Edward P. Lazear Stanford University
1
Performance Pay: Lessons from Personnel
Economics
Compensation, Work Practices and Productivity
SteyrJuly, 2003
Edward P. Lazear Stanford University
2
Today’s talk All pay is performance pay Traditional incentive pay
Uses - Sorting and incentives Results
Relative pay Intrinsic motivation
Edward P. Lazear Stanford University
3
All Pay is Performance Pay Issue of structure and formula
Piece rates, commissions Hourly wages, annual salaries
Which variables determine best structure? Measurability Heterogeneity Complexity
Edward P. Lazear Stanford University
4
Pay FormsPay on input or output?
Input• Hourly wage• Annual salary• Effort related bonuses
Output• Piece rates• Sales commissions• Output related bonuses• Team compensation
• Profit sharing• Stock and options
Edward P. Lazear Stanford University
5
Input or Output? Measurement costs and risk aversion
Output sometimes difficult to measure• Complex• Teams• Takes time to play out (R&D)• Education (earnings)
• Proxies – Test scores Output highly variable (risky projects)
When risk or measurement is an issue, pay on input Input imperfect but often more available measure of
what is desired Input is non-distorting
Edward P. Lazear Stanford University
6
Discrete or Continuous?
Compensation
Outputor effort
e*
W
0
Discrete
Continuous
Edward P. Lazear Stanford University
7
Both Create Incentives Discrete is strong in narrow range Continuous is weak in any one range,
but strong overall Not “high-powered” v. “low powered” Discrete well-suited to homogeneous
workforce Both induce sorting
Edward P. Lazear Stanford University
8
Choice of Pay Structure
Continuous (Heterogeneous Workforce)
Discrete (Homogeneous or idiosyncratic workforce)
Input (Output difficult to observe)
Hourly wage paid management consultant
Retainer paid to a lawyer
Output (Output easy to observe)
Piece rate workers
Non-commission salesperson
Edward P. Lazear Stanford University
9
A European Issue Termination difficult Suggests use of continuous over
discrete
Edward P. Lazear Stanford University
10
An Example: Team Incentives European Oil Company
Oil field on Arctic Ocean – great success
Head office in Anchorage – total failure
The lesson Production versus managers? No Peer pressure effective small groups
Edward P. Lazear Stanford University
11
Results on “Performance Pay” for Managers Abowd (1990)
16,000 managers at 250 corporations Additional 10% performance 400 to 1200 basis
points on stock return Leonard (1990)
Companies with long-term incentive plans had significantly higher R.O.E.
Kaplan (1994) Japanese executives whose compensation is tied to
performance are associated with companies with higher subsequent performance
Almost never find the reverse
Edward P. Lazear Stanford University
12
Managerial Performance Pay The “Puzzle”
If effective, why is the relation of pay to performance closer to zero than one?
• Jensen & Murphy, • Hall and Leibman• Graziano and Parigi (Italian firms have similar low elasticities)
Explanations Straw man
• Every manager cannot own firm• Groves and other mechanisms have problems
Enough incentives if managers are risk averse (Haubrich) Another reason for performance pay: Information for investors
Edward P. Lazear Stanford University
13
Put Your Money Where Your Mouth Is or “Skin in the Game”
Venture capitalists require low base + performance pay
Wage = a + b (Output) Incentive solution
b=1, a<0 Sorting solution: Homogeneous managers
b very small, a slightly less than alternative Always accepts positive profit project Never accepts never profit project
Much evidence for this view (new, high tech firms use more)
Edward P. Lazear Stanford University
14
Another Form of Performance Pay: Tournaments
Almost all managers face implicit if not explicit tournaments
Tournament defined as pay or non-monetary compensation function of relative performance
Edward P. Lazear Stanford University
15
Tournaments Goal of tournament theory
To explain salary hierarchy To explain different patterns across industries
and countries Metaphor of tennis match
Compensation based on relative position The larger the spread, the more effort There is an optimal spread
• Too much variation creates recruitment and cooperation problems
Edward P. Lazear Stanford University
16
Mathematics
(W1 W2)g(0) C(µj)
a. qj µj εjb. qk µk εk
Maxµj
W1P W2(1 P) C(µj)
P Prob (µj εj>µk εk) Prob (µj µk>εk εj) G (µj µk)
(W1 W2)Pµj
C (µj) 0
Edward P. Lazear Stanford University
17
So… Bigger raises create greater
incentives Use larger spread in “noisier”
environments Higher wage differentials in new and/or
risky industries Higher wage differentials in US than in
Europe But bigger raises less teamwork
Edward P. Lazear Stanford University
18
Evidence on Tournaments Eriksson (1999) finds strong support of
implied wage structure O’Reilly, Crystal and Main: Promotion
rates Sports: Ehrenberg and Bognano and later
Bronars, golf Experiments: Bull, Schotter and Weigelt
(1987) and Fehr and Falk (2002), Freeman, et. al. (2003), European Science Days students
Edward P. Lazear Stanford University
19
More Evidence on Tournaments Knoeber: Chickens Drago and Garvey (1998)
Large raises for promotionbetter attendance in 23 Australian firms
Large raises associated with less sharing
Edward P. Lazear Stanford University
20
Intrinsic and Extrinsic Action on the margin At some levels, makes no sense
Cut commissions so will work harder?
Edward P. Lazear Stanford University
21
A Traditional Human Resources View
The fundamental economic theory of motivation is based on assumptions of effort aversion (people will not expend effort unless paid to do so), opportunism (people, in the pursuit of their own interests, will often misrepresent their true preferences and engage in guile and deceit), and a lack of goal alignment (employees in organizations have different agendas than the owners and, therefore, incentive systems need to be designed to force people to do what is right for the good of the organization). In the economists’ view, people are assumed to be lazy, dishonest, and at odds with the goals of the managers.
Although each of these assumptions may be valid in a specific situation, or for a particular individual (for instance, when managing economists themselves), none is likely to be right in most settings with normal human beings.
- Charles O’Reilly and Jeffrey Pfeffer
Edward P. Lazear Stanford University
22
Courses Taught
Salary
A
1 2 3 4 5
Equilibrium Number of Courses
An Example: Lazear’s Teaching
Edward P. Lazear Stanford University
23
Demand and Supply Version
Number of courses
WageLazear’s supply
Demand for Lazear courses
Intrinsic range
Extrinsic Range
Edward P. Lazear Stanford University
24
Conclusion Performance pay: Not whether but what form
Input or output Discrete or continuous Relative or absolute
Output-based performance pay may be as important for selection as incentives
Rank-order based incentives are pervasive for managers Manifestations
• Promotions• Bonuses• Status and other non-monetary factors
Adverse effects on cooperation Implicit if not explicit almost everywhere so understanding
essential