before, but some ltigherqualtty workers are now willing ►o take the job. " This is done so that no mechanical connection between low output per week and separation would exist as a result of leaving in the middle of a week. 19 Note that the turnover rates in Table 4 are lower than the one reported in Table 1. This is because in order to be

4970

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 10 of 17 1355

lG4ZFAR: PERFORMANCE PAY AND PRODUCTlV1TY

VOL 90 NO. S

TABLE 4-SarnsnToN RATES By Recta AND D>:cus

PPP regime

Houly regime

Standard error

Difference

Standard error

1,285 1.491 1,625 1.691

0.005 0.005 0.005 0.005

-0.002 -0.006 -0.005 --0.002

0.007 0007 0.007 0.007

0.034

1,693

0.004

-0.003

0.007

0.04

1,792

01005

0.002

0.007

0:004 0.005 0.006 0.004

0A3 0.03 0.022 0.027

1,777 1,879 2,169 339

0.004 0,004 0.003 0.009

0.005 0.001 -0.008 -0.007

0.006 0:006 0.007 0.009

0.002

0.036

15,741

0.002

0.003

0.002

Standard error

Separation rate

1.641 1,465 1,358 1,245

0.005 0.005 0.005 0.005

0.039 0.038 0.037 0.037

Number of observations

Difference between PPP and hourly separation. rates

Docile

Separation rate

Lowest 0 1 2 3

0.041 0.043 0.042 0.039

4

0.037

1,282

0.005

5

0.038

1,279

0:005

6 7 8 9

0.025 0.029 0:03 0.033

1,223 1,135 880 2,437

Highest Overall

0433

13,945

the change is not surprising since a major change in the pay system may make some of the incumbents unhappy enough to leave or may signal that the firm has become less tolerant of low productivity. Second, theory predicts that those at the higher end of the ability spectrum should see turnover rates that declit►c, Although the highest output deciles are the ones that experience the largest declines in separation rates, the differences are not statistically sij*ficant. IV. Fued Effects Some of the theoretical predictions can be tested by estimating person-specific fixed effects. Since the data set consists of multiple observations on a given individual over time and under different regimes, person specific effects can be estimated. Fixed effects are estimated from a regression of the log of outputper-worker-per-day on tenure and time dummies. Should this be done using data from both regimes combined or from one or the other? Some workers were employed in both hourly

in the sample for Table 4; the worker must have been with the firm during the previous month as well. Thus, those who leave during their first month are included in Table I but not in Table 4.

Number of observations

wage and piece-rate regimes whereas some worked in only one regime. The theory implies that incentives are muted during the hourly wage period, so it is not clear that fixed effects based on output during the hourly wage period are good proxies for ability. This might suggest using the fixed effects estimated during the piece-rate regime for those who worked in both regimes. But then separation behavior over the two regimes cannot be examined since no one who worked in both hourly wage and piece-rate regimes left the firm during the hourly wage regime. An alternative is to use the hourly wap regime estimated fixed effects, based on the argument that fixed effects are highly correlated across periods. Indeed, there is evidence of strong correlation. Figure 2 shows the scatterplot, which reveah the pattern. The correlation between the fixed effect from the hourly wage period and that from the piece-rate period is 0.72 with 1,519 observations. T7t.is. correlation is high, but not .perfect. There are some workers who performed relatively better under the hourly wage system than under the piece-rate system and vice versa. A regression of the fixed effect from the piece-rate regime on the same individual's fixed effect from the hourly wage regime yields a coefficient of 0.700 with a standard error of 0.017. The constant term is -0.04 with a standard error of 0.01. The effect of

4971

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 11 of 17 DECEMBER 2000

7t`fE AMERICAN ECONOMIC REVIEW

1356

TABLE S-VAm'rtort ttr FtxeD E»m

raw dFw 6® peas+.r ,.bsso^ -^

14 .^.ra

°

o ° ^•° 8 ^'^. a

-J.J477i

Difference

o,m

Oq

Q

4

a

o

•

Regime Hourly wage Piece rate

Number of individuals

Standard deviation in fixed effects

between 90th and 19th parcentile in fixed effects

1.519 1,519

0:65 0.64

1.28 1.12

°°

°

e

F*$d.a,M*,,. ►..1Y

rq*W

FIGURE 2. ScATfE2r1AY OF FIXED EFFECTS FROM THE Two R8(iUm

ability on effort is attenuated during the hourly wage period because there is ►ess incentive to put forth effflrt. If the fixed effect of output in the piece-rate period measures true ability, wherea,c the fixed effect during 60 hourly wage period measures ability only impWwdy, then the coefficient in the regression of piece-rate fixed effects on hourly wage fixed effects is biased toward =o.t9 The fact that it equals 0.100 suggests that workers do reveal their abiiities to a large extent even during the hourly wage period.-I This evidence provides a rationale for using the hourly wage-period fixed effects to examine turnover. The medi level of fixed effect for those who leave no later than two months after the start of the piece-rate system (the leavers) is 0.15 with an upper bound of the 95-percent confidence interval of 0.19. The median level of fixed effect for those who stay beyond the initial two months (the sttaye.ts) is 0,22 with a standard error of lower bound of the 95-pdrxs confidence interval at 0.21. The medians are signift9 The bin is caused by the standard erm%-in-variables problern. where the observed independent variable Is not the true effect, but Instead the me effect plus measurement error. 20 The relation of ability to output need not be nwnotonic, especially during the hourly wage period. Since the lowest- and highest-ability workets. .!10 and A,,, earn no mats, tbey should be least concerned about losing their jobs. Middle-ability workers earn mats and may themfore put forth additional effort to reduce the likelihood of a termination.

icantly different, with the more able, as measured by pre-period fixed effects, being mom likely to There is no evidence that the stayers have higher variance in ability than the leavers. The standard deviation of the fixed effects for the stayers is 0.68 and that: for the leavers. is 0.89, With number of observations equaling 1,511 and 659„ respectively. More evidence on this point is presented in Table 5, where lized effects estimated-on hourly Wage-regime data are computed for those individuals who worked in both regimes. Ajaio, the results of Table 5 suggest that dte Prediction about variance in ability finds no support in the fixed effects results= 'I he standard deviation in fixed affects among piece.-rate wo&m Is virtually identical during the piece rate and hourly wage regime. The 90-10 percentile is higher during the hourly wage regimeAlthough T.able 2 reveals an increase in the variance in output when the firm switches from to piecework, the increase in varis hourly wges ance does not reflect an obvious change in the dispersion of underiyiag ability. Summarizing, it is clear that person-specific effects are important. They play a significant role in the interpretation of the results of Table 3, and their pattern is consistent with the theory in that their mean levels tend to rise as t3he firm goes frocu time rates to piece rates. They pro. vide no support for the hypothesis that variance in underlying ability increases when the firm switches from time rates to piece rates. Ability 21 Part of this diffexentx may reflect pure selection that would occur even in the absence of a regime change. Presun>atbly, the tenure variable included in the output regression controls for most of the regime-itxltpendent sorting. 22 The difference between this sample and the previous one is that the former sample included those who left before piece-rate-based fixed effects could be estimated.

4972

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 12 of 17 1357

L,AZEAR: PERFORMANCE PAY AND PRODUGT7V#TY

VOL 90 NO. 5

Tnscs 6-AKwssm Res[n.is o.oe Reipasion number

R3

Description

OAb$ (0.005)

0:06

D'Imrtdes for month and year included

0.099

0.76

Dummies for month

PPP dtimmy

°+a` ^•

OAB ^

9 •

0,041

•

4 s•

^•

a

2

°s

(0.004) •

° a

O.Q! i ed

° •

and year workerspecific dummies included (2,755 individual workers)

OpO^aa^I!!.N••u••eosa°

1

2

3-^ s s 7 s a to ou^v^

Flottde 3. RawM DOMMS IN 1tt& Two RtCIOM

is higher among those who work at the end of the sample, period than among workers prownt at the beginn4 of the ample period. Most of the increase in ability is a result of selection through the hiring process that occurs after piece rates are adopted. The effect of differential changes in turnover rates, hiring policy, and incentives can be snmrnarmed by the kernel densities of output shown in Figure 3. The two disaftudoes look rather similar, but it is clear that the On*-rata distribution lies to the tight of the hourly wage digtribution. purthm the peak vidue of the density function during the pieecworh regime is lower than that of the hourly wage regime. Them is less concentration of output around the modal value under piece rates than there is under hourly wages.23 V. Fay and Profitability The effect of the program on pay can be traced also. Table 6 reports the effects of the switch to the PPP regime. The log of pay-per-worker went up by 0.068, implying about a 7-percent increase in compen23 7Tie model in Figure 1. taken Gtaaily, implies that there should be holes in the data, wbicltua not fbatul. Tbare am a number of possible explimstions. First, workers may try to get into >begiewc9te range and fi& Second. aim the unit of measurement is a month, there may be some weeks during wltieh die worker bits the piece-rate range and others where he does not, averaging out to some amount between 2.5 and 5 units. lbicd, do worker may to guess eo perfeetly, and this cream variance around sp. Finally, time may be other reasons to exceed the minimum level of output.

Notes: Standard errors are reported in parentheses below the coefficients. Dependent ,variable: In pay-pa-dityNumber of observations: 29.837.

sation. Recall that the increase in productivity for the firm as a whole was 44 petceatt.Regresslou 2 of Table 6 implies that the log of pay for a given worker rose by 0.099, implying a 10.6 pexcent gain in earnings. This is jug under-half the increase in per-worker productivity. Thus, the firm passes along some of the benefits of the gain in productivity to its existing workforce. The effect without worker duounies is smaller than that with worker dumtmes because the newer workers are paid less than the more senior workers whom they replam Further, 92 percent of workers experienced a pay increase, with a quarter of the workers receiving incream at least as large as 28 percent. Did profits rise? This depends on the increase in productivity relative to the increase in labor and other costs. Given the numbers (44-percent increase in productivity, 7-percent increase in wages), it is unlikely that other variable costs of production ate up the margin still given to the flmL The piwA-tate plan seems to have been implemented in a way that likely made bofh capital and labor better off.24 There is one cost that has been ignored throughout. Piecework requires measurement of output. In Safelite's case, the measurement comes about through a very sophisticated information system. But the system involves people and machines that are costly. Indeed, in equilibtium, firms that pay hourly wages or monthly salaries are probably those for whom

24 The 6rm•s earnings we up substantially since the switch to piece xaas, but this could reflect other faetors as well.

4973

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 13 of 17 1358

THE AMERICAN ECONOMIC AEWEW

measurement costs exceed the beriefits; from switching to outprit-based pay. In this case, the gains in productivity were very large. Further, the information systems were initially put in place for reasons other than monitoring worker productivity, having to do with inventory control and reduced installation lags. The economies of scope in information technology, coupled with the labor productivity gains, are probably large enough to cover whatever additional cost of monitoring was involved.' V1. Quality

One defect of paying piece rates is that quality may suffer.26 In the Safelitte case, most qualiry problems show up rather quickly in the form of lrolten windshields. Since- the guilty installer can be easily identified, there is an efficient solution to the quality problem The installer is required to reintatt the windshield on his own time and must pay the company for the replacemeat glass before any paying jobs am assigned to him. This induces the installer to take the appropriate amount of care when installing the glass in the first place.2' Initially, 5afe}ite used another system that relied on peEd' presst1re.2S When a customer

repotud a defect, the job was randomly assigned to a some worker in the shop that was responsible for the problem:" The worker assigned to do the re-do was not necessarily the worker who did the original installation and the

25 This was not always so. Whenever a firm switches from one pay system to another, it is almost certain that one system does not maximize profit. 26 See Lamar (1986) and Baker (1992). " Both J. Asch (1990) examines the effects of compensation schemes on military recruiter performance with a focus on quality dimensions. As recram receive incentive pay for signing up recruits, there is a tendency to take lowenquality applicants. 2' Sea Eugene Kandel and lazear (1942) for a discussion of the effects of peer pressure and norms in an c►[ganization. ^ ram are two advantages of assigning the re-wark to the shop rather than the individual. First, the customer gets faster service since it is unnecessary to wait for the availability of the original installer. Second, some workers will have already separated from the firm before the defect is noticed. Assigning the work to the shop deals with this problem. Neither argument provides a rationale for forcing the re-work to be done by others without pay.

DECEMBER 2000

worker was not paid for doing the repair work. But workers knew the identity of the initial installer. If one installer caused his peers to engage in too many re-dos, his coworkers pressured him to improve or resign. More rr0ofttiy, the system was changed to assign ro-do work to the worker who did the initial installation. Workers are not paid for the re-do, but they are not charged for On wasted glass or for other costs associated with the re-do, as a folly efficient system would require. The outcome has been that quality has gone up after the switch to PPP. rather than down. The firm surveys its customers on their satisfaction with the job. The customer satisfaction index rose from slightly under 90 percent at the beginning Of the saMple period to aroutid 94 percent by the end of the sample period. Because re-dw are costly to the worker, he is motivated to get it rot the first time around. Y1I. PkeewerY Is Not AWM -Profit" It is interesting that the productivity gains are so large for this particular firm. Of course, this is only one data point and, it is we where the case for Piece rata; seems especially Strong Output is easily measured, quahty problems am readily detead, and blame is assMud* Manageria3 and professional jobs may not tae as well suited to piecework. The fact that the productivity gains are so large in this case is worthy of attention, but these results do not imply that all firms should switch to piece-rate pay. Piece-rate pay, defined narrowly, is used sparingly in the United States. Although it is difficult to obtain data on the distribution of piece-rate work, one national survey, the National Longitudinal Survey of Youth,3° asked whether a worker was on piece rates up through 1990. Results from this survey are shown in Table 7. In a subsatnple of 7,438 workers, 3.3 percent reported being on piece rates. The number varies significantly by oocupqpon. As expected, managers, whose output is difficult to measure,

1O The NLSY is a nationally representative sample Of 12,686 young men and young women who were 14 to 22 years of age when they were first surveyed in 1979.

4974

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 14 of 17 VOL 90 NO. S

lA7lsAR: PERFORMANCE PAY AND PRODUCTIVITY

Tns18 7--Pi8^RATia PaopoRmons IN rHeNnnorreu. LoNomuotar, SuRvey op Yom

0004ation ProfessiauVtectmical Managerstof6cials/proprietars Sales Cteticat Craftsmen Operatives Nonfetm labor Farm labor TOTAi.

Nmber in occupation

Percent pica

in NLSY

rate

1r848 468 477 1,322 2,278 296 543 6 7,438

1.4 0.9 1.3 1.3 3.6 9.8 13.8 16.7 3.3

are least likely to be on piece rates. At the other extreme, about 14 percent of laborers are paid piece rates. The use of incentive pay. broadly defined, is more widespread. Pencavel (1978 p. 228, Table 2) reports a peak of 30 percent of workers in rnanttfaeturittg; who received incentive pay in the United States in 1945-1946, with a downward trend afterward. ne relative paucity of plece-rate pay in the United States is not particularly disturbing for this study. For one thing, piece ram remain more prevalent in other countries. For example, using a data set on manufacturing in Sweden (which accounts for 20 percent of the workforce), it is found that 22 percent of workers received piece-rate pay as late as 1990. But even were this not the case, the experiment would be relevant. As far as the workers are concerned, the effect of a dtange in compensation was exogenous, and the data consist of about 3,000 independent worker responses to that common change. The implication of the study is not that firms should switch to piecework, but rather that when workers faced a new compensation scheme, they responded by altering effort, turnover, and labor-supply behavior in the way predicted by theory.

1359

does not imply that profits must rise. Market equilibrium is characterized by firms that choose a variety of compensation methods. Firms choose the compensation scheme by comparing the costs and benefits of each scheme. The benefit is a productivity gain. Costs may be associated with measurement difficulties, undesirable risk transfers, or quality declines.

The theory above implies that average output per worker and average worker ability should rise when a firm switches from hourly wages to piece rates. The mini um level of ability does not change, but more able workers, who shunned the firm under hourly wages, are attracted by piece rates: As a result of incentive effects, average output per worker rises. Thus, average ability and output, as well as variance in output and range of ability, should rise when a firm switches from hourly wages to piece rates.

VllI. &==wry and Conclusion

The effects of changing the compensation method were estimated using worker-level, monthly output data from SafeIite Glass Company. The primary predictions of the theory are borne out. Moving to a piece-:rate regime is associated with a 44-percent increase in productivity for the company as a whole. part of the gain reflects sorting, part reflects incentives, and some may reflect the pattern in which the scheme was implemented. The incentive effect of the piece-rate scheme accounts for an increase in productivity of about. 22 percent. The rest of the 44-percent increase in productivity is a result of sorting toward. more able workers or possibly some other factors. Sorting occurs primarily through the hiring process, where a disproportionate share of new hires come from higher ability groups after the switch to piece rates. There is no strong evidence that the change to piece rates increases separations relatively more among lower-output workers. Nor is there evidence of an increase in range or variance in underlying ability after the switch to piecework.

The results imply that productivity effects associated with the switch. from hourly wages to piece rates are quite large. The theory implies that a switch should bring about an increase in average levels of output and in its variance. These predictions are borne out. The theory

Since the data measure actual productivity, tenure effects on productivity (rather than wages) can be estimated. Tenure effects on productivity are found to be large. Part reflects learning on the job, but a significant fraction reflects sorting that induces the least productive workers to leave first. Also, time since the

4975

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 15 of 17

introduction of the piecework scheme is positively associated with productivity. Workers capttttrd some of the return from moving to piece cates. The average incumbent worker's wages rose by just over 10 percent as a result of the switch. Over 90 percent of the workers had higher pay dueing tfie piece-raxe period than they did during the hourly wage period.

PROOF OF PROPOSIITON I: Output cannot fall below eo because of the firm-imposed constraint at co. But output may exceed ea if for some A. Ap S AS Ah, (Al)

DECEMBER 2000

THE AMERICAN ECONOMIC REVIEW

1360

U(W, Xo(A)) < .U(bf(X*(A), A) - K, X* (A))

A0 is willing to work for W at effort efl because A0 worked under these term before Fw:khermore, since the guarantee has not been made any more attractive, no one with A < A0 is willing to work for the guaranteed wage. Since the lower bound on ability remains the same and the upper bound does not fall and generally rises, average ability does not decrease and generally increases after the switch to piecework. PROOF OF PROPOSITION 3: From the Proof to Proposition 2, A a:;- A. But A. cannot rise because the wage guarantee is still available so A. remains willing to work. This is sufficient to imply that range or variance in ability rises. Also, since all workers choose to produce eo under the hourly wage, but some produce in the piece-rate range with the new scheme, positive variance in A implies positive variance in e under piece rates. REFERENCES

where X*(A) is the effort level chosen by worker of type A given piece rate t ►. As #ong as thsre is some We A for wbom output rims, average output must rise. PROOF OF PROPOSITION 2: If any choose to work in the .piece-tate range, then surely the worker with the highest ability chooses to work in this range. But the highestability worker cannot, except in the rarest coincidence, be Ah. If A* chooses to Wodk in the pi,sce-rate range, then Ah, who was indiffirent to working under hourly wages, is at worst indiffet:ent to working under piece rates, but more generally, strictly prefers the piece rate. If A. earns rents under the new plan, then A,h is no longer the marginal worker. That exists an A*h with A> Ak who would now be the marginal worker, i.e., the worker for whom U(b.`(X*(Ar*,), Ah) - K. Ak) = U(*(A), I(A)) where X* ( A h) is defined as the effort for type A; under piece rates b and where VV, t are the wage and effort on the alternative job.

Also, if any accept the wage guarantee, then surely A0 accepts the guarantee. We )mow that

Ascm„ Beth 3. "Do Incentives Matter? The Case of Navy Recruiters:' Industrial and Labor Relations Review, February 1990. Spec. tss., 43(3), pp. S89-106. AsbeeMfsac, Orley and Pomvd, John K "A Note on Measuring the Reladombip between Changes in Earnings and Changes in Wage Rates." British Journal of Industrial Reladrns, Mtuncir. 197b.14(9.), pp. 70=76. Baw, George. Incentive Contracts and 1'wftirmatsce Measurement" Journal of Political Eeonvmy, June 1992. 100(2X pp. 598-614. Buoth, Allison and FraW4 JeM "Performance Related Pay." Unpublished manuscript, University of Essex, 1996. Brown, (harks. "Firms' Choice of Method of Pay." Industrial and Labor Relations Review, Febaiary 1990; Spec. Iss., 43(3), pp. S265-82.

."Wage Levels and Methods of Pay." Rand Journal. Autumn 1992, 23(3), pp. 36675. Brown, Martin and P4Wps, Peter. "The Decline of Piece Rates in California Canneries: 1890-1960." Industrial Relations, Winter 1986, 25(1), pp. 81-91.

Dad, E. L. "Effects of Externally Mediated Rewards on Intrinsic Motivation." Journal of Personality and Social Psychology, January 1971, 18(1), pp. 105-15.

4976

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case Page 16 of 17 VOL 90 NO. 5

LA?lAR: PERFORMANCE PAY AND PRODUCTIVrI7

Drago, Robert and .Heywood„ John S. '"The Choice of Payment Schemes: Australian Establishment Data." Industrial Relations, October 1995, 34(4), pp. 507-31. Fama, Eupoe F. 'Time, Salary, and Incentive Payofk in Labor Co^cts." Joumal of Labor Economics, January L991, 9(1), pp. 25-44. Feraie, Sue and MeIprK, Dava "It's Not What You Pay It's the Way That You Pay It and "["hat's 'What Gets Results: Jockey's Pay and Performance." Centre for Economic Performance Discussion: Paper No. 295, London School of EconDmies, May 1996. Geidin, Claudia. -Monitoring Costs and Occupational Segregation by Sex: A Historical Analysis:' Journal of I,abor,Fconomics, January 1986, pp. 1-27. Ktnrdel, fluveoe and Lawar, Edward P. "Peer Pressure and Partnerships." Journal o,f ,i°toliticai Economy, August 1992,100(4), pp. 8()117. l,azesrr, Edward P. "Salaries and Piece Rates.' Journal of Business, July 1986, 59(3), pp. d05-31.

1361

. "Performance Pay and Productivity." National Bureau of Economic Research (Cambridge, MA) Working Paper No. 5672, July 1996. Lepper, Mark 84 Greene, David and Nisbrst, Rkhard E. "Undermining Children'slnbistsic Interest with Extrinsic Reward: A Test of the 'Over,jusdflccaalaiton' Hypotheds." Journal of Personality and Social Psychology, ftbruary 1973, 28(1), pp. 129-37.

Pawrseh, Bury J. and Sbeprer, Bruce S. "Fixed Wages, Piece Rates, and Intertemporal Productivity: A Study of Tree Planters in British Columbia." Unpublished manuscript, 1996. Paacavel, J" "Wort Efiort. On-ft-lob Screening,. and Alternative Methods of Remufteratiott." Research in Labor Economics, January 1978, 33(1), pp. 225-58. Sever, Farlc. "Piece Raw vs. Time Rate: The Effect of Incentives on Earnings." Review of Economics and Statistics, August 1984, 66(3), pp. 363-75. Slichter, Suaainer. Modern economic society. New York Henry Holt and Co., 1928.

I

4977

WP/JCH Testimony-fns 6 & 10 2009 TX Rate Case

Page 17 of 17 Copyright of American Economic Review is the property of American Economic Association. The copyright in an individual article may be maintained by the author in certain cases. Content may not be copied or emailed to multiple sites or posted to a Iistserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

4978

This page has been intentionally left blank.

2009 ETI Rate Case

WP-1-99

4979

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 1 of 30

CEO Compensation and Stakeholders' Claims* ALKA ARORA, University of Wiseonsin -Whitewater

PERVAIZ ALAM, Kent State University Abstract The traditional view that a corporation exists solely to serve the interests of the firm's shareholders has given way to a changing view that recognizes the importance of corporate constituents in addition to shareholders. Prior studies demonstrate a significant association between the sensitivity of CEO compensation and a firm's stock prices. However, the association between CEO compensation and the claim of other primary stakeholders (customers, employees, suppliers) has not been examined. The purpose of this study is to investigate whether the adoption of long-term incentive plans aligns the interest of the CEO with the interest of the primary stakeholders in the firm. Using the fixed-effect regression, our results indicate a significant association between the change in CEO compensation and the claims of the customers, shareholders, and employees. We contribute to the literature by demonstrating that the managers are accountable not only to the shareholders but also to primary stakeholders. Keywords CEO compensation; Corporate constituents; Long-term incentive plans; Stakeholders JEL Descriptors G30, J33, M41, M52

R6mundration des chefs de la direction et revendications des parties prenantes Condensd FonciZrement, Is majoritb des Etudes Ecottomiques empiriques portant sur Is r6mun6ration des cadres prdsupposent que lea regimes de rEmun6ration des cadres sont adoptds pour harmoniser lea intbrgts de la direction et ceux des actionnaires. Bien que les recherches comptables pr6c6dentes fassent Etat d'une relation significative entre les changements dans la r6munEration des chefs de Is direction et Is performance du cours de 1'action de l'entreprise (par exemple, Murphy, 1986). la relation entre la rdmunEaadon des chefs de Is direction et Its revendications des autres parties prenantes de 1'entreprise a essentiellement W ignorde. Les ouvrages de gestion dEfinissent lea parties prenantes comme f.tattt des personnes ou des groupes qui possbdent ou revendiquent des droits de propriEtb ou des intbr8ts dans une •

Accepted by Peter Clarkson. We are especially thankful to Peter Clarkson and two anonymous reviewers whose comments were instrumental in developing the paper. We would like to thank Peter Barton, Christine Clements, Robert Gruber, Rosemerry Rudesal, and Roy Weatherwax for their support. We also appreciate Jan Winchell's computational assistance.

Contemporary Accounting Research Vol. 22 No. 3 (Fall 2005) pp. 519-47 ® CAAA

2009 ETI Rate Case

WP_1-100

4980

WP/JCH Testimony-fn 2009 TX Rate Case Page 2 of 30

520

Contemporary Accounting Research

entreprise et ses activitEs. Selon Clarkson (1995), un groupe de parties prenantes principales est un groupe sans la collaboration constante duquel une entreprise ne peut survivre et poursuivre son exploitation. Les parties prenantes principales sont ddfinies comme 6tant les clients, les employEs, les fournisseurs et les actionnaires. La t8che consistant A dquilibrer les objectifs des parties prenantes est un processus complexe, compte tenu de l'eventail des besoins et des attentes conHictuelles que les parties prenantes entretiennent it I'Egard d'une organisation. $tant donn6 sa position au sommet de la hiErarchie de 1'entreprise, le chef de la direction est Is personne la plus susceptible d'&re chargee de la t8che dEcisionnelle d'Equilibrer 1'ensemble des exigences variEes et conflictuelles des parties prenantes (Andrews, 1971 ; Lenz et Lyles, 1986). Les hauts dirigeants, en raison de leur « position strat6gique» (Herman, 1981) relativement aux dEcisions clEs de I'entreprise, peuvent gtre consid6rEs comme des agents contractuels de 1'entreprise. Lon peut done envisager cette derni8re comme un « resew de contrats » entre les hauts dirigeants et les parties prenantes (Jones, 1995). La prEsente etude a pour but d'explorer si 1'adoption d'un programme il long terme d'incitation au rendement permet d'harmoniser les intE.rBts respectifs des parties pn:nantes principales de 1'entreprise et du chef de la direction (c'est-h-dire du cadre supdrieur). Les programmes 81ong terme d'incitation au rendement ne se sont rEpandus dans les entreprises qu'au ddbut des ann6es 1970. B s'agit de programmes it long terme reposant sur des bases comptables, qui visent it rE.tribuer Its cadres pour 1'am6lioration du rendement sur une pbriode de trois A six ans. Les objectifs de ces programmes sont Enonc6s sous forme d'indicateurs comptables (par exemple, la croissance du bEnEfice par action et celle du rendement de 1'actif), et le nombre d'actions octroyE.es aux gestionnaires d6pend de la performance il long terme de l'entreprise (Pavlik et a1.,1993). Les faits constat6s indiquent que les programmes d'incitation au rendement Wont pas remplacE ou affaibli la structure de remuneration en place (qui comprend les salaires, les primes, les avantages indirects et les options d'achat d'actions), mais se sont greffis au regime de remuneration existant des cadres (Kumar et Sopariwala, 1992). De nouveaux regimes de rEtnun6ration sont susceptibles d'8tre adoptds par les entreprises qui ne sont pas convaincues que leur structure de remuneration offre aux cadres les encouragements qu'il faut pour maximiser la valeur pour 1'actionnaire (Brozobsky et Sopariwala, 1995). Le point de vue traditionnel selon lequel 1'unique but de 1'entreprise est de servir les intf,r8ts de sea actionnaires (Friedman, 1962) a cEdE la place it la perspective dans laquelle est reconnue ['importance de toutes lea autres constituantes de l'entreprise. En 1950, le g6nEral Robert Wood, leader de la croissance rapide de Sears aprt:s la guerre, a 6num6rE les « quatre parties prEsentes dans toute entreprise, par ordre d'importance *, soit « lea clients, les employes, Is communautb et les actlonnaires» (Preston, 1990). Si Its besoins et les int6r8ts appropriks des trois premiers groupes sont satisfaits de manibre efficace, a-t-il afBrmE, les actionnaires de I'entreprise en tireront profit. Dans cet esprit, les auteurs estiment que, dans son souci de maximisation de la valeur pour 1'actionnair+e, le chef de la direction dynamique ne saurait ignorer lea intErBts relatifs des autres parties prenantes principales, soit Ies clients, les employEs et les fournisseurs. En gEnEral, la communautd est dbfinie comme 6tant un groupe de personnes qui poursuivent un but commun. Outre les clients, les employds et les actionnaires, les fournisseurs d'une entreprise out un intErBt dans la survie de cette dernibre. C'est pourquoi les auteurs comptent les fournisseurs an nombre des parties prenantes dans leur analyse. CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-101

4981

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 3 of 30

CEO Compensation and Stakeholders' Claims

521

Les auteurs se demandent si un programme d'incitation au rendement nouvellement adoptd harmonise ou non les intdrgts du chef de la direction et les inMts, complEmentaires mais diffdrents, des parties prenantes principales. Ils formulent 1'hypothbse selon laquelle la relation entre la rEmundration du chef de la direction et les parties prenantes principales de I'entreprise (soit ses clients, ses employbs, ses fournisseurs et ses actionnaires) est sensiblement plus marqufe dans Is pdriode suivant I'adoption que dans la pdriode prdcesdant I'adoption du programme. Les auteurs ont recours dans leur analyse h un mod8le de regression 3 effets fixes. La variable dd-pendante est dEfinie comme 6tant la variation du logarithme naturel de la rEmun6ration totale du chef de la direction. Les r6sultats obtenus par les auteurs rdvirlent une intensification apprEciable de Is relation entre la variation de la rbmunbration du chef de la direction et les revendications 1) des clients, 2) des actionnaires et 3) des employds, entre la p8riode prEcEdant I'adoption et la p6riode suivant ]'adoption. Les rEsultats montrent igalentent une intensification non significative de la relation entre la rdmun6ration du chef de Is direction et les fournisseurs de 1'entreprise. De plus, au chapitre de l'importance relative des intfatsts des parties prenantes principales, les int6r8ts des actionnaires ont pr6s6ance sur ceux des clients, suivis de ceux des employf,s. L'6tude des auteurs dE.montre donc que 1'entreprise tente non seulement d'harmoniser les int6a+8ts des gestionnaires avec ceux des actionnaires, mais aussi avec ceux des autres parties prenantes principales - clients et employEs.

1. Introduction Fundamentally, most empirical economics-based research on executive compensation presumes that executive compensation plans are adopted to align the interests of management and shareholders. Although prior accounting literature documents a significant association between changes in chief executive officer (CEO) compensation and corporate share price performance (e.g., Murphy 1986), the association between CEO compensation and the claims of other stakeholders of the firm has been largely ignored. Management literature identifies stakeholders as persons or groups who have or claim ownership rights or interests in a corporation and its activities. Clarkson (1995) states that a primary stakeholder group is one without whose continuing cooperation the corporation cannot survive as a going concern. Primary stakeholders are defined as customers, employees, suppliers, and shareholders. Balancing stakeholder goals is a complicated process given the range of needs and conflicting expectations that stakeholders have of an organization. Given his or her position at the apex of the corporate hierarchy, the CEO is the most likely to be concerned with the decision-setting task of balancing the varied and conflicting set of demands of the stakeholders (Andrews 1971; Lenz and Lyles 1986). Top managers, because they have "strategic position" (Herman 1981) with regard to key decisions of the firm, can be considered contracting agents of the firm. Thus, the firm can be viewed as a "nexus of contracts" between its top managers and its stakeholders (Jones 1995). The purpose of this study is to explore whether the adoption of a long-term incentive performance plan (LTIP) aligns the relative interests of the primary stakeholders of the firm with the interests of the CEO (the top manager).

Long-term incentive performance plans' were not widely used by firms until the early 1970s. Available evidence indicates that performance incentive plans CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-102

4982

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 4 of 30

522

Contemporary Accounting Research

have not replaced or weakened the existing compensation structure (salary, bonus, perquisites, and stock options) but have been added to the existing executive compensation package (Kumar and Sopariwala 1992). New compensation schemes are likely to be adopted by companies that are not satisfied that their existing compensation structure provides their managers with appropriate incentives to maximize shareholder wealth (Brozobsky and Sopariwala 1995). The traditional view that corporations exist solely to serve the interests of the firm's shareholders (Friedman 1962) has given way to a changing view that recognizes the importance of corporate constituents in addition to shareholders. In 1950, General Robert Wood, who led Sears' rapid postwar growth, listed the "four parties to any business in order of their importance" as "customers, employees, community, and stockholders" (Preston 1990, 1). He maintained that if the appropriate needs and interests of the first three groups were looked after effectively, the company's shareholders would be beneficiaries. In line with this research, we believe that in maximizing shareholder wealth, a dynamic CEO cannot ignore the relative interest of other primary stakeholders - namely, customers, employees, and suppliers.2 We investigate whether a newly adopted performance incentive plan aligns the interests of the CEO and the complementary but differential interest of the primary stakeholders. We hypothesize that the association between CEO compensation and the firm's primary stakeholders (defined as customers, employees, suppliers, and shareholders) is significantly higher in the post-adoption period than in the pre-adoption period. A fixed-effect regression model is used in the analysis. The dependent variable is defined as the change in the natural log of total CEO compensation. Our results show a significant increase from the pre- to the post-adoption period in the association between the change in CEO compensation and the claims of customers, shareholders, and employees. Our results also document an insignificant increase in the association between CEO compensation and the firm's suppliers. Moreover, in terms of the relative importance of the interest of the primary stakeholders, the interest of the shareholders takes precedence over that of customers, followed by the employees' interest.

The remainder of the paper is organized as follows. Section 2 presents prior literature and develops the hypothesis. The procedure used to obtain the sample is described in section 3. Section 4 specifies the empirical model that is used in the analysis and the measure of CEO compensation. In section 5, we describe the nature of the primary stakeholders' claims. Results of the analysis are given in section 6. Section 7 presents specification tests and section 8 presents the conclusion. 2. Prior literature and development of research hypothesis Freeman (1984) indicates that stakeholders include any constituency that is affected by the effort to achieve the firm's objectives. The stakeholder concept is widely accepted in management literature. Corporate laws have been changed in 12 states to include the stakeholder perspective (Preston and Sapienza 1990). The stakeholder theory indicates that the CEO balances the interests of various constituencies and manages responsibilities to various stakeholder groups, and the CEO's decisions reflect the varying degrees of importance assigned to different stakeholder CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-103

4983

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 5 of 30

CEO Compensation and Stakeholders' Claims

523

groups (Ansoff 1984; Freeman 1984). Lerner and Fryxell (1994) argue that firms adopting the stakeholder perspective may actually enhance firm performance. Thus, a firm intending to stay afloat as a going concern must continually address the interest of primary stakeholders. On the basis of a 10-year research program, Clarkson (1995) developed a stakeholder framework for analyzing and evaluating corporate social performance. Clarkson defines a firm's stakeholders as primary stakeholders and secondary stakeholders. He states that the corporation's survival and continuing success depend on the ability of the managers to create sufficient wealth, value, or satisfaction of the primary stakeholder group. There is a high degree of interdependence between the corporation and the primary stakeholders. If a primary stakeholder group becomes dissatisfied and withdraws from the corporate system, in whole or in part, the corporation will be seriously damaged or unable to continue as a going concern. Primary stakeholders include employees, customers, suppliers, and shareholders. Secondary stakeholders are defined as individuals or groups that influence or affect the corporation, but are not essential for its survival. Freeman's 1984 landmark work has provided a solid foundation for defining and building stakeholder model and theories. Freeman argues that a firm is characterized by relationships with many groups and individuals. Each group or individual has (1) the power to affect the firm's performance, and/or (2) a stake in the firm's performance. In many cases both conditions apply. Thus, a firm's stakeholders include, but are not limited to, shareholders. Eisenhardt (1989) opines that a contract is an appropriate representation of the relationship between the firm and its stakeholders. Ample precedent exists for a broad definition of "contract" (Dunfee 1991; Jensen and Meckling 1976; Ross 1973; Williamson 1984, 1985). Contracts can take the form of exchanges, transactions, or the delegation of decision-making authority, as well as legal documents. The firm can thus be seen as a "nexus of contracts" (Jensen and Meckling 1976) between itself and its stakeholders (Jones 1995). Top managers contract with stakeholders on behalf of the firm. The compensation contract of the top managers signals the ability of the two parties, stakeholders and managers, to agree on appropriate compensation for managers. Unusually high executive compensation is suggestive of inefficient contracting (Jones 1995). Preston and Sapienza (1990) report that major corporations strive to achieve a satisfactory level of performance for all primary stakeholders. This is usually achieved by negotiating efficient contracts with the primary stakeholders. Jones (1995) states that top managers, because they (1) contract with all stakeholders either directly or indirectly through their agents and (2) have a "strategic position" (Herman 1981) regarding key decisions of the firm, can be considered the contracting agents for the firm. The firm is thus recast as a nexus of contracts between its top managers and its stakeholders. The task of balancing stakeholder goals is a complicated process. Stakeholders have a variety of needs and, often, conflicting expectations (Lerner and Fryxell 1994). The CEO is most likely to be concerned with the direction-setting task of balancing this varied and conflicting set of demands (Andrews 1971; Lenz and Lyles 1986). Thus, for long-term success, CEO compensation contracts should align the interests of the CEO with the interests of the primary stakeholders .3 CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP_1-104

4984

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 6 of 30

524

Contemporary Accounting Research

Agency literature describes how managerial interests are aligned with the interests of shareholders (a primary stakeholder group) through the design of executive compensation contracts. However, extant literature ignores the relation between CEO compensation and the implicit claims of the customers, employees, and suppliers, also referred to as primary stakeholders. Top managers, including the CEO, increasingly operate from a perspective that recognizes the importance of managing or balancing the needs and demands of a variety of stakeholders. The extent to which management responds to and fulfills stakeholder needs is a measure of firm performance (Freeman 1984; Hambrick 1989). We contend that boards of directors consider the competing interests of primary stakeholders (explicit and implicit claimants) in designing executive compensation contracts.

Thus, the adoption of LTIP is expected to significantly increase the association between CEO compensation and empirical proxies measuring implicit and explicit stakeholder claims of the primary stakeholder groups from the pre- to the postadoption period. Before the introduction of performance incentive plans, executive compensation consisted primarily of salary, short-term bonus plans, and stock option plans (Fox 1980; Cook & Co. 1981; Peat, Marwick, Main & Co. 1982). Since 1970, many firms have added LTIP to their existing compensation schemes. One reason for this is that the board of directors, believing that no one performance measure could provide all the necessary incentives to align managerial and stakeholder interest, added LTIP to the existing compensation package to make it more reflective of the managers' actions. Recent studies examining the effect of performance incentive plans have found mixed evidence regarding an association between performance incentive plan adoption and post-adoption capital expenditures (Larcker 1983; Gaver and Gaver 1993; Enis 1993), as well as mixed evidence regarding the stock market reaction to the announcement of performance incentive plan adoption (Larcker 1983; Brickley et al. 1985; Gaver, Gaver, and Battistel 1992; Kumar and Sopariwala 1992). Recent studies have also found that companies with performance incentive plans achieve positive excess returns on announcements of mergers and selloffs (Tehranian, Travlos, and Waegelein 1987a, 1987b). Brozobsky and Sopariwala (1995) found that performance incentive plans have been adopted primarily because low growth in market value may not have afforded the executives of performance incentive plan adopting companies adequate market-based long-term compensation, and that performance plans are likely to be adopted by firms that have older managers who underinvest in research and development (R&D) and have limited investment opportunity sets.

One may argue that managers may be able to exercise significant influence over their compensation contracts by initiating or strongly endorsing the adoption of an LTIP when managers believe that the future prospects of the firm are good. However, this management strategy may not work if the compensation committees and the boards of directors duly recognize that the interest of the firm is best served when the competing interest of the primary stakeholders is protected. To examine whether adoption of the LTIP was exogenous to the model, the proxy statements of the 68 sample firms were examined. For all 68 firms the executive CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-105

4985

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 7 of 30

CEO Compensation and Stakeholders' Claims

525

compensation committee was composed only of external directors, and the committee met annually to evaluate the CEO and other executives. In order to determine whether the LTIP should be adopted, the committee reviewed information provided by an independent consultant. The consultant was appointed to determine the competitiveness of the CEO's total compensation package with that of a peer group. In a1.168 cases, one of the reasons given for adopting the LTIP was that the base salary and long-term incentives of the CEO were below the median of the peer group. Hence, we believe that the managers' self-dealing is limited under the existing corporate governance structure. Thus, we treat the adoption of LTIP as exogenous to our stakeholder model. Consistent with prior literature (see Pavlik et al. 1993), we assume that firms adopt performance incentive plans to better align the interests of the manager and the shareholders. In order to maximize stakeholders' return, a CEO must increase stock returns, satisfy consumer needs, encourage optimal employee performance, and ensure an adequate supply of materials to meet production and/or sales needs. Thus, subsequent to the adoption of LTIP, CEO compensation will be aligned not only with shareholder return but also with variables measuring customers, employees, and suppliers' claims. Our hypothesis, stated in the alternate form, is as follows: HYPOTHESIS. For firms that adopt a long-term incentive plan, the association between CEO compensation and the claims of primary stakeholders in the post-adoption period is significantly higher than in the pre-adoption period. 3. Sample selection This section details the selection and elimination process followed to identify the sample of firms used in this study. The step-wise process used in selecting the sample is outlined in Table 1. An initial sample of firms adopting performance incentive plans was compiled by conducting a keyword search of proxy statements on the LEXIS Financial Information Database. The time frame for executive compensation plan adoption examined was 1988 to 1993. The keywords used to identify sample firms were "CEO compensation" and "adoption of performance incentive plans". This search resulted in a total of 289 observations. To ensure that results from the analysis were not spurious, each of the 289 observations was classified as either "clean" or "contaminated". An observation was "clean" if the proxy statement contained no sponsored agenda item other than the proposed adoption of a performance incentive plan, or the election of directors and/or the ratification of auditots.4 An observation was considered "contaminated" if (1) the proxy statement included any sponsored agenda item other than those identified under a "clean" plan; (2) the firm adopted more than one type of executive compensation plan; (3) the firm modified an existing compensation plan while simultaneously adopting a performance incentive plan or adopting or modifying a stock-based compensation plan; (4) the firm adopted any plan or modified any existing compensation plan for a period of three years prior to the adoption of a performance incentive plan; or (5) the firm adopted any plan or modified any existing plan CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-106

4986

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 8 of 30

Contemporary Accounting Research

526

for a period of three years subsequent to the adoption of a performance incentive plan. Almost all proxy statements included agenda items to elect directors and to ratify auditors; some proxy statements also included shareholder proposals. Because shareholder proposals are usually defeated unanimously, such proposals were not classified as contaminated. If the firm changed the CEO, the previous year's compensation data would relate to a different CEO. Thus, firms that changed CEOs were excluded from the analysis because the level of CEO compensation is determined by a variety of CEO-specific factors in addition to performance (Sloan 1993). Thus, to ensure that the analysis captured the effect of performance incentive plan adoption on the compensation-stakeholder association, the proxy statement and the Wall Street Journal Index (WSJI) were scrutinized for other contaminating events (merger or a change in the CEO). A firm was excluded from the analysis if in the four years preceding the adoption, or for a period of three years subsequent to the adoption of the performance incentive plan, the firm either changed the CEO or participated in a merger. An examination of proxy statements and WSJI for the calendar years 1984 through 1996 provided 185 contaminated observations and 104 clean observations. TABLE 1 Sample selection criteria and sample size, 1988-93 Description Initial keyword search on Lexis Contaminated plan' Clean plant Plan adopters with insufficient data Final sample#

Number of firms

Firm years

289 185 104 36 68

1,750 1,065 685 209 476

Notes: A plan was considered contaminated if (1) the proxy statement included any other board-sponsored agenda item other than identified under a "clean" plan; (2) the firm adopted more than one plan; (3) the firm modified an existing plan simultaneously with the new plan adoption; (4) the firm adopted another compensation plan or modified an existing plan within a period of four years prior to and three years subsequent to new plan adoption; (5) the firm changed the CEO within a period of four years prior to and three years subsequent to new plan adoption; or (6) the firm was involved in a merger within a period of four years prior to and three years subsequent to new plan adoption. A plan was considered clean if the proxy statement included no other board-sponsored agenda item except the adoption of a proposed compensation plan, the election of directors, or the ratification of auditors. Final sample firms were required to have (1) COMPUSTAT annual data for a period of four years prior to plan adoption and for a period of three years subsequent to plan adoption and (2) CEO compensation data. CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP_1-107

4987

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 9 of 30

CEO Compensation and Stakeholders' Claims

527

To analyze the compensation-stakeholder relation in the post-adoption period, data from a period of three years subsequent to the adoption of the performance incentive plan were compared with data from three years prior to the adoption of the performance incentive plan. The compensation variable used in the analysis (described in section 4) is the change in CEO compensation. This requires compensation data to be available for four years prior to the adoption of the performance incentive plan and for three years subsequent to the adoption. CEO compensation data were obtained from either the Forbes annual compensation survey or the firm's proxy statements. Data on perquisite consumption were obtained from the Forbes annual compensation survey, the firm's 10K statement, or the firm's proxy statements. The financial press rarely discloses adoption of managerial compensation plans. The event dates examined in earlier studies (e.g., Brickley et at. 1985; Gaver et at. 1992) included the board of directors' meeting date, the Securities and Exchange Commission (SEC) stamp date, and the shareholders' meeting date. The LEXIS data base reports the SEC stamp date and the shareholders' meeting date. This study identifies the shareholders' meeting date as the date of the performance incentive plan adoption. Annual data for a period of three years preceding the shareholders' meeting date were classified as data from the pre-adoption period. Annual data for a period of three years subsequent to the shareholders' meeting date were categorized as data from the post-adoption period. Annual data were obtained from COMPUSTAT. For the period 1988-93, 68 firms met our sample selection criteria .5 Similarly, a change in CEO compensation requires compensation data to be available for three years prior to the adoption of the performance incentive plan as well as for three years subsequent to the adoption. Thus, a sample size of 476 firm-years was used to compute the dependent variable change in CEO compensation, and a sample size of 408 um-years was used to estimate the regression parameters.

Selection of control firms Similar to the Larcker 1983 study, control firms were matched with treatment firms on the basis of the following criteria: (1) similar industry (Standard Industrial Classification [SIC] code), (2) similar size (measured by corporate sales in the year prior to adoption of a performance incentive plan by the treatment firm),6 (3) no formal adoption of a performance incentive plan over the test period, (4) similar fiscal year to the treatment firm, (5) no participation in a merger over the test period, and (6) no change in the CEO over the test period. Table 2 presents an industry profile of the firms in our treatment and control samples. Given the constraints, we found 29 matches at a four-digit SIC level, 14 matches at a three-digit SIC level, and 25 matches at a two-digit SIC level. The sample is distributed over 19 industries. There is no indication of industry or event clustering in any SIC group.

Table 3 presents statistics for total sales (in millions of dollars) and leverage for the treatment and control sample. Similar to Larcker's 1983 study, leverage is defined as long-term debt divided by market value of equity. Long-term debt is measured by book value. Market value of equity is computed by adding (1) the CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP-1-108

4988

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 10 of 30

528

Contemporary Accounting Research

average of high and low common stock price over the fiscal year multiplied by the average number of shares of common stock during the fiscal year and (2) the book value of long-term debt. The summary statistics for sales indicate that treatment firms are larger than control firms, but control firms are more leveraged than treatment firms. However, these differences are not statistically significant. 4. Model specification With pooled data, the assumption of constant intercept and slope may not be appropriate when using ordinary least squares (OLS). Pindyck and Rubinfeld (1991) suggest using the covariance model (also known as fixed-effect regression) in such situations. Thus, dummy variables for time and firm effect are added to the OLS model. The number of degrees of freedom, however, is considerably reduced in a fixed-effect model: NT - N - T, where N represents the individual firm and T represents the number of time periods. In this study, to investigate the potential shift in the association between CEO compensation and stakeholder wealth from the pre- to the post-adoption period, the fixed-effect regression model estimated for the treatment and control samples (firm and year coefficients not shown) is as follows: TABLE 2 Distribution of pair-wise industry matches between treatment and control firms

wo-digit industry codes

Number of two-digit matched pairs

Number of three-digit matched pairs

Number of four-digit matched pairs

Total number of matched pairs

10 (Mining) 13 (Oil and gas) 16 (Heavy construction) 20 (Food manufacturing) 21 (Tobacco manufacturing) 22 (Textile production) 26 (Paper) 27 (Printing) 28 (Chemical) 29 (Petroleum) 30 (Footwear) 32 (Clay and glass) 35 (Machinery) 36 (Electrical equipment) 37 (Transportation equipment) 38 (Medical equipment)

2 1 1 1 2 2 2 2 2 1 1 1 0 0 1 3

0 0 1 2 0 0 0 1 0 0 1 2 0 3 0 1

0 4 1 1 0 0 1 2 2 2 3 1 3 2 1 1

2 5 3 4 2 2 3 5 4 3 5 4 3 5 2 5

40 (Rail transportation)

1

2

1

4

1 0 14

1 3 29

4 3 68

48 (Broadcasting) 50 (Wholesale trade) Total

2 0 25

CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-109

4989

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 11 of 30

CEO Compensation and Stakeholders' Claims COMPtt = Po +P 1Dit +P2STAKE,t +sg3Dit STAKE;t + e;t

529 (1),

where COMP,t = change in CEO compensation for firm i in period t, described below; STAKE;t = stakeholder variables for i in period t, described further below;

Dit

= 0 if firm i's data are from the period prior to the shareholders' meeting date;

D;t

= 1 if firm i's data are from the period subsequent to the shareholders' meeting date; and

E;t

= a random disturbance term for firm i in period t.

From the specification of (1), P2 is the compensation-stakeholder association coefficient for the pre-adoption period and P2 + P3 is the compensation-stakeholder association coefficient for the post-adoption period. For firms adopting performance incentive plans, our hypothesis predicts that the post-adoption compensationstakeholder association will be larger than the pre-adoption association. Hence, ceteris paribus, Pg captures the incremental effect of the plan adoption? On the basis of our hypothesis, we expect P3 > 0. In this study, if a firm adopted a performance incentive plan in 1988, the pre-adoption period covered the years 1985-87 and the post-adoption period included the 1989-91 period, and so on. Thus, the 1984-87 period is always in the pre-adoption period, while the 1993-96 years are always in the post-adoption period, and the 1988-92 years can be in either period depending on the adoption year. To control for the effects of overall trends in the economy, all numeric values used to analyze stakeholder variables (described TABLE 3 Summary statistics on total sales and leverage

Sales (in millions of dollars) Treatment Control Leverage* Treatment Control

n

Mean

Standard deviation

t-value

68 68

821.29 776.54

576.71 498.23

0.47

68 68

0.06 0.08

0.12 0.10

0.73

Note: '

Leverage is defined as long-term debt divided by market value of equity. Long-term debt is measured by book value. Market value of equity is computed by adding (1) the average of high and low common stock price multiplied by the average number of shares of common stock during the fiscal year and (2) the book value of long-term debt. CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-110

4990

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 12 of 30

530

Contemporary Accounting Research

below) are measured in terms of constant 1990 dollars prior to computing the change variables used in regression analysis. Measure of CEO compensation Subsequent to the adoption of LTIP, a CEO is rewarded for improved accountingbased performance through the allocation of shares of stock. Thus, long-term incentives are an important component of the CEO's compensation package. Bushman et al. ( 1998) report that data from Hewitt Associates indicated that long-term compensation comprises an average of 35 percent of total executive compensation over the period 1987-94. The dependent variable in the regression analysis, COMP, is computed as the change in the natural logarithm of total CEO compensation for firm i in year t. The total CEO compensation variable is defined as the sum of annual salary, annual bonus, stock gains, and other compensation (for example, perquisites) received by the CEO.

5. Nature of primary stakeholders' claims Explicit claimants on the firm's resources are the shareholders who have the legal right to the firm's net assets. Prior literature considers customers, employees, and suppliers as implicit claimants on the firm's revenues (Bowen, DuCharme, and Shores 1995). Although implied commitments cannot legally be enforced and may be breached by either party, they may be viewed as self-enforcing relational contracts (MacNeil 1978). Tesler (1980) and Bull (1987) state that firms self-enforce these contracts because there is an implied promise, which if reneged upon by the firm could cause serious damage to the firm's reputation. Maunders and Foley (1974), Cornell and Shapiro (1987), and Maksimovic and Titman (1991) suggest that accounting numbers are used by shareholders in their assessment of the firm's ability to fulfill its implicit claims. Customers Generally, customers purchasing products are also purchasing implicit claims related to a specified quality of performance and availability of continuing parts and service over the life of a product. In today's global economy, a firm's longterm growth and survival depend on a dedication to R&D and a strategy that adopts technological advances to deliver quality products and services with a reputation for value (Clarkson 1995; Lerner and Fryxell 1994; Sougiannas 1994). Bowen et al. (1995) and Titman and Wessels (1988) identify uniqueness of a product as a characteristic that is likely to be indicative of implicit claims related to ongoing relations between a firm and its customers. In particular, they argue that it is likely to be difficult for customers to find alternative servicing for relatively unique products and that employees of firms with unique products are likely to have job-specific skills. Consistent with these studies, we use R&D intensity (RDI) as a proxy for uniqueness of a product. Thus, RDI, defined as R&D expenditures of firm i for period t minus R&D expenditures of the firm for year t-1, divided by net sales for year t-1, is identified as the proxy for customers' implicit claims.

Results presented in Table 4, panel A reveal a statistically significant increase in RDI from the pre- to post-adoption period for firms that adopt performance incentive CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-1 11

4991

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 13 of 30

CEO Compensation and Stakeholders' Claims

531

plans. Table 4, panel B reveals no statistically significant differences in RDI from the pre- to the post-adoption period for control firms. These results provide preliminary evidence that is consistent with the notion that the adoption of a performance incentive plan is successful in aligning the interests of managers and customers. Employees The success of a CEO depends on his or her ability to attract and retain the "best" employees. Employees are implicit stakeholders, and mutually trusting, cooperative relationships between a firm and its employees provide a competitive advantage (Jones 1995). In general, employees who are paid well, who are given continuous education or training, and who receive good benefits are likely to be more productive (Clarkson 1995; Lerner and Fryxell 1994). The proxy for employees, LABOR, is measured as the number of employees of firm i for period t minus the number of employees of the firm for year t - 1, divided by the number of employees for year t - 1. Results presented in Table 4, panel A reveal a statistically significant increase in LABOR from the pre- to the post-adoption period. Table 4, panel B reveals no statistically significant differences from the pre- to post-adoption period for control firms. These results are consistent with the notion that the adoption of the performance incentive plan is successful in aligning the interests of managers and employees. Suppliers The "community" as a stakeholder includes a wide variety of participants interested in the firm. We chose to use "suppliers" to represent "community". The level of mutual trust and cooperation between a firm and its suppliers can significantly affect its cost structure (Jones 1995). In general, the implicit claims of suppliers on a firm are based on timely payment and continuing demand for the suppliers' product (Clarkson 1995). Although firms have relations with a variety of suppliers, they are likely to have significant implicit claims with suppliers of inventory (Bowen et al. 1995). Our proxy for the implicit claims of suppliers, APT, is defined as accounts payable turnover of firm i for period t minus accounts payable turnover of the firm for year t-1, divided by accounts payable turnover for year t-1. Accounts payable turnover is defined as the cost of goods sold divided by the average accounts payable. Results presented in Table 4, panel A, as well as results presented in Table 4, panel B, do not reveal any statistically significant difference from the pre- to the post-adoption period for either the treatment firms or the control firms.

Shareholders Several accounting researchers have examined the role and efficacy of performance incentive plans on shareholder wealth (see Pavlik et al. 1993). Kumar and Sopariwala (1992) state that new compensation schemes (for example, LTIP) are likely to be adopted by companies that are not satisfied with their existing compensation structure. In addition, a positive stock market reaction to an announcement of a performance incentive plan adoption is interpreted as an improvement in shareholder wealth, an improvement in contracting between managers and shareholders, or a reduction in the agency problem (Larcker 1983; Brickley et al. 1985; Kumar and Sopariwala 1992). CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-1 12

4992

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 14 of 30

Contemporary Accounting Research

532

eo O ^

.!.

C In

O ,^o,

ry u 0

0% N O O N1 ^;

^ O :. ..Mr

C C C O O O -• fr1

as 0o 00 83 00 'N 00 00

V

o

0o

00

00

O^

C

^N

l^

00

00

o^ 28 a2 N;; c0 cc cc cc

0

> ^ pq •

Q ^

o

^ °

tI1Ii U llll

i

r^o U

W

N

^

CA i\ ol. 22 Ni P. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-113

4993

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 15 of 30

5; 33

CEO Compensation and Stakeholders' Claims

117, "^,^ t^ CG

^ y^

^ C O ^

C

O

N N O RN OoO CO CC

C C u ^

i: ^

te+,'^i-^SIN^ NON^

E ° C

00

^ v

0o

00

II .^

Q

= '^

O

^

11

C

N

ciaN

C30 O- 00 00

00

.S

V (^

V^ •^

u .5

.•^ V1

.•^ M .•^ ^G

^^8a 00 00 00 0- oc

•^ ^^

y

C^1 ^O

v

9 n g8=N°

o^ o0 00 00 0^ o-^ A

^:,g8Ngo^ ^00 ^ A NB•^&o`^ o0 o•^ o0 00

3S ^

00 00

w

IM =MAsN^^^^s^g$^go^g 06

It ac

00

60

O

d

O

O

O

O

O

O

^ C ^= b

.V..

.a o

U

u

_ [-^ a >

so ° ci a a

•

A x

a o

CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-1-1 14

4994

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 16 of 30

534

Contemporary Accounting Research .. b:Z:

.. ^p ^ v^`iON 66

•O

M8H -4 o

u a

-6

1

^OO^OOONN O C 60 cc

i H

0 N^O^NO^O1g-N 8O^ 00 C^+ C^

E ^O 00 00 0- c--^ o-

8^$8&$8^g8^g8^A 8og O r O O C-^ O o O O o -

V1-

V1

+N

0o 0o

o--

NO00 ^+

en 00M N

o0 00 0-; do

. Uv

a

9

Q V

en 00

8 d N N O^^ 800 d^^ O^ O 9 OknO N O O^^00 t "1 00 00 00 00 0-• o0 00

_ a

u

^Z O

Q p ^p ^ ^^ p s0 U^ ^ W^ ^ ^^• Q y^ V

CA 3ol. 22 N o. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-115

4995

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 17 of 30

CEO Compensation and Stakeholders' Claims

535

TABLE 4 (Continued) Notes: Proxies for stakeholders are: l. customers (RD!), defined as the R&D expenditures of firm i for period t minus the R&D expenditures of the firm for year t-1, divided by net sales for year t-1; 2. employees (LABOR), measured as the number of employees of firm i for period t minus the number of employees of the firm for year t-1, divided by the number of employees for year t-1; 3. suppliers (APT), defined as accounts payable turnover of firm i for period t minus accounts payable turnover of the firm for year t-1, divided by accounts payable turnover for year t-1; accounts payable turnover is defined as the cost of goods sold divided by the average accounts payable; and 4. shareholders (GROWTH), defined as the market value of firm i for year t minus the market value of the firm for year t-1, divided by the market value for year t-1. Each cell presents the Pearson correlation, its associated p-value, and the number of observations. D = I in the period subsequent to adoption of the performance incentive plan and D = 0 in the period prior to plan adoption. Primary stakeholders are defined as customers, employees, suppliers, and shareholders. All numeric values used to analyze the stakeholder variables are measured in terms of 1990 constant dollars prior to computing the change variables. Significant at the 0.001 level.

In general, the interest of shareholders is served if the market value of the firm increases (Lerner and Fryxell 1994). Kumar and Sopariwala (1992) indicate that adopting LTIP encourages managers to make strenuous efforts on behalf of the company to increase shareholder wealth. Similar to the proxy used in Kumar and Sopariwala 1992, the proxy to capture the effect of the adoption of a long-term performance incentive plan on shareholder wealth is growth in the market value of the firm. Market value of firm i is defined as the number of outstanding common shares at the end of the fiscal year t times the market price of the common shares at the end of the year. Growth in market value, GROWTH, is defined as the market value of firm i for year t minus the market value of the firm for year t-1, divided by the market value for year t-1. Results presented in Table 4, panel A show a statistically significant increase from the pre- to the post-adoption period in the growth in the market value for firms that adopt long-term performance incentive plans. Table 4, panel B reveals no statistically significant differences from the pre- to the post-adoption period for control firms. These results are consistent with the notion that the adoption of the long-term performance incentive plan is successful in aligning the interests of managers and shareholders. Table 4, panel C presents the correlation coefficients among the stakeholder variables for the treatment firms. Table 4, panel D presents the correlation coefficients among the stakeholder variables for the control group. CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP_1-116

4996

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 18 of 30

536

Contemporary Accounting Research

In both tables, the pair-wise correlations are not significant at either the 0.05 or 0.01 level, which suggests the absence of multicollinearity. 6. Results Table 5, panel A reports parameter estimates for regression (1) for the treatment firms (firm and year effect coefficients not shown). In Table 5, panel A, column 2, STAKE is defined as the claims of the primary stakeholder group, customers. The proxy used to represent the implicit claims of customers is RDI. Results from this column indicate that P3 is significantly positive (83 = 2.11, t-value = 4.54). These

TABLE 5 Regression analysis of change in CEO compensation on variables representing primary stockholders (sample size = 408 observations) COMP =flo+P ID +P 2STAKE+P3DSfAKE+ e Panel A: Treatment firms Coefficient ^p All P2 P3 F-value Adj. R2

Shareholders (GROWTH)

Customers (RD!)

Employees (LABOR)

Suppliers (APT)

0.12 (1.04) 0.35 (1.37) 0.65 (0.77) 2.11 (4.54)' 5.46 0.21

0.09 (1.54) 0.34 (0.67) 0.16 (1.09) 1.73 (5.64)' 5.86 0.06

0.01 (0.86) 0.64 (1.11) 0.07 (0.59) 0.34 (0.11) 0.58 0.01

0.34 (0.65) 0.43 (1.18) 2.18 (3.84)' 7.98 (9.76)' 10.62 0.27

Customers (RDI)

Employees (LABOR)

Suppliers (APT)

Shareholders (GROWTH)

0.02 (1.24) 1.15 (1.47) 0.45 (0.79) 0.51 (1.22) 1.56 0.03

0.01 (1.12) 0.06 (0.61) 0.13 (1.17) 0.25 (0.74) 0.91 0.01

0.02 (1.35) 0.03 (0.17) 0.07 (0.63) 0.16 (0.76) 1.68 0.00

0.16 (0.58) 1.24 (1.10) 1.18 (1.64) 1.29 (1.17) 1.92 0.04

Panel B: Control firms Coefficient AD A P2 P3 F-value Adj. R2

(Me table is continued on the next page.)

CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP-1-1 17

4997

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 19 of 30

CEO Compensation and Stakeholders' Claims

537

TABLE 5 (Continued) Panel C: Matched-pair sample Coefficient

j3p

P2 P3 F-value Adj. R2

Customers (RDI)

0.00 (0.01) 0.05 (1.07) 0.11 (1.02) 1.02 (2.89)' 7.23 0.04

Employees (LABOR)

0.12 (1.04) 0.04 (0.72) 0.09 (1.07) 0.53 (3.98)' 6.18 0.03

Suppliers (APT)

0.00 (0.77) 0.02 (0.43) 0.01 (0.61) 0.07 (1.21) 1.18 0.00

Shareholders (GROWTH)

0.04 (0.91) 0.58 (0.79) 0.85 (1.59) 3.42 (4.38)' 6.11 0.09

Notes: COMP is computed as the change in the natural logarithm of the total CEO compensation of firm i in year t. D=1 in the period subsequent to adoption of the performance incentive plan and 0 in the period prior to plan adoption. Primary stakeholders are defined as the customers, employees, suppliers, and shareholders. All numeric values used to analyze the stakeholder variables are measured in terms of constant 1990 dollars prior to computing the change variables. Proxies for stakeholders customers (RDI), employees (LABOR), and suppliers (APT) are as defined in Table 4. The proxy for shareholders (GROWTH) is computed as (1) the market value of firm i, defined as the number of outstanding common stock at the end of the fiscal year t times the market price of common stock at the end of the year; and (2) GROWTH, defined as the market value of firm i for year t minus the market value of the firm for year t-1, divided by the market value for year t-1. In panel C, the dependent and independent variables in the regression are computed by subtracting the variable attributable to the control firm from the corresponding variable associated with the treatment firm. The number of observations in the preadoption period is 204, and the number of observations in the post-adoption period is also 204. The number of degrees of freedom (334) in this estimation is computed as (NT - N - T), where N represents the individual firm and T represents the number of time periods. Significant at the 0.001 level. '

results indicate that the association between CEO compensation and the firm's customers significantly increased from the pre- to the post-adoption period .8 In Table 5, panel A, column 3, STAKE is defined as the claims of the primary stakeholder group, employees. The proxy used to represent the implicit claims of employees is LABOR. Results from this column indicate that P3 is significantly positive (83 = 1.73, t-value = 5.64).9 These results indicate that the association between CEO CAR Vol. 22 No. 3(Fall 2005)

2009 ETI Rate Case

WP_1-118

4998

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 20 of 30

538

Contemporary Accounting Research

compensation and the firm's employees significantly increased from the pre- to the post-adoption period. In Table 5, panel A, column 4, STAKE is defined as the claims of the primary stakeholder group, suppliers. The proxy used to represent the implicit claims of suppliers is APT. Results from this column indicate that P3 is positive but not significant (8g = 0.34, t-value = 0.11).t0 In Table 5, panel A, column 5, STAKE is defined as the claims of the primary stakeholder group shareholders. The proxy used to represent the implicit claims of shareholders is GROWTH. Results from this column indicate that Pg is significantly positive (8g = 7.98, t-value = 9.76). These results indicate that the association between CEO compensation and the firm's shareholders' interest significantly increased from the pre- to the post-adoption period for firms that adopted LTIP. The results also show that, among stakeholders, shareholders have the highest association with CEO compensation, followed by customers and finally employees (#shoreholders = 7.98, ^customen = 2.11, Pemployees = 1.73). This result establishes that CEO interest is mostly aligned with those of shareholders, followed by customers, and employees. Table 5, panel B presents parameter estimates for regression (1) for the control firms. Although Pg, the compensation-stakeholder association coefficient, is positive when stakeholders are defined as customers, employees, suppliers, or shareholders, 03 is not significant. Table 5, panel C presents parameter estimates for regression (1) for the matched-pair sample. Similar to Larcker's 1983 results, variables used in the regression model are computed by calculating the difference between each pair-wise observation. Specifically, the compensation variable and the stakeholder variable used for each regression model are computed by subtracting the variable attributable to the control firm from the variable attributable to the treatment firm. The dummy variable, D, equals 0 if the firm's data fell in the period prior to shareholders' meeting date, and D equals 1 if the firm's data fell in the period subsequent to the shareholders' meeting date. These results are consistent with those presented in Table 5, panel A and also show that, among stakeholders, shareholders have the highest association with CEO compensation, followed by customers and then employees (^sharcholdcrs = 3.42, ^costo^►urs =1.02, Pe loyeeS = 0.53). Moreover, these results establish that subsequent to the adoption of performance incentive plan, the CEO interest is mostly aligned with the interests of shareholders, followed by customers, and then employees.

Thus, when stakeholders are defined as customers, employees, and shareholders, the results are consistent with our hypothesis that for firms that adopt performance incentive plans, the compensation-stakeholder relationship is significantly stronger in the post-adoption period than in the pre-adoption period. However, when stakeholders are defined as suppliers, the results do not support our hypothesis. 7. Specification tests To examine whether our results are affected by extraneous, omitted, or alternative specifications of the dependent variable, we performed specification tests. For this purpose, we searched prior literature for variables that could influence the compensation-stakeholder relation. This section reports the results of these specification tests. CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP-I- 119

4999

WP/JCH Testimony-fn 8 2009 TX Rate Case Page 21 of 30

CEO Compensation and Stakeholders' Claims

539

Alternative specification of the compensation variable In prior literature (e.g., Sloan 1993), CEO compensation is often measured as the change in the natural log of salary and bonus compensation. To examine whether our results are robust with respect to this alternative specification, we estimated the parameters of (1) using the change in the natural log of salary and bonus compensation as the measure of the dependent variable. Our results (not presented) are similar to the results presented in Table 5. The P3 coefficient remains significant in the same manner for the treatment firms and insignificant for control firms.

Factors related to compensation plan adoption Prior literature shows that managers adjust accruals in reaction to bonus plans (Healy 1985; Scott 1991; Holthausen, Larker, and Sloan 1995), and compensation committees take an executive's personal stockholding in a firm's stock into consideration in the design of executive compensation contracts. Hagerman and Zmijewski (1979) found that managers' stockholding and firm size were important variables in managers' choice of income-increasing accounting methods. To assess whether discretionary accruals, CEO stockholding, firm size, and other factors (INTANGIBLE, TXCR, and CAP are defined in the following paragraph) are possible omitted variables, the parameters of the following fixed-effect regression are estimated (firm and year coefficients not shown):

COMP;t = Pp +P ID +WAKE +P3DSTAKE + PqDA,t +P5STOCK;, +PyS1ZE;r + p7INTANGIBI.Eir + psTXCRtr + /3OCAP;r + -cir (2), where (other variables as defined previously) DA,r

= discretionary accruals estimated using the Jones 1991 model for firm i in period r,

STOCK,t

= the dollar value of CEO stockholding divided by annual CEO salary and bonus compensation for firm i in year t•,

SIZE;r

= the natural logarithm of net sales for firm i in year t•,

INTANGIBLE;t = intangible assets divided by total assets for firm i in year t, TXCR,l

= 1 if the firm received tax credit for increasing R&D and 0 otherwise for firm i in year t; and

CAP;t

= 1 if the salaries of the CEO or other top executives exceed $1 million and equals 0 otherwise for firm i in year t.

The claims of the primary stakeholders, STAKE, and the dummy variable, D, are defined as above. To control for discretionary accruals, DA, the Jones 1991 model, is used to compute the discretionary accruals for sample firms. CEO stockholding, STOCK, is defined as the dollar value of CEO stockholding divided by annual CEO salary and bonus compensation. This measure is similar to the one used by Sloan 1993. In line with Bowen et at. 1995, to control for the size of the CAR Vol. 22 No. 3 (Fall 2005)

2009 ETI Rate Case

WP_1-120

5000