Intuition from lectures / tutorials

Question 1

How is implicit incentives defined in the lecture on career concerns?

Answer 1

Informal agreements that are self-enforcing within a repeated interaction because parties want to influence future outcomes in a favourable way.

Question 2

What is the key idea behind career concerns?

Answer 2

People are motivated to exert effort because good performance today may result in better employment opportunities in the future. Future employers try to deduce information on agent’s ability from his past performance. The higher the agent’s expected ability, the higher will be future wage offers.

Question 3

What is “signal jamming”?

Answer 3

Because performance is also increasing in effort, the agent exerts effort to appear more able than he actually is.

Question 4

How does effort impact wages?

Answer 4

An effort increase leads to higher expected future income. The larger t (the later in career), the smaller will be the impact of an effort increase. Thus, an agent’s optimal effort choice is decreasing in t.

Question 5

When are effort incentives high?

Answer 5

Effort incentives are high when A can strongly influence the market’s beliefs about his talent by increasing effort (signal jamming). In every period, the market obtains an additional signal on talent, resulting in more precise beliefs on talent that are harder for A to influence.

A probably chooses inefficiently high effort at the beginning and inefficiently low effort towards the end of his career.

Question 6

In equilibrium, the labor market perfectly anticipates A’s effort choice. Why does A nevertheless choose high effort levels when starting his career?

Answer 6

In its most basic version, the career concerns argument goes as follows. An agent (e.g., a manager) can take an action (exert costly effort) to improve the signal (output) that principals (potential employers) receive about her productive abilities. Even though all parties are symmetrically informed, the agent has an incentive to interfere with the principals’ updating by exerting effort to make the signal revealed to them look more favorable, i.e., engage in signal jamming. In equilibrium, principals take this into account and the agent’s signal is discounted accordingly. As a consequence, the agent is trapped into providing the effort to meet principals’ tougher standards and not to look worse than she actually is.

Question 7

What is optimal effort in second period when there is no noise term in the output? (Output = talent + effort)

Answer 7

When there is no noise term in output, it consists of ability plus the agent’s effort. Since the firms anticipate the agent’s effort perfectly in equilibrium and adjust their beliefs over ability accordingly, it follows that after observing output once, they perfectly know the agent’s ability. Thus, optimal effort = 0 and optimal wage = talent.

Question 8

What role does delta play in career concerns issues?

Answer 8

Delta captures the agent’s time preferences, and thus, a higher delta means that the agent values the future higher. Agents undertake effort today to improve tomorrow’s wage, and if they care more about the future (higher delta), they are more willing to endure effort costs today to increase future wages.

Question 9

What are the two different forms of human capital discussed in the lecture on human capital?

Answer 9

General human capital: an investment in general human capital raises the worker’s productivity at the firm he currently works for and at some other firms as well (by an identical amount).

Firm-specific human capital: an investment in firm-specific human capital makes a worker more productive at the firm he currently works for, but has no effect on his productivity elsewhere. Examples: sales representative who collects information about individual clients, knowledge of a product sold only by the current employer.

Question 10

Can you explain Gary Becker’s suggested solution for how to share the costs and benefits of training between employee and employer?

Answer 10

Picture the graph in the slides. At the beginning, both parties share the costs: worker accepts a wage below the market wage, and the firm pays the worker more than his productivity. At some point, the value inside the firm exceeds the value outside the firm due to acquisition of firm-specific HC. Then, the worker receives a wage that is above his market average but below his productivity inside the firm.

Question 11

What is a compressed wage structure?

Answer 11

Wage increases in training but less steeply than productivity. (Effekten av trening på produktivitet er større enn effekten av trening på lønn)

Question 12

What type of labor market imperfections can lead to wage compression?

Answer 12

(1) Asymmetric information on talent / ability
(2) Firm-specific human capital
(3) Moral hazard
(4) Minimum wages

Question 13

If a worker’s production function is f(t , n) where t = training and n = talent, what wage does the general labor market offer?

Answer 13

V(t) - the worker’s reservation wage, because the labor market only observes training. The current employer, however, observes both talent and training at the end of the first period and offers w(t , n).

Question 14

How do you calculate a worker’s equilibrium reservation value v(t)?

Answer 14

Training * (probability of talented worker quitting / probability of any worker quitting)

Question 15

Will employer invest in general human capital if talent and training are substitutes (f(t,n)=t+n)?

Answer 15

No, because then, training is equally effective for both types of workers - outside wage offers increase at the same rate as inside-firm productivity in training intensity. No wage compression.

Question 16

What is inequality aversion?

Answer 16

People get disutility from earning less than others and, possibly, also from earning more than others.

Question 17

What is positive (negative) reciprocity?

Answer 17

People exhibit positive (negative) reciprocity when they reward (punish) proposer with good (bad) intentions.

Question 18

What are the three legs of the three-legged stool illustrating organisational architecture?

Answer 18

Monetary rewards, performance evaluation and decision rights

Question 19

Why would employers want to delegate authority? (What’s the rationale for doing so?)

Answer 19

Relevant information is often dispersed across lower hierarchy levels, and there is a cost of accurately transferring relevant information from the source to the decision maker.

Question 20

What is a central problem with delegation of authority?

Answer 20

Those with the relevant information may have goals that don’t align with those of the company, which potentially leads to inefficient decisions.

Question 21

Delegation can be decomposed in what four steps according to Fama and Jensen (1983)?

Answer 21

(1) Initiation
(2) Ratification
(3) Implementation
(4) Monitoring

They claim that these steps should not be delegated to the same person (to avoid abuse of power).

Question 22

Which advantages of delegation are highlighted by Again and Tirole (1997)?

Answer 22

Motivate the employee, make employment more attractive, increase the employees’ incentives to communicate relevant information

Question 23

What does the incentive view of delegation say about the effect of delegation on effort?

Answer 23

The delegation of formal authority to the agent raises his effort but lowers the effort of the principal.

Question 24

What does the participation view of delegation say about whether decision j should be delegated?

Answer 24

Decision j should be implemented if (1) the decision is relatively unimportant to P or (2) the congruence parameter is sufficiently large? For example, “what to wear at work” should be up to A if it is not important to P, but it should not be delegated to A if it is important to P and A would like to do otherwise.

Question 25

What does an agent do with prior knowledge on projects under P-formal authority (A-formal authority)?

Answer 25

Under P-formal authority, A communicates his information if and only if congruence is sufficiently high and A’s marginal cost of effort are high.

Under A-formal authority, A always communicates his information.

Question 26

Can you mention two problems with incentive provision?

Answer 26

(1) In some tasks, effort is more easily monitored and/or output is more easily measured than in others

(2) Compensating employees based on what is measurable can make them focus on these tasks and neglect the remaining tasks

Question 27

What are the main characteristics of (1) partial delegation (PD), (2) non-specialised complete delegation (NSCD), and (3) specialised complete design (SCD)?

Answer 27

(1) partial delegation (PD): principal performs one task himself and delegate the other one

(2) non-specialised complete delegation (NSCD): both tasks are performed by a single agent

(3) specialised complete design (SCD): tasks are delegated to different agents

Question 28

Will employer invest in general human capital if talent and training are substitutes (f(t,n)=t+n)?

Answer 28

No, because then, training is equally effective for both types of workers - outside wage offers increase at the same rate as inside-firm productivity in training intensity. No wage compression.

Question 29

Explain task interaction in terms of c.

Answer 29

Substitutes if C > 0
Complements if C < 0
Independent if C = 0

Question 30

If tasks are technologically interdependent, how can incentives for a given task be provided?

Answer 30

Increasing the reward for the task

Substitutes: decreasing the reward for the other task, thereby decreasing the opportunity cost for the given task

Complements: increasing the reward for the other task, thereby indirectly decreasing the cost for the given task

Question 31

When should an outside activity be allowed at work?

Answer 31

When the benefit is larger than its cost

vk(tk(b)) > p*tk(b)

Question 32

How does the incentive parameter b affect whether an outside activity should be allowed?

Answer 32

The larger the b, the more outside activities should be allowed.

Question 33

How are agency costs (loss for the firm due to the moral hazard problem) affected by agents’ risk aversion?

Answer 33

They are increasing in the agents’ risk aversion r and the uncertainty of the process sigma squared.

Question 34

Will employer invest in general human capital if talent and training are substitutes (f(t,n)=t+n)?

Answer 34

No, because then, training is equally effective for both types of workers - outside wage offers increase at the same rate as inside-firm productivity in training intensity. No wage compression.

Question 35

What are potential problems of principal-supervisor-agent relationships?

Answer 35

Supervisor has no incentive to report honestly to principal. Supervisor’s and principal’s goals might not be aligned.

Question 36

Is it optimal from the principal’s point of view to induce the first-best effort levels?

Answer 36

In an ideal scenario where we have no financial constraints or any frictions, first best effort levels would be optimal. Information asymmetry, moral hazard, agency costs and other factors might however make it less optimal.

Question 37

What is first best effort?

Answer 37

In the context of principal-agent relationships, the first-best effort level is the level of effort that the principal would like the agent to exert in order to achieve the best possible outcome for the principal’s interests. The principal’s goal is to align the agent’s efforts with their own objectives to maximize overall utility or performance.

Question 38

Discuss one advantage and one disadvantage of a change from complete to partial delegation (NSCD to PD).

Answer 38

A disadvantage of PD is potential free-riding. Free-riding is a common consequence of trying to incentivise two or more people with a share of the same output.

If the tasks complement each other, it is disadvantageous to split them as this will cause inefficiencies. You want to bind complementary tasks together, because effort in one task helps the other task. If they are substitutes, splitting them is advantageous. Putting substituting tasks together will increase costs.

Wether the principal should be involved in the production process depends on risk preferences.

Question 39

Suppose that k = 0 (no interaction between tasks) and that the principal has access to signals for each individual task, x1 = e1 + ε1 and x2 = e2 + ε2, in addition to output. How do the incentive parameters associated with each signal vary with the agent’s task allocation? How does this change as r tends toward 0?

Answer 39

Free-riding problem disappears when we have separate signals that the payment can depend on. For a risk averse agent, payments should be conditioned on any signal which is informative about the agent’s task. For such agents, we want to include as much information as possible into the contract. If we have an informative signal about the agent’s effort, we always want to include it into the contract. When r tends toward 0 we can achieve the first-best solution.