Real estate projects are big, complicated and involve many actors. Therefore, success often depends on a legally binding contract to ensure that things get done with as few disputes as possible.
The first barrier to finding an answer is to describe the situation accurately: e.g.
- What is the source of uncertainty? If there is none then it is too easy to detect misbehaviour.
- are there only two outcomes (“success” and “failure”) or a continuum between really good to really bad?
A clear theory helps immensely.
The second puzzle is that changing the incentive changes the anticipated behaviour, which changes the outcome and the degree of risk. Think how the behaviour of a real estate agent might change if, instead of being paid five percent of the selling price, they were paid 50 percent of the price if the price were to exceed a reserve price (to be negotiated in the agency contract) or if the percentage fell as time passed. (The law on real estate contracts in Ontario (Section 36.2) says that this commission scheme is illegal, which raises an interesting puzzle about who benefits from the current rules and who bears the costs.) Or, think how the behaviour of banks might change if, when deciding on a high LTV mortgage and similar to what happens with co-payments in health insurance, a mortgage insurer required them to accept the first 10 or 20 percent of any shortfall (as opposed to CMHC’s current policy, which exposes taxpayers to the cost of that risk).
Hart and Holmstrom identified a number of other features which a contract must navigate.
If you believe my previous suggestion that one party to the contract should be given high powered incentives to behave well then it is important to make sure that their job is simple. In complex team projects, co-operation has a bigger effect on the outcome than any one dimension. It is important to have the electricians, plumbers, drywall installer, and painter work with each other, as opposed to each selfishly bumping into one another to complete their part of the job quickest.
Or, incentives can seem like a good way to encourage a supplier to lower costs but, if costs are manipulable in hidden ways then increasing the incentive can change “lowering costs” into “cutting corners”. A key insight is that, if the incentives are not compatible with the anticipated actions then there will be problems.
The Laureates’ work also shows that a contractible variable is a good choice if it passes the “informativeness principle”. For example, paying a CEO based on the company’s share price may seem reasonable, because the actions of the CEO affect the company’s success in ways which benefit shareholders. That intuition is not wrong but deeper thought shows that this intuition reveals a lack of imagination: the contractible variable may not be as informative as you hope. The share price is also affected by other things which the CEO does not control, such as general trends in the stock market. It would be more informative to pay a CEO based on the share price relative to competitors  .
In our imperfect world, a well-crafted contract adds value to a project by having the parties work toward a common goal while acknowledging that they do not share a common interest.