L17 - Asymmetric Information : Incentives I Flashcards
RECAP:
What is private information?
What is a moral hazard?
What is adverse selection?
- Private Information –> Information that only a subset of players know
- Moral hazard –> ASym info when one side of the market cannot observe all actions
- Adverse selection –> ASym info when one side of the market cannot observe all ‘ types’
what is the point of incentives?
- In many situations, one party wants to prompt another party to complete a task in a certain way.
- Market: employers and workers, shareholders and CEO, manufacturer and retailer
- Government: contractors, tax evaders, criminals, general public Covid, Climate Change
- Social: parent and child, landlord and tenants, boyfriend and girlfriend
- In all of these relationships, there are a number of key characteristics:
- • A principal delegates a task to an agent
- . • The principal collects all the revenue or wider benefits – he is the residual claimant.
- • However, the agent has control of some decisions. Hence, there is a separation of ownership from control.
- • There is a divergence in objectives
How can the principal best design a set of incentives to influence the agent?
What are some examples of incentives in a commercial context?
In commercial contexts, we see a variety of different incentive systems and payment methods. As part of this topic, we will learn more about their relative merits:
- Pay per hour: e.g. salaried work, casual labour, some construction jobs
- Pay per unit of output: e.g. agriculture, production
- Fixed payment contracts: e.g. IT projects, construction jobs, tenders
- Revenue Sharing: e.g. agriculture, commission, informal payment arrangements
- Franchises: e.g. McDonalds.
- Salary (and Bonuses): e.g. Bankers
What is the setup for the basic incentives model?
- Incentives I and II will be focusing on the incentive design under Full Information
-
Payments are only transfers
*
![](https://s3.amazonaws.com/brainscape-prod/system/cm/365/639/926/a_image_thumb.png?1638721550)
What is the social optimum of the incentive model?
Note, we will assume S(x*)> U0, otherwise, the Principal would never want the agent to incur any effort.
![](https://s3.amazonaws.com/brainscape-prod/system/cm/365/640/285/a_image_thumb.png?1638723837)
How do you solve for x* in the incentives model?
Within our example , where C(x) = cx and E(x) = x0.5, verify that x* = (1/4c2)
- Sub in for S(x) = R(x) - C(x)
- Then maximise (differentiate)
![](https://s3.amazonaws.com/brainscape-prod/system/cm/365/644/558/a_image_thumb.png?1638724126)
How can the principal design the incentives for the agent?
- Suppose the owner wants to induce the farmer to exert some effort level, x’.
- What general features must any effective payment schedule exhibit?
- First, the payment schedule must prompt the farmer to prefer exerting x’ rather than taking his outside option, 𝑈(𝑥′) = 𝑧(𝑥𝑥′) − 𝐶𝐶(𝑥′) ≥ 𝑈0.
- This is called a Participation Constraint (PC).
- Second, the payment schedule must prompt the farmer to prefer exerting x’ rather than any other level of effort, x. That is, 𝑈(𝑥′) = 𝑧(𝑥′) − 𝐶(𝑥′) ≥ 𝑈(𝑥) = 𝑧(𝑥) − 𝐶(𝑥) for all 𝑥 ≠ 𝑥′.
- This is called an Incentive Compatibility Constraint (ICC).
- We now consider the properties of some possible payment schedules, z(x)
- Take it or leave it Offer
- Revenue Sharing (Sharecropping)
- Renting/ Franchising
What happens under the Take it or Leave it incentive structure?
- As a simple option, the owner could set z(x) in the form of a payment, B’, if and only if the farmer puts in an effort of x’. #Provided B’ is sufficiently large, the farmer will prefer x’ to any other level of x, as all other effort level remain unpaid and so the Incentive Compatibility Constraint will be satisfied.
- But in order to prefer x’ to taking the outside option, requires U(x’) = B’- C(x’) ≥ U0 (PC)
- Therefore, the owner can profitably set the minimum possible payment, B’ = C(x’)+ U0, to ensure this is just satisfied.
- At this level, the payment is just enough to compensate for the effects of the outside option and the effort costs associated with choosing x’.
- Under this contract:
- The farmer is just willing to work, U(x’) = B’- C(x’) = C(x’)+ U0 - C(x’) = U0.
- Just indifferent from working for me or walking down the road and working for someone else
- While, the owner receives V(x’) = R(x’) – B’ = R(x’) - C(x’) - U0 = S(x’) – U0
- Revenue - benefit payment –> left with total surplus - utility I’m giving to the farmer
- The farmer is just willing to work, U(x’) = B’- C(x’) = C(x’)+ U0 - C(x’) = U0.
- This is very intuitive:
- • The owner need only give the farmer just enough surplus for the farmer to prefer x’ to the outside option, which gives the farmer a surplus of U0
- • As the owner is the residual claimant and receives any other generated surplus, he then gains the remaining S(x’) - U0.
- But which level of effort does owner actually wish the agent to exert?
- As the owner gains S(x’) – U0, the owner would like the agent to select the effort level that maximises total surplus, S(x’), and so he prefers x’=x*.
- Hence, in eqm, the owner receives S(x*) – U0 and the farmer receives U0.
- The contract is efficient in the sense that it maximises social welfare
What would happen under the Revenue Sharing incentive design?
** HOW DOES THIS SATIFY THE PC AND ICC
- x* maximises the social differences between the revenues and the cost curve
- With revenue sharing, there is fewer incentives to put in work as they no they will not get the full return for their work, only a fraction
- So they will only maximise at x(hat) instead
- With revenue sharing, there is fewer incentives to put in work as they no they will not get the full return for their work, only a fraction
![](https://s3.amazonaws.com/brainscape-prod/system/cm/365/792/627/a_image_thumb.png?1638820366)
What would happen under the Renting/Franchising incentive design?
- Like a franchise, you keep all the revenues you are going to make, but you have to pay me to work on the land
- Therefore, in this case, the owner is now in a position to actually use F as a fee rather than payment as S(x*) > U0
- F can even be negative ( farmer pays the owner to actually do it as working here is much better than going down the road and getting just U0)
Franchises include McDonald’s, Subway, Pubs (Marestons?), Shell garage, Car dealerships, Government sometimes franchise thing (trains franchised to the highest bidder), chairs in a hairdresser, American Footballl
![](https://s3.amazonaws.com/brainscape-prod/system/cm/365/793/931/a_image_thumb.png?1638820991)
Evidence from China about incentive Design
- As we have shown, theory suggests that the effort of workers should depend on the type of their payment contract. T
- he 1978 Agricultural reform in Chinese rural communes, allowed farmers to sell any production beyond a set quota on the private market.
- (Beyond the quota), this had the effect of increasing incentives to produce efficiently, much like the renting/franchise contract.
- Figures suggest that agricultural production rose by 61% from 1978-1984.
- McMillan et al (1989)0F 1 estimated that three-quarters of this rise was due to the increased incentives, rather than other changes in price controls
From Varian book
Evidence - Foster and Rosenweig (1994)
- Measuring how effort varies with contract type is hard.
- However, these authors form a proxy of effort by exploiting the biological relationship between energy intake, energy expenditure and body mass index
- Using farmers in the Philippines, they take regular measurements of food consumption and body mass index to infer energy expenditure. The results are consistent with theory.
- After controlling for other activities, they find that piece-rate workers (that are paid by output) deplete 10% more of their BMI than time-rate workers (that are paid by the hour).
- Moreover, self-employed workers (who receive all revenues) deplete 13 % more of their BMI than time-rate worker