CH. 11. PAY AND PRODUCTIVITY: WAGE DETERMINATION WITHIN THE FIRM

CH. 11. PAY AND PRODUCTIVITY: WAGE DETERMINATION WITHIN THE FIRM

COMPENSATION ISSUES TO BE CONSIDERED.•The basis of compensation •The level of compensation relative to that for comparable workers elsewhere.•Sequence of pay.

Productivity and the Basis of Annual Pay

• Types of Incentive Pay.– Piece rate:

• Pay tied to amount of output produced.• Examples: sales commissions; garment workers.

– Gainsharing: • Pay tied to gains in measures of group “success”

(productivity, costs, quality, etc.)– Profit-sharing or Bonus Plans: Pay tied to

profits of firm.

Piece Rates

• Employee preferences.– Holding the “expected” level of pay constant,

employees prefer time-based over piece rate.– Piece rate is “riskier”

• Production process is uncertain (e.g. Sales commissions.)

– Piece rate will cause self-selection on productivity and risk aversion

Piece Rates

• Employer preferences and piece rates.– Advantages:

• Productivity is increased among a given work force.• Attract most productive workers.

– Disadvantages:• quantity vs. quality• willingness to help coworkers• treatment of equipment• cost of monitoring output

– sales: who made the sale?– Is there a single “output” that can be measured?

• how to set a “competitive” rate.

Empirical evidence on piece rates

• Lazear (2000):• When glass installer switched from hourly

to piece rate, • individual output of employees that stayed with the

firm rose 20 percent.• because the “slowest” employees left, average

increase in productivity was 36 percent.• average pay of workers rose 9% (both employees

and firm benefitted).

Lazear, EP. “Performance Pay and Productivity.” American Economic Review, v. 90 issue 5, 2000, p. 1346.

Empirical evidence on piece rates• Jirjan & Stephan (2004)

– Women are more likely to be paid piece rates than men

– Hypotheses:• shorter expected tenure than men. • greater demand for flexibility in work hours.• Women prefer because subject to less wage discrimination

when objective performance measures are available.

– Authors conclude explanation is primarily the last hypothesis.

Jirjahn and Stephan. “Gender, piece rates and wages: evidence from matched employer–employee data.” Cambridge Journal of Economics, v. 28 issue 5, 2004, p. 683.

Empirical evidence on piece rates

• Goldin (1986).– Female manufacturing workers around 1900 were far more likely

to be paid by the piece– Occupational segregation by sex and differences in earnings

result even if workers are equally productive. – Feminization of clerical sector explained by improved ability to

monitor output with new office technology.

Goldin, C. “Monitoring Costs and Occupational Segregation by Sex: A Historical Analysis.” JOURNAL OF LABOR ECONOMICS, v. 4 issue 1, 1986, p. 1.

Empirical evidence on piece rates• Paarsch & Shearer (2000).

– Study of tree-planting firm– Piece rate incentives caused 22 percent increase in

productivity.– Improvements in productivity partly offset by

decreased quality.

Piece Rates, Fixed Wages, and Incentive Effects: Statistical Evidence from Payroll Records. Harry J. Paarsch; Bruce Shearer International Economic Review, Vol. 41, No. 1. (Feb., 2000), pp. 59-92.

Group Incentive Plans• Employee preferences:

– “free-rider” problem.– risk sharing.

• Employer preferences: – monitoring individual versus group output– encourage team work

• Group incentives in a large firm– free rider problem is amplified.– can be used to persuade union to implement

productivity enhancing rules.• Evidence of effect of profit sharing on productivity

is very mixed.

Merit Pay

• If costs of monitoring and risk aversion make piece rates or group incentive plans undesirable, may choose to pay on basis of time worked and reward based on “merit”.

• Still have problem with monitoring “output” and the extent to which it is affected by “external forces”.

• Can remove effect of common external forces by using – relative performance rankings– executive pay and relative performance ratings.– tournaments

• Problems with relative performance rankings:– the incentive to “sabotage”.– perceived “fairness” of ratings.

Efficiency Wages

• Increases in the level of pay may increase productivity by:– increasing the quality of worker that is hired.– increasing the productivity of a given worker.

• reduces the probability of “shirking”• can reduce monitoring costs.• reduces turnover and saves on hiring/training costs.

• “efficiency wage premium”– a premium above and beyond what a worker can earn

elsewhere that is justified by an increase in worker productivity.

Efficiency Wages• The trade-off between wages and monitoring

• q=per period probability that a worker is detected shirking

• p=efficiency wage premium per period• L=expected length of job• Expected loss from shirking in a given

period=q*p*L

• A worker will shirk if expected gain from shirking > q*p*L

Efficiency Wages• To maintain a given level of shirking, a firm can trade-off

between more monitoring (increases q) and a lower efficiency wage (lower p). – This yields an “iso-shirk curve”. – Iso-shirk curves further to the NE represent lower levels of

shirking since they are associated with a higher value of q*p*L).

Iso-shirk (q*p*L=constant)

Q (prob of detection)

P (eff w premium)

Efficiency Wages

• If a firm spends $1 less on efficiency wages, they can spend $1 more on monitoring and keep the same cost.

• As more is spent on monitoring, the probability of detection rises. This yields an “iso-cost curve”.

Efficiency Wages• Iso-cost curves further to the SW are lower levels of

costs • Iso-cost curve will be steeper at firms where an extra $1

spent on monitoring has a larger effect on the probability of detection.

Iso-cost (q*p*L=constant)

Q (prob of detection)

P (efficiency w premium)

Efficiency Wagesq (prob. of detection)

p (eff. w premium)

Iso-shirk (q*p*L=constant)

Isocost1Isocost2

q*

p*

For a given level of shirking,

•the firm with isocost1 is minimizing its costs with q* and p*.

•the firm with isocost2 can reduce costs by moving to the NW along the iso-shirk curve.

Efficiency Wages

Factors influencing combination of efficiency wages and monitoring

• advancements in technology for monitoring

• changes in expected length of job.

• increase in cost of shirking

•Capital intensity

•Reputation

• firm size

Empirical studies

• Raff and Summers (1989) – Did Henry Ford Pay Efficiency Wages?– Introduction of the five-dollar day in 1914 – strongly supportive of the relevance of

efficiency wage theory.• evidence that the five-dollar day resulted in

substantial queues for Ford jobs. • significant increases in productivity and profits at

Ford accompanied the introduction of the five-dollar day.

Efficiency wages

• Cappelli (1991). – Comparison of workers across plants of a

single firm.– Plants with higher wages had fewer

disciplinary dismissal– Disciplinary dismissals also drop when cost of

dismissal rises (poor job alternatives).

Cappelli (1991). “An Interplant Test of the Efficiency Wage Hypothesis.” The Quarterly Journal of Economics, v. 106 issue 3, 1991, p. 769

Deferred Pay

Deferred pay may increase lifetime productivity by:

• reducing quits (selection effects enhance this)

• reducing monitoring costs

Necessary conditions for deferred pay:

• PV of lifetime compensation must be at least equal to that in alternative jobs.

• “underpayment” followed by “overpayment” relative to MRP.

Deferred Pay

Risks to employee

• firm has an incentive to fire once wage exceeds MRP.

• reputation effects may prevent firm from violating implicit agreement.

•age discrimination laws may also prevent firm from violating agreement.

Deferred Pay

Risks to employer

• worker may want to postpone retirement beyond “implicit” agreement.

• firm may respond with “mandatory retirement” (no longer legal).

• firm may respond with pension that encourages retirement (DB plan, not DC).

Deferred Pay vs. Efficiency Wages

• Both can be used to reduce shirking and turnover.

• Deferred pay is “cheaper” because PV of pay is same as best alternative, whereas efficiency wage pays more than best alternative.

• Drawbacks to deferred pay are listed above (risks to employee)

• Efficiency wage stories have been used to justify trade protection, affirmative action, minimum wages, wrongful discharge legislation, and a wide range of other government policies.

Deferred Pay vs. Efficiency Wages

• Efficiency wages may lead to employer discrimination against

•Elderly•Women

•Deferred pay

• May require legal intervention by government to make contract credible.

• employment law regarding wrongful discharge• ERISA

Tournament Theory

• Suppose that employees produce output according toQi= a*ei + ui + v

ei = person i’s effort

ui = random error affecting person i’s output

v = random error affecting all workers’ output e,u and v are unobservable to employer.

Tournament theory• How should firm compensate worker?

– Pay according to Q?• Pro: encourages effort by employee• Con: risk averse workers will avoid; must pay risk

premium. More costly as variance of e and u increases.

– Pay by time and attempt to monitor output• Pro: no risk premium• Con:

– costly to monitor if variance of e or u is high;– Incentive to provide effort is reduced.

Tournament theory• Pay according to relative performance

– Eliminates risk in pay due to common shocks (v)

– Most beneficial when Var(v) is large, Var(u) small.

• Possible problems– If different ability among workers, low level

workers may • Follow risky strategies• Not put forth effort if no chance of winning

• Hybrid schemes: tournament plus time pay.

Tournament TheoryKnoeber and Thurman:• use data on broiler producers facing both

tournament and linear performance evaluation compensation structures to test three predictions

• In mixed tournaments, more able players will choose less risky strategies

• tournament organizers handicap players to avoid the disincentive effects of mixed tournaments..

Testing the Theory of Tournaments: An Empirical Analysis of Broiler Production. Charles R. Knoeber; Walter N. ThurmanJournal of Labor Economics, Vol. 12, No. 2. (Apr., 1994), pp. 155-179.

Tournament Theory