T01 Huck et al. (2011): Deferred Compensation Flashcards
What is a deferred compensation contract?
Huck et al. (2011): Deferred Compensation
… workers are underpaid during the early part of their career and overpaid during the later part of their career.
Why would a manager choose a deferred compensation scheme?
Huck et al. (2011): Deferred Compensation
- Observability of effort: Effort is costly to observe or cannot be observed.
- Long-term relationship: Worker-relationship is expected to be long-term.
- Reducing turnover rates: It motivates people to stay in the firm.
- Discount rate: Attracting workers with low discount rates
What are the incentives of a deferred compensation scheme for the worker and the firm? What is the implication and what are the solutions?
Huck et al. (2011): Deferred Compensation
- Worker increase their effort in order to maintain in the firm because future expected pay exceeds outside option.
- Firm reneges on late-career overpayment because older workers are paid above their marginal productivity
- Implication: If there is no commitment mechanism there will be no deferred compensation scheme
- Solutions: legally binding contracts and reputation building
Why did the authors use a labarotary experiment, why is randomness important and what is the main disadvantage of a lab experiment?
Huck et al. (2011): Deferred Compensation
-
Testing the theory on observational data is difficult:
- Steep earnings due to monitoring difficulties
- Pensions designed to retain and motivate
- Flatter wage profiles due to mergers or high bankruptcy costs
- Productivity increase in later career
- Importance of random assignment:
- subjects differ exclusively by treatment
- allows to draw causal inferences
- Disadvantage: Lack of external validity => simulate real environment
What was the experimental design? (5 aspects)
Huck et al. (2011): Deferred Compensation
- 10 subjects randomly assigned ‘worker’ or ‘firm’
- 20 rounds á 3 career stages
- Firm offers wages W=(W1,W2,W3)
- Worker chooses e1,e2,e3=(low, medium, high)
- If e=low, worker may get fired
What are the treatment groups?
Huck et al. (2011): Deferred Compensation
- FCT (Full Commitment Treatment): Firms can fully commit to W2 and W3.
- NCT (No Commitment Treatment): Firms cannot commit to W2 and W3.
- CFT (Computer Firm Treatment): Computers offer the same W2 and W3 as humans in FCT.
- RT (Reputation Buliding Treatment): Firms cannot commit to W2 and W3 but worker know the history of the firm.
What are the orthodox predictions for the 4 treatments?
Huck et al. (2011): Deferred Compensation
- FCT:
- Never e=high, e=medium if…
- Stage 1: 2W2+W3≥80 or W2+W3≥60
- Stage 2: W3≥40
- NCT:
- Always low effort
- CFT:
- Same as FCT
- RT:
- depending on modeling approach
Explain the 2 social preferences used in the paper and how they measure it.
Huck et al. (2011): Deferred Compensation
- Reciprocity (pos. W11-coefficent): In response to friendly actions, people are much nicer and more cooperative than in orthodox model. Conversely, much more nasty and even brutal in response to hostile actions
- Inequity aversion (neg. Fehr-Schmitt rationalizable): People resist inequitable outcomes. They give up own payoff in the direction of more equitable outcomes.
What are the behavioral predictions for the treatments?
Huck et al. (2011): Deferred Compensation
- FCT: Reciprocal workers may exert e=high if W1,W2,W3 are perceived as a gift:
- W1>0
- 2W2+W3>80 or W2+W3>60
- W3>40
- FCT: Inequity averse workers may exert e=low even if
- 2W2+W3≥80 or W2+W3≥60
- W3≥40
- holds but W1+W2+W3 < 100.
- CFT: Reciprocal and inequity-averse workers behave as predicted by orthodox model
- Firms pay W1=0 and W2,W3=(60-100,0)
Sketch the predictions graphically.
Huck et al. (2011): Deferred Compensation
What is the econometric method the authors used?
Huck et al. (2011): Deferred Compensation
Multinomial logit regression
How to read the table?
Huck et al. (2011): Deferred Compensation
- W11: Reciprocal behavior (+ and sign.)
- W12+W13: Incentive effect of deferred pay on effort
- Lazear line: Increase in effort @Lazear line?
- Lazear line x (W12+W13): Incentive effect of deferred pay if medium effort is incentivized
- Fehr-Schmitt-Rationalizable: Inequity aversion (if -)
What does the table look like?
Huck et al. (2011): Deferred Compensation
What are the 5 results of FCT?
Huck et al. (2011): Deferred Compensation
- Workers exhibit reciprocal behavior
- Deferred pay affects e1
- Lazear line matters for effort decision
- Deferred pay is less effective if medium effort is incentivized
- Workers exhibit inequity aversion, which may lead to e=low
What are the 4 results of CFT?
Huck et al. (2011): Deferred Compensation
- Workers show no reciprocal behavior
- Deferred pay affects e1
- Increase in effort @Lazear line
- Deferred pay is less effective if medium effort is incentivized
- Workers show no inequity aversion towards computers