Session 5: Agency Theory II: Applications

Question 1

JENSEN & MURPHY (1990): Purpose of paper

Answer 1

The purpose of the paper is to estimate the magnitude of the incentives provided to CEOs

Question 2

JENSEN & MURPHY (1990): Findings

Answer 2

Shareholders and CEOs: The estimates of the pay-performance relation for CEOs indicate that CEO wealth changes $3.25 for every $1,000 change in shareholder wealth.

CEO ownership: In the sample, CEOs hold a median of about 0.25% of their firms’ common stock.

Company size: In large firms, CEOs tend to own less stock and have less compensation-based incentives than CEOs in smaller firms.

Bonuses: Although bonuses represent 50% of CEO salary, such bonuses are awarded in ways that are not highly sensitive to performance measured by changes in market value of equity, earnings, etc

Question 3

JENSEN & MURPHY (1990): Why so little use of performance-based pay?

Answer 3

J&M’s explanation: The public (including regulators) disapproves of high salaries  Hence, in order to attract CEOs, boards have to limit pay cuts and firings.

CEO salary public? If CEO pay was not public, pay would be higher. Data for CEO pay for the 1930s confirm this (in these years, CEO pay was not public)

Question 4

JENSEN & MURPHY (1990): Why CEOs are paid so well

Answer 4

• Conspiracies between well-connected aspiring CEOs to not underbid each other.
• Imitation: Very high exec pay is a managerial practice that gives you credibility as a firm.
• Skimming: CEOs are better able to manipulate their pay (e.g., in many stock option program, base pay wasn’t adjusted)
• Much greater risk for being dismissed, i.e. they have to be paid more to take on risk.
• Productivity: CEOs are talented and the competitive labor market rewards CEOs for their contribution to performance.
• Agency perspective: CEO actions have massive consequences for wealth creation, thus CEO interests must be aligned with shareholders

Question 5

JENSEN & MURPHY (1990): Performance implications (Do J&M show that executive pay cause firm performance?)

Answer 5

• Difficult to estimate what part of performance CEOs actually contribute to, unless they are evaluated based on the entire firm’s profitability.
• CEOs choose to work hard even though they are not given a high reward.
• The regression cannot show causes (Only correlation)

Question 6

JENSEN & MURPHY (1990): Efficiency Implications (In your view, do they show that the US corporations they studied were run inefficiently?)

Answer 6

• It may be considered attractive for shareholders to minimize executive pay, assuming that performance remains high.
• They show that (b) is very small; Agency theory would tell us that all else equal, this would be bad
• There were limits to the pay to be paid to CEOs, because the public does not want salary inequality

Question 7

JENSEN & MURPHY (1990): Conclusions

Answer 7

The relation between CEO wealth and shareholder wealth is small and has fallen by an order of magnitude in the last 50 years. We hypothesize that political forces operating both in the public sector and inside organizations limit large payoffs for good performance

Question 8

Lazear (2000): Purpose

Answer 8

How sensitive is worker behavior to incentives and what specific changes in behavior are elicited?

Case-based: On data from Safelite Glass Corporation

Question 9

Lazear (2000): Compensation structure (before and after change)

Answer 9

Before the change: Fixed wage
Produce q_0, otherwise employees would get fired.
Effect is that everyone produced q_0, because there were no incentives.
After the change: Piece rates were introduced
Everyone still had to produce at least q_0.
Lazy employees would still produce q_0
Ambitious workers producing more than q_0, would get compensated per piece.

Question 10

Lazear (2000): Results

Answer 10

Firm-level output ↑ Profits and wages ↑
* Able/ambitious workers selected into the firm (“sorting”) (average ability ↑).
* Output per worker increased dramatically (44 %).
* Variance of (output/worker) ↑; more able/ambitious workers increased their output more.

Question 11

Lazear (2000): Conclusion

Answer 11

The results imply that productivity effects associated with the switch from hourly wages to piece rates are quite large

NB: Works only for companies where it is very easy to measure individual performance; Otherwise monitoring would be too high.