- Created by: Emily Moores
- Created on: 24-02-20 16:35
Personnel Economics Lecture 1
The Principle Agent Relationship
Why are labour markets not like other markets?
- Labour markets cannot be for very long periods of time
- Buyers and sellers have imperfect information - workers know there own quality and effort level but do not know about aspects of the job. Employers do not know employee's effort or quality.
- Relationship often long term- can change over time.
- Workers have specialised skills. Employers want workers to invest in specialised skills. Market for these very thin.
- Imply that the labour market does not satisfy the standard assumption of neoclassical markets.
- The law of supply and demand may broadly hold, but does not govern all worker and firm interactions.
What is Personnel Economics?
DEFINITION - A principal wants a particular job done but has insufficient time or skills to do so her self. Thus hires an agent to act on her behalf.
Personnel economics is the study of how the employer selects and motivates a worker.
Use economic theory to outline the nature of economic contracts, and use statistical analysis to examine the empirical relevance of different types of contracts.
- Choosing the right agent is difficult.
- Self-reported data can be unreliable- People lie on CV.
- References are meant to support a candidate not to distinguish between them.
- Agents have different reasons for looking for a job- advance career, do not get on with current employer.
- Job criteria hard to measure and unobservable- motivation, reliability.
- Education can be used as a screening mechanism, even when education does not directly affect the ability to perform a job.
- Drug testing screening.
The worker and employer may fundamentally differ with respect to:
- Attitude to risk: worker risk-averse, employer risk-neutral or less risk.
- Ultimate goal: Firm- maximise profit, worker- salary, effort level
- Time Horizon: Firm - short-run, Worker- long run(keep job)
- Information sets: Worker knows about day to day aspect of their job and their ability. Firm knows about business environment.
Because of these fundamental differences between employer and employee, it is difficult for firms to get optimal effort from their workers.
The problem of motivating worker effort
- Issue is that workers won't always follow the best interest of employer and courts won't enforce this.
- Opportunism - self-interest seeking - the worker tries to use their superior information to their advantage.
- Adverse selection: Hidden Information - agent has private information, principal can not determine if action was optimal without that information.
- Moral Hazard: Principle can not observe the agent's actions only the outcome. Effort can not be observed by the firm or courts. Cannot be deduced from outcomes- impossible to know if something fails if this is due to luck or lack of effort.
The fundamental issue of personnel economics is designing contracts that:
- Align the interests of employers and employees
- Can be enforced by a third party
Example: Landlord and tenant farmer
- A landowner lives in London, but owns a farm in France. He hires a farmer to work his land.
- Harvest depends on- weather, effort of the farmer.
- Landlord can never infer effort from output. Low output can be due to: Low effort or bad weather.
- This implies that effort is not contractable
- Therefore, need a self-enforcing contract.
Objectives of each party
Landlord wants to maximise profits. Max P = PQ(e) – w
- P is the price of the output
- Q is the amount of output
- Q is the compensation paid to the worker
- e is the effort level ( equal to 1 or 0)
Tenants objective function:
Max U(w, e) = w – 1.5e
Production Function: Q = a + be think of a as the weather
The principal-agent problem in this example:
1) Objectives are fundamentally different: Landlord cares about ouput, workers does not. Landlord values high effort, worker has a cost of providing effort.
2) Information is imperfect: Landlord and courts do not observe effort or a.
Thus if output is low it could be due to low effort or bad weather.
This implies that a court won't punish the worker for low effort if there is low output, so an effort based contract won't work.
The contracting solution
The landlord values high effort, the farmer values low effort.
High social optimal if
PH + UH > PL + UL i.e.
[P (a + b) – w] + (w – 1.5) > (Pa - w) + w
Solving this gives optimal high effort if:
Pb - 1.5 > 0
If low effort is socially optimal there is no contracting problem.
How can a firm motivate high effort?
Option 1: A Flat Wage
If a worker is paid a flat wage regardless of effort, them their utility is greater by exerting no effort:
U(e=0) = w > U(e=1) = w – 1.5
Option 2: Linear Wage Contract w = a + bQ
If the worker share (b) is optimal, the worker will choose high effort and both the worker and firm can be made better off.
The worker chooses high effort when marginal returns to effort is greater than the marginal cost.
U (E=0) = a + b(a) U (E=1) = a + b(a + b) – 1.5 To solve this set: U (E=1) > U (E=0)
Supply high effort is b > 3/2b Marginal returns to effort is greater than the marginal cost.
Different values of a and b in the linear wage contracts:(w= a+bQ)
- a> 0 and b=0 is a standard flat wage.
- a< 0 and b=1 is a standard rental contract. The farmer rents all the land from the landowner and keeps all the harvest. Risk associated with the harvest go to the farmer. The effort of the landlord my matter. There may be barns and fences that need replacing. If b=1 then the landlord has no incentive to supply effort.
- If 0<b<1 is a share contract: the risk associated with weather is shared between farmer and landlord. Both parties have incentive to supply effort as they earn more if larger harvest.
Reasons for incentive pay
In this example, incentive pay aligns the interest of principal and agent.
Another reason may be to attract the "right" employees
Consider two potential farmers: one with low ability and one with high ability.
The high ability farmer is more attracted to the share contract than the low ability because high ability means higher output and thus higher pay.
The low ability is more attracted to a fixed wage.
Worker selection in the farmer-landlord example
Suppose there are two farmers, i and k.
Assume i is more talented, his return to effort is twice that of k, bi = 2bk
Both have the same effort cost (1.5)
The share contract is w=a+bQ
The utilty is as follows:
Ui(E=0) = Uk(E=0) = a + b(a) Uk(E=1) = a + b(a + b) – 1.5 Ui(E=1) = a + b(a + 2b) – 1.5
Solve this for b (minimum share which will incentivise effort:
For k: b > 3/2b For i: b > 3/4b Thus if 3/2b > b > 3/4bi will supply effort, but k will not.
If, for example, the landlords offer a share of 0.9, worker i will agree but k will not.
Implication of landlord sucess
- Landlord prefers a wage just above 3/4b to one just above 3/2b because it costs them less.
- If Landlord x offers a wage of a+ 1/b and landlord Y offers a fixed wage a+ε , then i will take the offer from x, but k will take the offer from Y.
- This in turn means that landlord X will outperform landlord Y over time. Landlord X gets a better worker. Landlord X is able to motivate higher effort. The additional output will more than offset the additional wage cost.
Share contracts in modern employers
We don’t observe literal share-contracts
•A good car salesperson doesn’t receive a 10 car bonus •This is because workers don’t have a comparative advantage in selling or using the goods they produce Contracts with similar properties (pay depends on output) are very common: •Piece rates – pay per unit of output •Performance bonuses - individual (west) or team (Japan) based •Stocks and options - Primarily for upper management •Commissions - sales, cab drivers •Prizes (employee of the month) Other contracts reward long-term performance but do not tie pay to output •Deferred compensation and Pensions •Promotion “Tournaments”
So why do we ever observe wage contracts?
First order conditions (general form) •¶U/¶w > 0 and ¶U/¶e < 0 Second Order Condition •In the above example ¶2U/¶w2 = 0, this implies risk neutrality •Often in the real world, ¶2U/¶w2 < 0, workers are risk averse •Worker dislikes variance in pay beyond her control •Incentive contract Pay = f (output) = f (e, weather) •w = a + bQ = a + b(e, weather) in our example •To make a risk averse worker equally well off as with a fixed wage need higher E(W). Implies that incentive pay is more costly because the worker needs to be compensated for the added risk (due to weather)
If there is a strong relationship between the weather and output and/or a weak relationship between effort and output a wage contract may be better than a share
Problems with incentive pay
Risk - pay tied to factors beyond the worker's control
-"Second best policy": want to motivate effort, can't observe effort, tie pay to output, but output may only be loosely correlated to effort.
- Requires higher pay: workers are risk-averse, to make them equally well off with uncertain income, the values must be set higher.
- Poorly designed incentive pay can lead to workers taking too much risk.
Multitasking and getting incentives right.
-workers do more than one task
-If the are incentivised for only one, they put in too much effort on that task.
Risk and Incentive Pay
Example 1: Why aren't production line workers given large bonuses if company profits are high?
-Little control over company profits, pay tied to something out of the control of the worker.
-Without bonuses need different incentives e.g pay tied to promotion, service at the company, employee of the month.
CEO pay is almost always partly in stocks or options.
-Greater control over outcomes
-Maybe less risk-averse because of much higher income.
Risk and Incentive pay
Example 2: A Large mining corporation
-Two Possible projects:
1) Safe project - low risk, low return 2) Risky project - high risk, high return
Shareholders prefer 2 - higher expected returns, diversify portfolio.
Manager prefers 1 - low expected return, care more about risk as can't diversify.
-Firm may need to guarantee the manager's wages and give them a share of funds to induce optimal decision making.
The Multitasking problem and perverse incentives
In the farmer landlord example - the share contract provided incentives for the farmer to do the right thing ( social optimal).
Poorly designed contracts can have the opposite effect.
The Multi-tasking problem:
-In most firms, workers are part-contributors to multiple outputs and most outputs have multiple dimensions.
-Much more difficult to write a contract for the multi-task worker
Perverse Incentives due to multitasking
-Pay football players based on goals - need to work as a team, everyone tries to score
-Pay bankers on new accounts received - take too much risk
-Pay factory workers based on output- increase quantity, decrease quality.
Because of this problem firms often tie pay to subjective measures of overall performance rather than objective measures.
Why do leniency/centralised biases exist and why do they matter?
These biases exist because of a behavioural regularity - line managers don't like writing negative reviews are singling out staff.
This suggests that they will happen less if there is a greater social distance between appraiser and appraisee.
These biases matter because it weakens the management's ability to provide incentives.
-Prevents management from getting an honest appraisal of staff.
-Can't tie pay to performance because the measure of performance is poorly correlated to actual performance.
Incentive Contracts for Upper Management
Payment in Stock •Good performance results in a capital gain Bonuses •May or may not be linked to company profits •May be linked to other indicators of performance •May be individual or team based Stock options–right to future purchase at a specified P •Removes downside risk •Reduces incentives or creates perverse incentives after a negative shock “Phantom stock” •Pay depends on stock price •Manager receives no stock
Model of Executive Stock
Definitions: •K – option price (1 share) •n – number of shares available at the option price •e – per share productivity effect of effort (effort is binary low or high, i.e.= (0; 1)) •C – cost to the worker of high effort (cost of low effort = 0) •u – per share random output level u N(u; u) •V – per share total productivity: V = e + u •The nature of the option is that the executive can buy shares at price K at some point in the future. She will exercise this option if the price of the share is greater than K.
Option value = e + u - K if V> K 0 otherwise Value of the option if effort is high: •RH = n (e + u – K) * ProbH(V > K) – C Value of the option if effort is low: •RL = n (u – K) * ProbL(V > K) The firm will choose K and n such that RH> RL
An option scheme is different from a share contract in two ways: •no downside risk and •the principle has two instruments, number shares and share price No downside risk is preferred by a risk averse employee and strongly preferred by a less averse employee The absence of downside risk may induce excessive risk taking if K is too high •Once it becomes apparent that with a ‘normal’ strategy V>K, there is no cost to the employee of ‘going for broke’ •This is the ‘rogue trader’ scenario Reducing K will a) increase the probability that the option will be taken and b) increase the value of the option if it is taken Increasing n will increase the value of the option if it is taken Executive options will tend to be more effective if e is large or su is small