Firms Wage Schemes: Implementation and Empirical Evidence

1
Topic 2 Part X
Firms Wage Schemes Implementation and Empirical
Evidence
2
Choice of Payment Scheme

• Output-based pay scheme is the most common form
  of individually-based-incentive pay for
  production workers
• For other type of economic activities, time-based
  pay and combinations of time-based pay with
  output-incentives and group-incentives are the
  payment schemes used

3
Time-Based Scheme

• Time-based scheme consists of compensating
  workers for the time they work
• This payment scheme reduces the workers risk of
  having his/her earnings vary on a weekly (or
  monthly) basis

4
Time-Based Scheme (cont.)

• Employers accept the risk of variations in
  productivity when workers are productive,
  profits increase and when they are less
  productive, profits decline
• Employers may be less anxious about this
  variation because they have more assets to deal
  with lean periods and because they have several
  employees and the chances that all will have low
  productivity at the same time is low

5
Implementation Problems

• But guaranteeing the worker a wage without
  reference to her/his actual output creates a
  problem of moral hazard why should the worker
  work hard if the effort is not rewarded?
• This danger generates the need for the employer
  to closely monitor the workers behavior. The
  problem with close supervision is that it is
  costly

6
Output-Based Scheme

• The output-based pay is also knows as
  pay-for-performance or piece-rate pay
• The most obvious way to motivate workers is to
  pay them based on their individual output
• Linking pay to output creates strong incentives
  for productivity

7
Output-Based Scheme (cont.)

• Under the output-based scheme, workers bear the
  consequences of low productivity
• Thus, employers can afford to spend less time
  screening and supervising workers
• But there are several implementation problems

8
Implementation Problems

• First problem the output-based scheme provides
  only weak incentives if the workers effort and
  the resulting output are not closely related
  (i.e., when workers output is influenced by
  forces outside his/her control)
• The worker will be willing for example to put
  forth 10 more effort if she can be sure her
  output (and pay) will rise by 10
• If due to random factors, the worker counts only
  with 5 increase, she may decide that the extra
  5 effort is not worth

9
Implementation Problems (cont.)

• The weak link between output and effort also put
  the workers earnings at risk of variation that
  she/he cannot control
• A risk-averse worker may be unwilling to take a
  job with such a pay scheme unless it pays a
  compensating wage differential for the
  uncertainty or a bonus
• Therefore, unless a workers output and effort
  are very closely associated, output-based pay may
  have small benefits to the employer and yet come
  at added costs

10
Implementation Problems (cont.)

• Second problem the need to choose an output
  measure that coincides with the employers
  ultimate objective but that can be objectively
  measured
• The need to link output-based to what can be
  objectively determined means that workers might
  be induced to allocate their efforts away from
  aspects of their performance that are not being
  measured

11
Implementation Problems (cont.)

• If workers get paid for the quantity of items
  that they individually produce or sell, they may
  have minimum regard for quality for example.
  Then, there is a need for costly monitoring
• Quantitative aspects of output are easier to
  measure than qualitative aspects (such
  helpfulness to customers) and yet the qualitative
  aspects are critical to building a loyal customer
  base
• Imperfectly designed performance measures can
  induce employees to allocate their effort toward
  what is being measured and away from other
  important aspects of their jobs for the success
  of the firm

12
Implementation Problems (cont.)

• Third problem workers may be induced to work so
  quickly that machines and tools are damaged
  through lack of proper maintenance or use
• This problem can be mitigated to the extent that
  production downtime can cause the workers
  earnings to drop
• Frequently, however, employers require
  output-based-pay workers to provide their own
  machines or tools

13
Earnings under the Two Schemes Empirical Evidence

• Seiler, E. (Review of Economics and Statistics,
  1984) studies earnings under output-based and
  time-based schemes
• Data individual, firm and industry level data on
  two US manufacturing industries (clothing
  industries)

14
Earnings under the Two Schemes Empirical
Evidence (cont.)

• Hypothesis to be tested Workers paid for their
  output earn more than those paid for their time,
  because the output-based pay
• motivates employees to work harder (create
  incentives)
• attracts the most productive workers (generate
  sorting)
• involves risk that may call forth a compensating
  wage differential

15
Earnings under the Two Schemes Empirical
Evidence (cont.)

• Findings
• Workers paid an output rate (piece rate) earned
  about 14 more than workers paid by hour
• About one-third of this disparity was a
  compensating differential (for risk)
• The remainder being related to the incentive
  (higher productivity) and sorting effects
  (attraction of productive workers)

16
Output-Based Scheme and Incentives Empirical
Evidence

• Paarsch, H. and Shearer, B. (Journal of Human
  Resources,1999) provide empirical evidence on the
  incentive effects of output-based schemes using
  data from a tree-planting firm in British
  Columbia
• Findings paying 1 cent more per seedling planted
  (on the basis of the average payment of 25 cents
  per seedling) increased average daily output by
  67 trees, holding planting conditions constant

17
Output-Based Scheme and Incentives Empirical
Evidence (cont.)

• Lazear, E. (American Economic Review, 2000)
  provides evidence on the effect on productivity
  of shifting from hourly wages to output-based
  rates
• Data worker data for a large US company that
  installs car windshields (Safelite Glass
  Corporation)

18
Output-Based Scheme and Incentives Empirical
Evidence (cont.)

• Findings the shift to output-based rate
  increased productivity (output per worker) by 44
• About half of this was due to the incentive
  effect of employees working harder because their
  effort was directly rewarded
• The other half was due to sorting the least
  productive workers left the company and more
  productive workers stayed or joined the company

19
Non-Monetary Considerations

• Issues of Fairness peoples concern about their
  treatment relative to others in their reference
  group means that fairness is an issue that
  pervades the employment relationship
• For example, a worker who obtains a 7 wage
  increase during a year in which prices and wages
  increase on average 4 might be quite happy until
  he finds out that a colleague working in the same
  job for the same employer received a 10
  increase
• Workers who feel unfairly treated may quit,
  reduce their effort level and even sabotage
  output to settle the score

20
Fairness and Frame

• Unfortunately for employers, the fairness of
  identical policy decision can be perceived
  differently depending on the context or frame
  of the policy

21
Fairness and Frame Empirical Evidence

• Kahneman, Knetsch and Thaler (American Economic
  Review, 1986) find that the framing determines
  the judgment of people about the fairness of a
  wage policy
• A sample of people (telephone surveys) was asked
  to consider the case of two small companies that
  were not growing as planned and therefore had a
  need to cut costs
• Each paid workers 10 per hour, but Employer A
  paid that in salary and Employer B paid 9 in
  salary and 1 in the form of a bonus

22
Fairness and Frame Empirical Evidence (cont.)

• Most respondents said it would be unfair for A to
  cut wages by 10, but they considered fair if B
  were to eliminate its bonus.
• Apparently, pay framed as salary indicates a
  greater entitlement (right) than pay framed as
  bonus

23
Alternative Pay Scheme Group-Output-Based Scheme

• When individual output is difficult to monitor,
  when individual incentive plans are detrimental
  to output quality, or when output is generated by
  teams of interdependent workers, firms sometimes
  adopt group- incentive pay scheme to more closely
  align the interests of employer and employee
• This plan tie at least a portion of pay to the
  group productivity

24
Implementation Problems

• Free-riding
• Decisions about shirking are taken at an
  individual level
• A person who works hard to increase group output
  or the firms profits ends up splitting the
  fruits of his effort with others
• Thus, free-rider opportunities give workers
  incentives to cheat on their fellow employees by
  shirking

25
Implementation Problems (cont.)

• Group incentives attract low productivity workers
  and generate incentives to leave on high
  productive workers
• In very small groups, however, cheating is easy
  to detect and peer pressure can be effectively
  used to eliminate it

26
Alternative Pay Scheme Time-Based Pay with Merit
Increases

• Given employee risk aversion and the problem of
  devising appropriate measurable outcomes for
  individual and group incentive plans, employers
  might opt for some form of time-based pay
• While satisfying employees desires for pay
  stability, time pay creates a moral hazard
  problem because compensation and output are not
  directly related

27
Alternative Pay Scheme Time-Based Pay with Merit
Increases (cont.)

• Employers deal with this problem by using an
  additional merit-pay plans, which award larger
  pay increases to workers whose supervisors rate
  them as the better performers
• The advantage of this scheme is that it creates
  superior incentives for workers because these
  plans take into account subjective aspects of
  performance (capacity of being a team player,
  etc.) that may be critical in the welfare of firm

28
Implementation Problems (cont.)

• If individual effort and output are not
  correlated due to forces beyond the control of
  worker, and supervisors use the worker
  contributions to actual output as a basis for
  rating, the incentives to the worker will not be
  the required
• To avoid this problem, supervisors should rate
  their subordinates relative to each other
  (according to a relative status), given that all
  are supposed to be equally influenced by external
  forces
• Relative status, however, can generate incentives
  to sabotage the work of others