incentives, monitoring, and employee stock ownership plans: new evidence and interpretations

I

R

, Vol. 45, No. 4 (October 2006). © 2006 Regents of the University of California Published by Blackwell Publishing, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

753

Blackwell Publishing IncMalden, USAIRELIndustrial Relations: A Journal of Economy and Society0019-8676© 2006 Regents of the University of CaliforniaOctober 2006454Original Article

New Evidence and Interpretations

A

P

Incentives, Monitoring, and Employee Stock Ownership Plans: New Evidence and

Interpretations

ANDREW PENDLETON*

This paper reviews the theory and evidence for agency theory-based explana-tions for employee stock ownership plans found in the financial participationliterature. The UK Workplace Employee Relations Survey 1998 is used to testwhether share plans substitute for direct monitoring and individual incentives.Contrary to some predictions in the literature, individual incentives are foundto be complements of share plans, while other measures of monitoring costsprovide mixed results. However, it is found that monitoring costs and a widerange of performance targets explain the conjunction of stock plans and indi-vidual incentives. It is suggested that share plans are used to mitigate dysfunc-tional effects of individual incentives by engendering cooperation and trust,and by broadening the range and time frame of desired performance outcomes.

Introduction

A

research into employee financial participation over the last twenty years orso. The starting point has been the agency problem inherent in the employ-ment relationship, whereby the interests of utility-maximizing employees arenot congruent with those of the firm

1

and where they have some capacityfor discretionary behavior. It has been argued that collective remunerationsystems such as profit sharing and stock ownership plans can ameliorateagency costs by aligning employee interests with those of the firm. The use ofpropositions derived from agency theory pervades the small, though growing,

* Andrew Pendleton is Professor of Human Resource Management at the University of York,United Kingdom. E-mail:

[email protected]

. I am grateful to participants at the annual HRM sem-inar of the European Institute for Advanced Studies in Management for comments on an earlier versionof this paper. Two anonymous reviewers provided some extremely useful comments and criticisms ofprevious drafts of this paper.

1

The literature on all-employee financial participation tends to sidestep the issue of agency costs inthe relationship between owners and managers, though this is central to the literature on executive shareownership and options.

754

/

A

P

literature on the characteristics of firms or workplaces with financial partic-ipation (“determinants”) (see, for instance, Beatty 1994; Drago and Heywood1995; Kruse 1996; Jones and Pliskin 1997) and a second body of literatureconcerned with the impact of financial participation on performance (seeBlasi, Conte, and Kruse 1996; Jones, Kato, and Pliskin 1997; McNabb andWhitfield 1998; Addison and Belfield 2000; 2001; Sesil et al. 2002).

The current paper focuses on the first of these topics by investigating thedeterminants of the use of all-employee stock ownership plans in the UK.Stock plans are seen as those providing for stock-based rewards to employeesbut which in most cases do not lead to substantial employee ownership andcontrol of the firm. The financial participation literature typically makestwo predictions about the use of broad-based stock plans, drawing on agencytheory. The first is that employee share ownership is more likely to be usedwhere employee performance is costly to observe because of certain featuresof work organization and job design (see Cheadle 1990; Kruse 1996; Jones,Kato, and Pliskin 1997). The second is that group-based rewards, such asshare ownership plans, will be attractive to firms when individual incentive-based pay is costly to operate (Jones and Pliskin 1997). In this literaturestock plans tend to be viewed as an alternative to individual incentives.Recently, a new strand of thinking has emerged within the agency traditionthat argues that firms may use multiple incentives, and which thus impliesthat stock plans may complement individual incentives (Holmstrom andMilgrom 1994; Gibbons 1998). The argument is that employees typicallyhave multiple tasks, with varying degrees of measurability and focus, andthus a single incentive may have strong distortion effects (Prendergast1999). In this type of reasoning, stock plans may mitigate the costs arisingfrom the use of other pay instruments by encouraging employees to focuson a broader range of tasks and outcomes. A similar view is also found inthe “balanced scorecard” literature (Kaplan and Norton 1996). Here theemphasis is on providing targets and incentives that encourage long-termperspectives and an emphasis on performance in a broad sense rather thanjust short-term financial or output goals.

The paper assesses these explanations of the determinants of the use ofbroad-based stock ownership plans, and engages in particular with thequestion as to whether stock plans are substitutes or complements to indi-vidual incentives. In the first stage of the paper, the theoretical rationale inthe financial participation literature for the use of instruments such asbroad-based stock plans is considered, followed by a review of the empiricalevidence. This evidence is found to be wanting in three respects. First, theproxies typically used to explain why financial participation is used inpreference to other incentives do not generate very compelling results.


/ 755

Second, empirical evidence indicates that broad-based stock plans are usedin circumstances where their incentive effects are likely to be very weak(c.f. Oyer 2004). Third, this literature does not directly test the theorizedsubstitution between financial participation and individual incentives.

The paper then reports the results of an empirical examination of theseissues using the UK Workplace Employee Relations Survey (WERS) con-ducted in 1998. Initially, the prevailing approach in the financial participa-tion literature is replicated. Here, some results provide some support forpredictions based on monitoring costs but other findings clearly cast doubton the incentive effects of broad-based stock plans. A proxy is then includedfor individual incentives and it is found that these are a covariate ratherthan a substitute for stock plans. These findings lead us to doubt whetherbroad-based stock plans are widely used as a “high-powered” incentive. Toconsider the function of stock plans further, given these results, the deter-minants of the copresence of individual incentives and stock plans areinvestigated with reference to those for individual incentives alone. Theresults are consistent with the predictions of the “balanced scorecard” and“new agency” literatures. The explanations based on job complexity foundin the financial participation literature also find clearer support than wherefinancial participation is considered in isolation.

The interpretation of these results is that stock plans help to meet a perceivednecessity to achieve high performance through performance managementinstruments, and at the same time secure employee cooperation. They may doso by mitigating the costs arising from the use of other instruments that providestronger incentives. They may assist in promoting truthful informationexchanges in subjective evaluations of employee performance (i.e., performanceappraisal) while ameliorating the distortions found in individual payment byresults based on objective measures of performance. In a sense, the merits ofstock plans in these circumstances are precisely those that constitute theirlimitations when considered as an incentive in isolation. They shift the stimuliemanating from stronger incentives in the direction of longer-term time frames,collective performance outcomes, and more cooperative behavior.

We believe the paper adds to the literature in several ways. At a generallevel, it provides further empirical evidence on the determinants of the useof employee share ownership, and provides an answer to the question “whydo firms use incentives that have no incentive effects?” (Oyer 2004). Itprovides new empirical evidence on the relationship between employeeshare ownership and other components of remuneration. In addition, itgives empirical support for the views found in the “balanced scorecard” andrecent agency-based literatures that firms will aim to use a balanced “portfolio”of incentives and monitoring instruments to mitigate the dysfunctional

756

/

A

P

effects of single, high-powered incentives (Holmstrom and Milgrom 1994;Gibbons 1998, Prendergast 1999, Ittner, Larcker, and Meyer 2003b).

In the remainder of the paper, the predictions and evidence found in thefinancial participation literature are discussed. We then provide informationon the data source and empirical approach before presenting results.Finally, we interpret these results in the light of recent theory.

Background

The prevailing orthodoxy in studies of financial participation emphasizesits apparent capacity to lower agency costs where contracts are incomplete.Since workers have discretionary capacity to expend effort, in a context wherethe interests of workers (agents) and the firm (the principal) may diverge, thefirm faces costs arising from moral hazard and adverse selection (Holmstrom1979; Eisenhardt 1989). The upshot is, as McNabb and Whitfield put it,“workers can vary their effort, their co-operation with colleagues andmanagement, and their contribution to improving the effectiveness of boththeirs and related inputs” (1998: 173). Managers typically deal with theseproblems in two ways (the “carrot and the stick,” in McNabb and Whitfield’swords). The first is direct scrutiny of workers’ job performance via directsupervision. However, direct monitoring incurs costs, such as the setup andoperating costs of monitoring systems (i.e., supervisors’ salaries) or thosearising from a violation of trust (as the job enrichment literature highlights).Furthermore, disentangling the contribution of individual employees inteam production systems can be immensely difficult (Alchian and Demsetz1972). The second approach is to align remuneration with worker output sothat the interests of the firm and the worker converge, as in performance-basedpayment systems such as piecework. This is essentially an output-basedapproach to securing desired behavior.

The “personnel economics” literature suggests that firms will more likelyadopt systems for linking individual’s pay to outcomes (performance pay orpayment by results) as direct monitoring of worker effort becomes morecostly (Drago and Heywood 1995: 507; Heywood, Siebert, and Wei 1997).However, there are also costs to output-based pay, as is well documented inthe long-standing industrial relations literature on payment by results andperformance-related pay systems (Lupton 1963). These include “smoothing”of output, sub-optimal trade-offs of effort and rewards, manipulation ofperformance standards, manipulation of output, and goal displacement (seeBaker, Gibbons, and Murphy 1994; Gibbons 1998). The encouragement of low-trust relationships is widely seen as a major weakness of individual incentive


/ 757

pay, whether it takes the form of piecework for manual workers (Lupton1963) or merit pay for white-collar employees (Marsden and Richardson1994; Heery 1998). Since trust appears to be pivotal to “genuine” teamwork,individual piecework is generally inimical to teamwork (Drago and Heywood1995: 511). A further critical issue is whether individuals’ output and con-tribution is readily identifiable. Where technology or service delivery is suchthat individual work tasks and units of output are highly interdependent ordifficult to quantify, individual payment by results seems to be inappropriate.

In principle, stock ownership and profit sharing schemes may mitigateboth the costs of monitoring individual performance and the dysfunctionalside effects of individual incentive pay. It has been suggested by some thatprofit sharing or share ownership may be used in preference to individualincentives in a number of contexts. Work environments where misuse ofcapital equipment is a potential problem (through worker attempts to reachindividual targets and maximize earnings in a PBR setting), where outputis not easily attributable to individuals, where setting piece rates is costly(because production technology is changing or products change frequently),and where individual output is costly to measure may all prefer to use acollective form of incentives rather than individual payment by results(Brown 1990; Kruse 1996; Jones and Pliskin 1997). Furthermore, workenvironments where work quality is especially important may requireincentives that encourage high standards of employee performance but donot have a narrow emphasis on output levels (Brown 1990; Drago andHeywood 1995). Contemporary developments in work organization, suchas the growth of teamwork, may intensify monitoring costs and thus also leadto a growing use of financial participation (Applebaum and Berg 2000).

Financial participation is therefore typically seen as a substitute for otherforms of monitoring and incentives when the nature of the work environ-ment inhibits use of these other instruments. Recently, however, a literaturehas emerged within the agency-based tradition that implies that financialparticipation and individual incentives may be complementary. This approachemphasizes the complexity of operating incentive plans, given the costs ofusing objective performance measures and the dangers of incentivizingemployees to focus on partial aspects of their jobs (Baker, Gibbons, andMurphy 1994; Gibbons 1998; Prendergast 1999; Ittner, Larcker, and Meyer2003b). The emerging orthodoxy suggests that firms will mitigate the adverseeffects of particular monitoring and incentive instruments by using combi-nations of instruments (Holmstrom and Milgrom 1994). A very similar lineof argument is found in the “balanced scorecard” literature (Kaplan andNorton 1996). This suggests that the short-run, relatively narrow incentivesprovided by individual performance pay need to be balanced by incentives

758

/

A

P

with a longer-term and broader dimension, if the potentially dysfunctionaleffects of the former are to be avoided. Cooperative behavior, engenderedby group rewards, is necessary to counter-balance the self-interested focusof individual incentives. Thus, an alternative scenario to that presented bythe financial participation literature is that various forms of monitoring andincentives may be complementary to each other, and hence financial parti-cipation may be found where close monitoring and individual incentives arealso in use.

Empirical Evidence in the Financial Participation Literature

Empirical assessment of the determinants of broad-based stock plans inthe financial participation literature has focused on the role of monitoringcosts and the limitations of individual incentives schemes in certain workenvironments. This section reviews the extant empirical evidence in relationto these issues. It is found that the use of financial participation to amelio-rate monitoring costs is less than compelling while there is little directempirical engagement with the relationship between stock plans and otherpay instruments. The section also reiterates the common finding that stockplans are found in circumstances where their incentive effects would appearto be weak.

Several studies have considered whether all-employee share ownershipplans are more prevalent where work activities are difficult to monitor. Acommon approach has been to use workforce composition as a proxy formonitoring costs. For instance, it has been suggested that a high proportionof managers (after controlling for size) indicates a high degree of hierarchyand thus high monitoring costs (Drago and Heywood 1995). The evidence,however, is rather mixed. While Cheadle (1989) found a negative relation-ship between the proportion of managers and the presence of profit sharing,Drago and Heywood (1995) found a positive significant relationship betweencollective incentives and the proportion of managers in the workforce. Theyalso interpret this is as evidence that collective incentives are used to lowermonitoring costs. Given that the argument can be made either way, thesuitability of this proxy for monitoring costs in the financial participationliterature is open to question.

Empirical findings relating to the nature of job tasks and work technologyprovide mainly unsupportive evidence for the view that share ownershipplans are used to ameliorate monitoring costs in complex work settings. Arange of proxies for work measurement costs have been used such as capitalintensity, inter-dependent technology and work organization, unstable or


/ 759

changing technology, and high task complexity (Kruse 1996). Festing et al.(1999) found that changes to widen job content were not significantly asso-ciated with the presence of share ownership. Using workforce compositionas a proxy for types of work environment, Cheadle (1989) failed to findsignificant relationships with financial participation. Using the proportionof unskilled workers as a measure of the level of technology, McNabb andWhitfield (1998) found negative but insignificant relationships with perfor-mance in firms with employee share ownership. Drago and Heywood (1995)found that both labor intensity and technological change were insignifi-cantly related to profit sharing, with the sign for labor intensity contrary topredictions. However, positive relationships were found between profitsharing and quality circles. In Jones and Pliskin (1997), the proxy for machine-paced work—the capital–labor ratio—was found to be positively rather thannegatively associated with stock purchase plans. The main exception tothese unsupportive findings is Kruse (1996), who found significantly highercapital intensity in firms with share schemes at the end of his observationperiod and in firms that adopted share plans during the period. However,he also found negative relationships between share plans and both teamworkand job enrichment plans.

The work settings referred to above are also likely to be ones whereindividual incentives are costly to operate, for the reasons outlined earlier(Jones and Pliskin 1997; Brown 1990). However, many studies of financialparticipation do not include any measures for individual incentives or elsedo not use it as an independent variable even though a substitution effectis posited in the theoretical rationale for stock plan usage. Where data isavailable on individual incentives, there has been a tendency to reportseparate estimates for financial participation and individual incentives, as inDrago and Heywood’s (1995) study of incentive plans in Australia. In somecases, this analytical approach is clearly justified by the distribution ofschemes. Heywood, Hubler, and Jirjahn’s (1998) investigation of the deter-minants of variable pay schemes in Germany found that only a fraction of1 percent of their firms used both profit sharing and piece rates. This impliesa substitution effect, especially as some key variables take different signsbetween the two systems.

Although few studies engage empirically or directly with the use of finan-cial participation to substitute for individual incentives, there is a range ofother empirical evidence that casts doubt on the substitution argument. Thisrelates primarily to the weakness of stock ownership plans as an incentive.An important limitation of stock plans is the “the line of sight” problem(see Conyon and Freeman 2004). This refers to the capacity of stock plansto provide a direct link between individual effort and the receipt of additional

760

/

A

P

rewards. Research on motivation and compensation tells us that pay is morelikely to be an effective motivator when employees feel able to influence payoutcomes. (Lawler 1971). Yet the “line of sight” appears to be especiallyindirect and obscured in the case of share schemes for several reasons. One,the empirical evidence indicates that equity-based plans are concentratedamong large firms, where the contribution of individual employees tooutput will typically is small and/or indirect (Kruse 1996; Pendleton 1997;Gregg and Machin 1988; Pendleton et al. 2001). Two, the concentration ofschemes among listed firms (see Pendleton 1997; Pendleton et al. 2001)means that the value of the reward will also be a function of exogenousinfluences, such as investor sentiment and stock price volatility. Three, thereis usually a considerable time lag between award of the stock and full acqui-sition of the benefit, either because of vesting periods (stock options) ordeferral periods (stock purchase and stock award programs).

A further limitation on the incentive power of broad-based stock plans isthe free rider or 1/n problem. Lazear (2004) and Oyer (2004) have arguedrecently that the free-rider problem fundamentally weakens the incentiveeffects of collective remuneration, except in the very smallest firms (see alsoPrendergast 1999). Yet the near unanimous empirical finding from theresearch literature is that all-employee stock plans are predominantly foundin larger firms, where incentive effects are likely to be weakest (Gregg andMachin 1988; Kruse 1996; Pendleton 1997; Pendleton et al. 2001). Admittedly,the 1/n problem is widely recognized in the financial participation literature.The response to it typically takes one or two forms: one emphasizes themagnitude of monitoring costs in large organizations (Drago and Heywood1995; Jones and Pliskin 1997) so that the size effect becomes ambiguous;the other, pace Weitzman and Kruse (1990), refers to the possibility of“repeated games,” i.e., as participation events occur over and over again,employees may come to engage in cooperative behavior despite weak directincentives to do so in the first instance. The precise contexts and circum-stances in which these repeated games may take place is not usually speci-fied but the literature has emphasized the importance of other forms ofemployee participation as complements to financial participation in thisrespect (Ben-Ner and Jones 1995; see also Levine and Tyson 1990). Otherobservers, less committed to the moral hazard explanation for stock plans,have argued instead that that their primary functions may be to signalmanagerial confidence to less-informed investors (Lazear 2004), or to pro-vide a flexible instrument for employee selection and retention (Oyer andSchaefer 2005).

These findings suggest that stock plans are unlikely to be used as a sub-stitute for individual incentives because their incentive power is likely to be


/ 761

too weak (cf. Oyer 2004; Oyer and Schaeffer 2005). There is a case to bemade, therefore, for an alternative view, that financial participation maybe used to complement individual incentives in certain circumstances. Anew strand of literature in the agency tradition emphasizes that themultiple task components of employees’ jobs are likely to require multipleremuneration instruments (Holmstrom and Milgrom 1994; Gibbons 1998).The “balanced scorecard” literature indicates that those firms who arefocused on long-term performance and achievement of corporate “mission”are more likely to use a balance of incentive systems (Kaplan and Norton1996: 221). The function of employee stock plans in these circumstancesmay be to broaden the narrow focus, often produced by higher-poweredincentives, to lengthen the time frame, and to counter the low-trust dynamicthat is often associated with individual performance pay. However, thoughthese recent literatures have emphasized how packages of incentives mayserve to mitigate dysfunctional elements emanating from individual schemes,there has been very little empirical engagement with this possibility. Whetherstock plans complement or substitute for individual incentives is thereforea timely topic for empirical investigation.

Data and Methods

Given the observations in the previous sections, the empirical strategytakes a number of stages. The first concern is to consider the determinantsof broad-based stock plans in a way that replicates the approach of previousstudies. Various proxies included in earlier literature, such as workforcecomposition, task discretion, and teamwork, are included as independentvariables. We then address more explicitly whether stock plans are a sub-stitute for individual incentives. A final stage of the analysis then assessesthe determinants of the combined use of individual incentives and stockownership plans.

The data source is the WERS of 2091 Workplaces, conducted in 1998.This is a nationally representative survey of UK workplaces with ten ormore employees. The main survey

2

relies on a management respondent (themost senior manager with responsibility for employee relations/HRMissues) for each workplace but asks mainly factual questions to minimizethe potential for respondent bias. It is a very well-established data sourcefor industrial relations analysis in the UK, and over the years an extensiveprogram of research has been based on this series of surveys (conducted

2

There is also a worker representatives survey and a survey of employees.

762

/

A

P

periodically since 1980) (see Millward, Bryson, and Forth 2000). Earlierpapers on financial participation using this survey include Addison andBelfield (2000, 2001), Beaumont and Harris (1995), Fernie and Metcalfe(1995), Gregg and Machin (1988), McNabb and Whitfield (1998), andPendleton (1997). Following McNabb and Whitfield (1998) and Addisonand Belfield (2000), we limit the analysis to trading sector workplaces andexclude workplaces in the public sector and those providing goods andservices to other workplaces in the same organization.

3

This leaves anunweighted data set of 1260 workplaces, of whom 325 (26 percent) havean all-employee share scheme. After weighting (large workplaces are over-represented in the survey) and excluding cases with missing values there are1308 workplaces, of whom 173 (13 percent) have a share plan.

4

The dependent variable in the first stage records the presence of an all-employee share ownership plan open to at least 80 percent of the workforce(share ownership plan). Although the precise characteristics of the schemesare not observed, in nearly all cases the share ownership scheme will be anInland Revenue-approved plan. The share plan most likely to be capturedby the WERS variable is Sharesave (SAYE): a stock options scheme intro-duced in 1980 that provides a tax advantageous saving plan mechanism foremployees to accumulate the money necessary to exercise discountedoptions in 3, 5, or 7 years’ time. Although it uses an option structure, SAYEis more akin to a share purchase plan, especially as very few participantsfail to exercise their option and around 50 percent of employees retainshares for some time after acquisition (see Pendleton 2005). The shareownership plan dummy is created from two questions: one on the presenceof a scheme, and the other on the extent of eligibility among the workforce.The variable is coded as 1 if a plan is present and if more than 80 percentof the workforce is eligible to participate.

5

We used a set of variables to control for organizational size (Medium, Large,Very Large), the form of ownership (Listed), and the country of ownership(Domestic). WERS measures organizational size in categories: these aregrouped here into Medium (200–999 employees), Large (1000–4999 employees),and Very Large (5000 or more employees). Contrary to practice in some ofthe literature, size is not viewed as a proxy for monitoring costs, as it is notself-evident that monitoring costs per employee are higher in larger firms.

3

The reason for excluding nontrading workplaces is that WERS does not ask them about marketstructure and opportunities. This reduces the initial data set by 257 cases.

4

The weights attached to each workplace in the sample are the inverse of the probability that it wasselected from the national population. The weights were attached by the survey designers. See Forth(2001) for further details.

5

WERS has a set of categories for the eligibility rate, with each covering 20 percentage points.


/ 763

However, it may be used to measure the weakness of incentives because thefree-rider effect grows in magnitude with company size.

The following measures are used to test whether employee share owner-ship is used in contexts where there are likely to be high costs attached todirect monitoring or individual incentives.

Workforce Composition.

Three variables measure the proportions ofmanagers (Managers), technical staff (Technical) and routine operatives andassembly staff (Operatives) in the workforce. The proportion of managersin the workforce proxies the extent of the managerial hierarchy, bearing inmind the view that a higher proportion indicates higher monitoring costs.The measure of technical staff is included because the complexity and timeframes of technical work makes it particularly difficult to monitor. Thus, apositive relationship is expected between the use of share schemes and thesize of the technical workforce. By contrast, a negative relationship isexpected between the proportion of routine assembly workers and shareschemes on the grounds that the work tasks of this occupational group willbe less costly to monitor, and that individual incentives could be usedinstead to incentivize this group.

Indicators of Work Organization and Job Design.

WERS asks whethermembers of the largest occupation group

6

have discretion over the methods(Discretion) and pace of work (Control). WERS uses a four-point scale,with answers ranging from “a lot” to “none.” The agency-based approachesin the financial participation usually predict a positive relationship betweenmeasures of this sort and the presence of an employee share plan since thelatter may provide a substitute for direct monitoring. We also include ameasure for the extent of team work (Teamwork). WERS asks what pro-portion of the largest occupational group works in formally designated teams,with answers recorded in one of seven categories, ranging from 0 to 100 percent.This variable may capture costs of monitoring individual work behaviorand output, and the prediction is that the difficulties of monitoring and ofusing individual incentives in group-based work situations will encouragethe use of employee share schemes (c.f. Applebaum and Berg 2000).

Other Measures of Monitoring Costs.

The Financial Economics literaturepredicts that agency costs are potentially higher when there are growth

6

WERS uses the “largest occupational group” as the base for many of its questions about workpractices on the grounds that diversity in practices may be observed in many workplaces. Reference tothe largest group provides a consistency of approach and also helps to capture the dominant practices.

764

/

A

P

opportunities for the firm because the decision set for managers will be larger(see Smith and Watts 1992). Both Ittner, Lambert, and Larcker (2003a) andCore and Guay (2001) found that the investment opportunity set is a sig-nificant determinant of stock option grants for all classes of employees.Since the market-to-book ratio is not available in WERS, we use a measureof anticipated product market growth instead (Growth Opportunities). Thisis a four-item question where respondents were asked whether the marketfor the main product or service is growing, mature, declining, or turbulent.If the market is growing, our variable is coded as 1; otherwise it is zero.

Measures of Incentives.

A critical measure focuses on the extent to whichindividual incentives are in use. If share plans substitute for individualincentives in certain types of work environment, then a negative associationbetween share plans and individual incentives should be found. To testthis we include a measure of the presence of performance pay. Incentivesis derived from a seven-item, unbalanced category question that askswhat proportion of the nonmanagerial workforce has received individualperformance-based pay in the last twelve months.

Finally, a measure of union strength (Union Density) is included to testwhether unions have any influence on the presence of share plans, alongwith a set of controls for industrial sector.

Summary statistics for each variable are shown in Table 1 while the cor-relation matrix for all variables can be found in Appendix 1.

The statistical approach utilized is logistic regression. Three models aretested. The first includes those measures that the literature has suggestedproxy for monitoring costs, and that relate primarily to the costs of mea-suring worker effort. The second includes instead measures that relate to theprovision of incentives and the costs of measuring output. The final modelcombines both sets of variables. There are no major problems of covariancebetween variables, though trade union density and some industry dummiesexhibit quite high correlation coefficients.

Results

Results are shown in Table 2. Overall, there is mixed support for the firstproposition found in the literature and some important counter-evidence tothe second. The most important and novel finding is that all-employee shareplans and individual incentives are complementary rather than substitutesfor each other. We discuss this finding in more detail shortly, after reviewingthe overall pattern of results.


/ 765

It is readily apparent that structural features are the most importantpredictors of share plans. Most important of all is organizational size. Theodds of having a share scheme grows with each step in organizational size,and the single most important determinant of the presence of a share schemeis very large organizational size (i.e., over 5000 employees. Forty-one percentof workplaces that belong to these organizations have a share scheme, andthe odds of plan presence is forty-six times greater than in organizations withless than 200 employees. Since the free-rider issue suggests that group-based

TABLE 1

S V

Variable Type of variable Mean Standard deviation

Managers % 11.54 10.70Technical % 4.98 12.36Operatives % 11.05 22.72Listed Company 0,1 0.41 0.49Domestic 0,1 0.90 0.30Medium 0,1 0.10 0.30Large 0,1 0.11 0.31Very Large 0,1 0.24 0.43Discretion 1–4 2.90 0.91Control 1–4 2.82 0.92Incentives 0–6 0.72 1.77Teamwork 1–7 4.29 2.50Growth 0,1 0.47 0.50OpportunitiesUnion Density % 11.21 23.28Performance factor 0 1Management constructSectorsManufacturing 0,1 0.20 0.40Utilities 0,1 0.002 0.05Construction 0,1 0.05 0.21Wholesale 0,1 0.27 0.45Hotels and 0,1 0.10 0.30CateringTransport 0,1 0.05 0.21Financial Services 0,1 0.05 0.21Business Services 0,1 0.12 0.33Independent variablesShare Ownership 0,1 0.13 0.34PlanIndividual 0,1 0.16 0.37IncentivesIncentive 0,1 0.05 0.22Combination

766

/

A

P

incentives are strongest in smaller organizations, these results cast doubt onthe importance of “direct” incentives as a primary factor behind the use ofshare plans, and suggest some other rationale for their use. The separateimportance of stock market listing indicates that the potential for liquidityis a powerful influence on the incidence of share schemes, whil the importanceof domestic ownership is to be anticipated, given that the tax breaks areavailable only in the UK.

Workforce composition provides very mixed support indeed for thepredictions in the financial participation literature. In the literature, theproportion of managers in a workforce has been viewed as an indicator ofmonitoring costs, and hence a predictor of financial participation. However,this measure is negatively related (significant at 0.05) to share plan presencehere. This result is nevertheless somewhat inconclusive in a cross-sectionalstudy as the lower proportion of managers could reflect the use of share plansas a substitute for direct monitoring instruments. More compelling evidenceagainst the mainstream view is provided by the results for routine, assembly

TABLE 2

T I M C I U E S

O P (L M)

Variable B Odds B Odds B Odds

Listed 1.828*** 6.219 1.699*** 5.466 1.724*** 5.605Medium 1.963*** 7.118 1.645** 5.182 1.935*** 6.925Large 2.944*** 18.984 2.740*** 15.487 2.966*** 19.420Very Large 3.837*** 46.393 3.632*** 37.794 3.879*** 48.364Domestic 0.905* 2.473 0.858* 2.359 0.953* 2.593Managers −0.035* 0.966 −0.029* 0.972 −0.031* 0.969Technical 0.042*** 1.043 0.036*** 1.037 0.042*** 1.043Operative 0.027*** 1.027 0.019** 1.019 0.025*** 1.026Discretion 0.290* 1.336 – 0.225 1.253Control 0.309* 1.362 – 0.338* 1.403Growth Opportunities 0.484* 1.623 – 0.408 1.503Teamwork – – 0.032 0.944 0.027 1.028Incentives – – 0.158** 1.171 0.147** 1.159Union Density 0.000 1.000 0.000 1.000 0.000 1.000Significant Industry Dummies Finance and

transportFinance and transport

Finance and transport

DiagnosticsChi-square 453.181*** 453.030*** 461.719***−2 log likelihood 556.158 585.859 547.621Percentage correct predictions 90.4 90.5 91.1N 1267 1308 1268

N: * = significant at 0.05; ** = significant at 0.01; *** = significant at 0.001.


/ 767

workers. Financial participation would not be anticipated for this group ofworkers because direct monitoring costs ought to be relatively low. Yet, thismeasure is positively and significantly related (at 0.001) to the use of shareplans, though admittedly the effect is fairly small. Meanwhile, the results fortechnical workers are supportive of the predictions made in the literature:workplaces with larger numbers of an occupational group whose workeffort and output appear difficult to measure are significantly more likely tohave share plans.

The measures for the nature of work tasks provide more consistent sup-port for the agency theory-based view found in the literature. In column 1,share plans are significantly more likely when employees have greatercontrol over the pace of work and have greater discretion over how work isperformed (both significant at 0.05). Share plans therefore appear to func-tion as a substitute for direct monitoring, as predicted. Growth opportunitiesare also related to share plan usage in the direction predicted. However,when individual incentives are entered (column 3), job discretion and growthopportunities become insignificant. The extent of work teams, potentiallymaking output difficult to measure, is positively associated with share plans,as anticipated, though the result is not significant.

The most interesting results of all are those concerning individual incentives.Contrary to the previous financial participation literature, these results indi-cate that the two forms of payment are complementary, with the measurefor individual incentives significant at 0.005. The odds of having a shareplan are 16–17 percent greater for each step increase in the use of individualincentives. The combination of individual incentives with a high proportionof employees undertaking routine operative tasks is a surprising set ofpredictors for employee share plans given the stance taken in the financialparticipation literature, though considered by themselves the conjunction ofroutine operative work, some worker control over work pace, and the useof individual incentives would be expected (Brown 1990). Since stock planswould not be expected to serve a useful incentive function in these circum-stances, the possibility arises that stock plans serve some other purpose.These results also suggest that the determinants of stock plans need to beconsidered in conjunction with those for individual incentives.

Finally, the regression results suggest that unions have no influence, eitherway, on the use of share plans once other influences are controlled for. Thiscontrasts with earlier UK findings that unions and stock ownership plansare positively related (Gregg and Machin 1988; Pendleton 1997). In fact, inthis version of WERS stock plans and union membership are positivelycorrelated to a significant degree (see Appendix 1). In a reduced form ofthe regression without industry dummies, “union density” is positive and

768

/

A

P

significant. The explanation of this change in effects is that union density ishighly unequal between industries.

7

The results so far, therefore, are notable for contradicting many of thepredictions found in the financial participation literature. There is a verypronounced size effect that calls into question the incentive power of shareplans (given the free rider effect). Workforce composition does not un-ambiguously indicate high monitoring costs. Share ownership plans areassociated with the presence rather than absence of individual incentives.Furthermore, the combination of routine work and individual incentivesindicates work environments where stock plans would not be expected to beused as an alternative source of incentives.

Further Analysis: the Determinants of Multiple Reward Instruments

In the remainder of the paper, we interpret these results further in the lightof recent writing in agency-inspired work on incentives and compensation,and present some further analysis. We focus especially on the conjunctionof share ownership plans with individual incentives, as this is the clearestand most surprising departure from the predictions found in the financialparticipation literature. The co-presence of individual and collective incentivesis consistent with the recommendations and predictions in the “balancedscorecard” and new agency literatures. The former suggests that multipleincentives will counter the gaming effects associated with single, high-powered incentives (especially those based on short-term performance andwith a single metric) (Kaplan and Norton 1996). In a similar vein, the newagency literature argues that multiple incentives are used to inhibit agentsfocusing their efforts on a restricted and incomplete range of job tasks. Thecore insight underlying this argument is that most employees’ jobs involvemultiple tasks, and these may differ markedly in their character, focus, andmeasurability (Gibbons 1998).

Clearly, not all firms choose multiple incentives. What influences thenumber of pay incentives in use? The “balanced scorecard” literature doesnot generate a set of testable predictions as such but the implication is thatfirms that are more attuned to the desirability of multifaceted high perfor-mance will more likely use multiple incentives. In the new agency literaturethe choice between substitution or complementarity of incentive systems isseen to be driven by the measurability of task performance, and the desiredbalance of employee efforts across their range of tasks (Holmstrom and

7

This is the only measure where industry dummies affect the results.


/ 769

Milgrom 1994). This view therefore incorporates “traditional” agency viewson monitoring costs and moral hazard.

We test these predictions by further developing the earlier empiricalanalysis. The variables previously used to proxy monitoring costs are againdrawn upon. These refer to occupational composition of workplaces, theextent of employee discretion over work pace and methods, and the extentof teamwork. Between them, they address the measurability issues identifiedin the new agency literature. Commitment to all-round performance (balancedscorecard) and the desirability of multitasking (new agency) are proxiedby a new variable that measures the breadth of performance targets andmonitoring.

WERS has an extensive range of questions in this area. These includequestions on the number and type of performance targets in use (e.g.,product quality targets, productivity targets, etc.); whether individuals aremade aware of their job responsibilities by targets and objectives, whetherperformance appraisal is used to establish pay levels, the type of instrumentsfor monitoring product/service quality, etc. Preliminary analysis indicatesthat many of these items are highly inter-correlated. The most appropriateway to proceed, therefore, is to reduce these items to one or more constructs.Three items were factor analyzed: the number of performance targets in use(0–6), the number of ways of monitoring product/service quality (0–6), andthe proportion of the workforce undergoing regular performance appraisal(seven categories based on percentages). These loaded onto a single factor,and a derived score was created from this (using a regression procedure).This construct is referred to as Performance Management (see Table 1 fordescriptives).

We then create a new dependent variable to indicate the combined use ofemployee equity plans and individual incentives. Combined Incentivesrecords the presence of an all-employee equity plan and the use of individualincentives for 40 percent or more of the workforce (equal to 1 whenboth conditions satisfied, 0 otherwise).

8

We also modify the measure ofIncentives used in Table 2 to create a new bipolar dependent variable forIndividual Incentives (equal to 1 if 40 percent of the workforce hasincentive pay, 0 otherwise). This provides a reference point for incentives towhich equity plans are then added. Since the correlation between Incentivesand Share Ownership Plans is fairly high (

r

= 0.233, see Appendix 1),workplaces with share plans are excluded from this reference-point model.The earlier regressions are re-run with the new dependent variables and the

8

We experimented with variants of this variable but resulting models were unstable. Key variableswere characterized by high odds and high standard errors.

770

/

A

P

omission of Incentives from the independent variables. Results are pre-sented in Table 3.

Comparison of the results for the two dependent variables indicates thecircumstances in which all-employee stock plans are added to individualincentives, and these results are supportive of the balanced scorecard andnew agency perspectives. Usage of individual incentives on their own is inthe main not strongly predicted by the variables selected. In particular, thevariables used to proxy monitoring costs are not significantly associatedwith individual incentives. Employees operating under individual incentivesdo not clearly have greater discretion over the way work is done or the paceof work. Nor is teamwork a feature of these workplaces. The significantlylower proportion of managers in incentive plan workplaces, however, mayindicate that individual incentives are used in preference to direct supervi-sion. The proxy for performance management is highly significant. Theoverall picture, then, is that individual incentives are used to ensure thatemployees, whose jobs are not especially difficult to monitor, hit perfor-mance targets.

TABLE 3

D U M I (L M)

Variable

Individual incentives Incentive combination

B Odds B Odds

Listed 0.463 1.590 1.848** 6.346Medium 0.214 1.238 1.795 6.018Large 0.654* 1.923 0.194 1.214Very Large −0.406 0.666 2.080* 8.005Domestic 0.442 1.556 −0.062 0.940Managers −0.041** 0.960 −0.017 0.983Technical 0.012 1.012 0.020 1.021Operatives −0.001 0.999 0.012 1.012Discretion 0.240 1.271 0.803** 2.233Control −0.176 0.839 0.602* 1.826Growth Opportunities 0.012 1.012 0.862* 2.369Teamwork −0.023 0.978 0.188* 1.207Union density −0.008 0.992 −0.010 0.990Management Performance 0.656*** 1.927 1.091*** 2.976Industry Dummies Yes YesDiagnostics

Chi-square 147.754*** 288.172***−2 log likelihood 670.102 237.304Percentage correct predictions 86.6 96.5N 1093 1266

N: * = significant at 0.05; ** = significant at 0.01; *** = significant at 0.001.

New Evidence and Interpretations / 771

When all-employee stock plans are added to individual incentives, the changein results is both striking and in accordance with predictions. Employeecontrol over pace and methods of task execution become significantlypositive, indicating that measurability of employee effort and performancebecomes more difficult. The odds of using both stock plans and individualincentives are approximately double for each step change in job control anddiscretion. Furthermore, workplaces with share plans and individual incen-tives are more likely to have teamwork than workplaces without or withindividual incentives only, implying that measuring employee contributionsto output will be more difficult. Overall, the costs of monitoring employeeperformance are clearly higher in these workplaces. An additional factor isthat these workplaces benefit from greater growth opportunities. The litera-ture suggests that in these circumstances, instruments will be used to ensurethat employees with high levels of discretion make the right decisions (Coreand Guay 2001; Ittner, Lacker, and Meyer 2003a). Finally, performancemanagement is strongly positive, with a higher level of odds of predictingcombined incentives than individual incentives alone. In line with predic-tions, workplaces with individual incentives and employee stock plans havea greater range of performance targets and measures than workplaces withjust individual incentives or no incentives at all.9 The odds of having stockplans and incentive plans are approximately half as much again as those forhaving individual incentives for each unit of performance management.

The results are therefore supportive of both strands of literature onmultiple incentives, and illustrate the circumstances in which all-employeestock plans and individual incentives accompany each other.

Discussion and Conclusions

The main objective of the paper has been to assess the theoreticalapproaches found in the financial participation literature toward the deter-minants of employee share ownership plans, utilizing evidence from the UKWERS. Contrary to the predictions and implications in this literature thatshare plans will be used where monitoring costs are high and individualincentives difficult to use, share plan workplaces are more likely to useindividual performance pay and to have some workforce characteristicsthat should be relative cheap to monitor. When the determinants of the

9 The mean number of performance targets is 4.8 for workplaces with combined incentives, 4 forworkplaces with individual incentives only, and around 3 for workplaces without any incentives.

772 / A P

combined use of individual incentives and stock ownership plans areconsidered, the proxies for monitoring costs become more consistent withprior expectations.

The financial participation literature has struggled with two difficult setsof findings in the past. Most important, stock ownership plans have nearlyalways been found to be far more prevalent in work settings where theirincentive effects are likely to be weak or nonexistent. Scholars in this traditionhave mainly sought refuge in the “repeated games” scenario, where overtime the use of participatory instruments, typically in conjunction withother forms of participation, is predicted to counter the weakness of initialdirect incentive effects (Weitzman and Kruse 1990). Yet, this argument alsofalls prey to the free-rider argument (Prendergast 1999) and it is difficult todemonstrate such repeated games empirically. For this reason, some recentcontributions in personnel economics have sought alternative explanationsfor the use of stock-based plans (Lazear 2004; Oyer 2004; Oyer and Schaefer2005). The second problem is that in many of the studies reviewed earlier,the proxies for monitoring costs do not work very well. Ours do not workwell either when the typical approach in the literature is replicated.

It will be apparent that combining individual incentives and stock planscuts through these problems. Our measures for monitoring costs work muchbetter in the final model when the determinants of both schemes areconsidered. Most importantly, the free-rider problem disappears. The expla-nation for stock plan presence no longer has to be rooted in the provisionof strong incentives. These are provided instead by individual pay incen-tives. The function of share ownership plans might be rather different. Theymight be seen as “weak incentives” aimed at achieving more diffuse or“softer” effects than “hard” incentives (Lazear 1989; Gibbons 1998).

The results in the paper are supportive of the arguments proposed in the“balanced scorecard” (Kaplan and Norton 1996) and “new agency” litera-tures. The latter in particular emphasizes the complexity and context ofcompensation contracts (Holmstrom and Milgrom 1994; Gibbons 1998;Prendergast 1999). The “balanced scorecard” perspective argues that a varietyof targets and instruments to achieve them will be used in organizationsfocusing on long-term high performance, while the agency literature empha-sizes that multitask work environments require a multifaceted approach tocompensation.

However, these perspectives discuss combinations of incentives in generalterms, and as yet, the literature has not considered the possible reasonsfor a specific complementarity between individual performance pay andbroad-based employee stock ownership plans. We speculate that stock planshave some specific features that may counter the well-known dysfunctional


affects of individual incentives and high-powered monitoring (see Baker,Jenson, and Murphy 1988). The latter can lead to a dysfunctional emphasison the targets chosen relative to other desirable outcomes, an emphasis onachievement of the immediate target rather than the underlying objective ofthe target, and low-trust relationships.

Share plans may ameliorate these problems in several ways. One, theymay help to win support for targets by signaling that firms will share thebenefits of good performance with employees. This signaling may counterthe low-trust effects of target setting and measurement (Gibbons 1998).Two, share plans may encourage “truthful” exchanges of information in theperformance appraisal process and other work monitoring procedures whenotherwise the linking of pay to performance provides incentives for adverseselection-type effects. It has been noted elsewhere that subjective forms ofwork measurement are used to mitigate the distortions induced by objectivesystems but are themselves prone to distortions (Baker, Gibbons, andMurphy 1994; Ittner, Lambert, and Larcker 2003b). Share plans may reducedistortions in subjective evaluations of performance by encouraging andsignaling cooperation. Three, share plans may counteract the tendency forperformance measures to focus workers’ attention on meeting the targetsspecified to the detriment of other desirable outcomes, and on meeting thespecific target rather than the objective underlying it. The linkage providedby the share plan to longer-term organizational performance, as reflected inshare prices, signals a wider and a longer-term performance metric thanbehavioral or workplace performance measures. Share plans may be seenas way of encouraging employees to meet performance targets but notat all costs.

Thus, the contribution of share plans to efficiency may emanate fromprecisely those features that appear to be fatal weaknesses when they areconsidered in isolation as alternative instruments to provide incentives.They may only provide weak direct incentives in themselves but they mayencourage employees to think and act more long-term, more broadly, andmore cooperatively than if they were paid solely according to their short-term individual performance.

R

Addison, John and Clive Belfield. 2000. “The Impact of Financial Participation and Employee Involve-ment on Financial Performance: A Re-Estimation Using the 1998 WERS.” Scottish Journal ofPolitical Economy 47:571–83.

——— and ———. 2001. “Updating the Determinants of Firm Performance: Estimation Using the1998 UK Workplace Employee Relations Survey.” British Journal of Industrial Relations 39:341–66.

Alchian, Armen and Harold Demsetz. 1972. “Production, Information Costs, and Economic Organiza-tion.” American Economic Review 62:777–95.

774 / A P

Appelbaum, Eileen and Peter Berg. 2000. “High-Performance Work Systems: Giving Workers a Stake.”In The New Relationship: Human Capital in the American Corporation, edited by Margaret Blairand Thomas Kochan, pp. 102–44. Washington DC: Brookings Institution.

Baker, George, Michael Jenson, and Kevin Murphy. 1988. “Compensation and Incentives: Practice vs.Theory.” Journal of Finance 43:593–616.

———, Robert Gibbons, and Kevin Murphy. 1994. “Subjective Performance Measures in OptimalIncentive Contracts.” Quarterly Journal of Economics 109:1125–57.

Beatty, Anne. 1994. “An Empirical Analysis of the Corporate Control, Tax and Incentive Motivationsfor Adopting Leveraged Employee Stock Ownership Plans.” Managerial and Decision Economics15:299–315.

Beaumont, Philip and Rosemary Harris. 1995 “The Pattern of Diffusion of Employee Share Schemes inBritain: Some Key Findings.” International Journal of Human Resource Management 6:391–410.

Ben-Ner, Avner and Derek Jones. 1995. “Employee Participation, Ownership, and Productivity: ATheoretical Framework.” Industrial Relations 34:532–53.

Blasi, Joseph, Michael Conte, and Douglas Kruse. 1996. “Employee Stock Ownership and CorporatePerformance among Public Companies.” Industrial and Labor Relations Review 50:60–80.

Brown, Charles. 1990. “Firms Choice of Methods of Pay.” Industrial and Labor Relations Review43:165–82.

Cheadle, Allen. 1989. “Explaining Patterns of Profit Sharing Activity.” Industrial Relations 28:387–400.———. 1990. “Why do Firms Adopt Profit Sharing? Evidence from Deferred Profit-Sharing Plans.” In

Profit Sharing and Gain Sharing, edited by M. Roomkin, pp. 46–64. Metuchen, NJ: IMLR Press/Rutgers University and the Scarecrow Press.

Conyon, Martin and Richard, Freeman. 2004. “Shared Modes of Compensation and Firm Performance:UK Evidence.” In Seeking a Premier League Economy: The Economic Effects of British EconomicReforms, 1980–2000, edited by D. Card, R. Blundell, and R. Freeman, pp. 109–24. Chicago:University of Chicago Press.

Core, John and Wayne Guay. 2001. “Stock Option Plans for Non-Executive Employees.” Journal ofFinancial Economics 61:253–87.

Drago, Robert and John Heywood. 1995. “The Choice of Payment Schemes: Australian EstablishmentData.” Industrial Relations 34:507–31.

Eisenhardt, Kathleen. 1989. “Agency Theory: An Assessment and Review.” Academy of ManagementReview 14:57–74.

Fernie, Sue and David Metcalf. 1995. “Participation, Contingent Pay, Representation, and WorkplacePerformance: Evidence from Great Britain.” British Journal of Industrial Relations 33:379–416.

Festing, Marion, Yvonne Groening, Rudiger Kabst, and Wolfgang Weber. 1999. “Financial Participationin Europe-Determinants and Outcomes.” Economic and Industrial Democracy 20:295–329.

Forth, John. 2001. Three Common Questions about the WERS98 Sample Design and its Implicationsfor Analysts. London: National Institute of Economic and Social Research.

Gibbons, Robert. 1998. “Incentives in Organizations.” Journal of Economic Perspectives 12:115–32.Gregg, Paul and Stephen Machin. 1988. “Unions and the Incidence of Performance Linked Pay

Schemes in Britain.” International Journal of Industrial Organization 6:91–109.Heery, Ed. 1998. “A Return to Contract? Performance-Related Pay in a Public Service.” Work, Employ-

ment and Society 12:73–95.Heywood, John, Stan Siebert, and Xiangdong Wei. 1997. “Payment by Results Systems: British Evi-

dence.” British Journal of Industrial Relations 35:1–22.———, Olaf Hubler, and Uwe Jirjahn. 1998. “Use of Variable Payment Schemes: Evidence from

Germany.” Kyklos 51:237–57.Holmstrom, Bengt. 1979. “Moral Hazard and Observability.” Bell Journal of Economics 14:74–91.——— and Paul Milgrom. 1994. “The Firm as an Incentive System.” American Economic Review

84:972–91.Ittner, Christopher, Richard Lambert, and David Larcker. 2003a. “The Structure and Performance

Consequences of Equity Grants to Employees of New Economy Firms.” Journal of Accountingand Economics 34:89–127.


———, David Larcker, and Marshall Meyer. 2003b. “Subjectivity and the Weighting of PerformanceMeasures: Evidence from a Balanced Scorecard.” American Accounting Review 78:725–58.

Jones, Derek and Jeffrey Pliskin. 1997. “Determinants of the Incidence of Group Incentives: Evidencefrom Canada.” Canadian Journal of Economics 30:1027–45.

———, Takao Kato, and Jeffrey Pliskin. 1997. “Profit Sharing and Gain Sharing: A Review of Theory,Incidence, and Effects.” In The Human Resource Management Handbook Part 1, edited by DavidLewin, Daniel Mitchell, and Mahmood Zaidi, pp. 153–73. Greenwich, CT: JAI Press.

Kaplan, Robert and David Norton. 1996. The Balanced Scorecard: Translating Strategy into Action.Boston, MA: Harvard Business School Press.

Kruse, Douglas. 1996. “Why do Firms Adopt Profit Sharing and Employee Ownership Plans?” BritishJournal of Industrial Relations 34:515–38.

Lawler, Edward. 1971. Pay and Organizational Effectiveness: A Psychological Approach. New York:McGraw-Hill.

Lazear, Edward. 1989. “Pay Equality and Industrial Politics.” Journal of Political Economy 97:561–80.———. 2004. “Output-Based Pay: Incentives, Retention or Sorting?” Research in Labor Economics

23:1–25.Levine, David and Laura d’Andrea Tyson 1990 “Participation, Productivity, and the Firm’s Environ-

ment.” In Paying for Productivity: A Look at the Evidence, edited by Alan Blinder, pp. 183–244.Washington DC: Brookings Institution.

Lupton, Tom. 1963. On the Shop Floor: Two Studies of Workshop Organization and Output. Oxford:Pergamon Press.

Marsden, David and Ray Richardson. 1994. “Performing for Pay? The Effects of Merit Pay on Moti-vation in a Public Service.” British Journal of Industrial Relations 32:243–62.

McNabb, Robert and Keith Whitfield. 1998. “The Impact of Financial Participation and EmployeeInvolvement on Financial Performance.” Scottish Journal of Political Economy 45:171–89.

Millward, Neil, Alex Bryson, and John Forth. 2000. All Change at Work? British Employment Relations1980–1998, as Portrayed by the Workplace Industrial Relations Survey Series.

Oyer, Paul. 2004. “Why do Firms Use Incentives that have No Incentive Effects?” Journal of Finance59:1619–49.

——— and Scott Schaefer. 2005. “Why do Some Firms Give Stock Options to All Employees? AnEmpirical Examination of Alternative Theories.” Journal of Financial Economics 76:99–133.

Pendleton, Andrew. 1997. “Characteristics of Workplaces with Financial Participation: Evidence fromthe Workplace Industrial Relations Survey.” Industrial Relations Journal 28:103–19.

———. 2005. “Sellers or Keepers? Share Retentions in Share Option Plans.” Human Resource Manage-ment 44:319–36.

———, Erik Poutsma, Jos van Ommeren, and Chris Brewster. 2001. Employee Share Ownership andProfit-Sharing in the European Union. Dublin: European Foundation for the Improvement ofLiving and Working Conditions.

Prendergast, Canice. 1999. “The Provision of Incentives in Firms.” Journal of Economic Literature 37:7–63.Sesil, James, Maya Kroumova, Joseph Blasi, and Douglas Kruse. 2002. “Broad-Based Employee Stock

Options in U.S. ‘New Economy’ Firms.” British Journal of Industrial Relations 40:273–94.Smith, Clifford and Ross Watts. 1992. “The Investment Opportunity Set and Corporate Financing,

Dividend, and Compensation Policies.” Journal of Financial Economics 32:263–92.Weitzman, Martin and Douglas Kruse. 1990. “Profit Sharing and Productivity.” In Paying for Produc-

tivity: A Look at the Evidence, edited by Alan Blinder, pp. 95–142. Washington DC: BrookingsInstitution.

776/

A

P

APPENDIX 1

C M V

1 2 3 4 5 6 7 8 9 10 12 13

1 Managers 1.0002 Technical 0.004 1.0003 Operatives −0.068 −0.079 1.0004 Listed 0.027 −0.026 −0.014 1.0005 Domestic −0.185 −0.065 −0.154 −0.066 1.0006 Medium −0.046 0.079 0.032 0.074 −0.102 1.0007 Large 0.004 0.015 0.021 0.090 −0.103 −0.118 1.0008 Very large −0.063 −0.128 −0.133 0.503 0.024 −0.191 −0.198 1.0009 Discretion 0.142 −0.007 −0.089 −0.130 0.017 −0.031 −0.061 −0.140 1 –10 Control 0.092 0.029 0.049 −0.095 0.005 −0.083 −0.082 −0.073 0.377 112 Incentives −0.049 −0.001 −0.078 0.167 0.026 0.021 0.070 0.178 −0.003 −0.013 113 Teamwork 0.025 0.130 −0.167 0.012 −0.020 0.046 0.025 0.017 −0.022 0.046 0.108 114 Growth opportunities 0.024 0.058 −0.036 0.119 0.008 0.016 0.057 0.122 −0.060 −0.086 0.051 0.04315 Union density −0.116 −0.014 0.096 0.112 0.014 −0.031 0.102 0.251 −0.076 −0.001 0.039 0.09316 Performance −0.002 0.031 −0.002 0.264 −0.112 0.090 0.142 0.336 −0.057 0.020 0.094 0.26217 Manufacturing −0.069 0.090 0.448 0.008 −0.089 0.064 0.033 −0.210 0.009 0.058 −0.119 −0.15718 Utilities −0.033 0.042 0.004 0.055 −0.021 −0.003 0.056 0.021 −0.008 −0.013 0.006 0.03919 Construction −0.085 0.044 −0.037 0.070 0.056 0.032 −0.033 −0.055 −0.060 0.031 −0.040 0.09620 Wholesale 0.074 −0.183 −0.159 0.010 0.019 −0.060 0.030 0.225 −0.069 −0.073 0.257 0.01321 Hotel/catering 0.014 −0.121 −0.068 0.188 −0.005 −0.040 0.000 0.238 −0.034 −0.039 −0.102 −0.06622 Transport −0.074 −0.008 0.179 0.014 −0.073 0.026 −0.031 0.016 −0.020 −0.043 −0.047 −0.02923 Financial services 0.063 0.034 −0.107 0.109 0.004 −0.051 0.044 0.206 −0.092 −0.084 0.217 0.14924 Business services 0.104 0.194 −0.111 −0.096 −0.038 0.090 −0.003 0.119 0.053 0.008 −0.042 0.00725 Share ownership plan −0.092 0.049 0.037 0.392 0.041 −0.042 0.078 0.447 −0.045 −0.010 0.231 0.08126 Individual incentives −0.049 −0.001 −0.078 0.167 0.026 0.021 0.070 0.178 0.005 −0.030 – 0.10627 Incentive combination −0.036 −0.012 −0.081 0.245 0.016 −0.026 −0.039 0.333 0.037 0.027 0.491 0.140

New

Evidence and Interpretations

/777

14 15 16 17 18 19 20 21 22 23 24 25 26 27

14 Growth opportunities 115 Union density −0.079 116 Performance 0.097 0.174 117 Manufacturing −0.023 0.112 −0.023 118 Utilities −0.010 0.144 0.033 −0.023 119 Construction −0.098 −0.037 −0.154 −0.111 −0.010 120 Wholesale −0.019 −0.106 0.105 −0.306 −0.028 −0.136 121 Hotel/catering 0.076 −0.135 0.019 −0.166 −0.015 −0.074 −0.205 122 Transport 0.018 0.169 −0.022 −0.111 −0.010 −0.049 −0.136 −0.074 123 Financial services −0.010 0.322 0.163 −0.109 −0.010 −0.049 −0.135 −0.073 −0.049 124 Business services 0.094 −0.131 0.005 −0.187 −0.017 −0.083 −0.231 −0.125 −0.083 −0.082 125 Share ownership plan 0.095 0.239 0.319 −0.039 0.075 −0.040 0.059 −0.022 0.077 0.239 −0.030 126 Individual incentives 0.051 0.039 0.250 −0.119 0.006 −0.040 0.257 −0.102 −0.047 0.217 −0.042 0.233 127 Incentive combination 0.084 0.151 0.242 −0.096 0.031 −0.051 0.119 −0.054 −0.028 0.337 −0.049 0.587 0.528 1

incentives, monitoring, and employee stock ownership plans: new evidence and interpretations

Documents