What if a loan is not repaidMaking promises credible
What is lent is today's grain; what comes back is only a promise about the future. If the borrower does not repay after the harvest and the lender has no recourse, no one will lend in the first place, and all the gains from lending are lost. There are three ways to make a promise credible: make the borrower care about the future, which is reputation; give the borrower something to lose, which is collateral; or let a third party enforce it, which is contracts and rulings. All three have costs, and all three shut some people out. Raising the interest rate cannot replace trust: the higher the rate, the more likely it is that those who come to borrow will not repay.
1. Loans not repaid: unable or unwilling
The island has grown more populous, and it has become customary for those who harvest in autumn to lend grain to those who harvest in spring. Yet when spring comes, some people always fail to repay. There are two kinds of failure. One is inability: the spring harvest was poor and the borrower has no grain to give. The other is unwillingness: the grain is in the store, but the borrower does not want to hand it over.
The two kinds of default differ in nature. Inability comes from uncertain harvests and is a matter of risk. This chapter assumes that harvests always arrive as expected and considers only unwillingness. Default then becomes a choice: the borrower weighs what repaying and not repaying would each bring, and decides.
Today's counterpart. When lenders assess a borrower, they usually look at two things separately: ability to repay and willingness to repay. The first is judged from income and debts, the second from whether the borrower has kept past agreements.
Unable to repay, unwilling to repay
Click a situation to see which kind of default it is. This chapter considers only unwillingness; inability comes from uncertain harvests and is left until risk is discussed.
Unable to repayHas nothing to give: a matter of risk
Unwilling to repayCould give, chooses not to: this chapter's problem
2. In a one-off encounter, not repaying is rational
Imagine two people who will never meet again. When spring comes, the borrower has two choices: repay and lose the principal and interest, or not repay and lose nothing. As long as not repaying has no consequences, it is the better choice. The lender can foresee this, so in autumn he does not lend. Both of them lose the gains from lending, gains that would have been enough to make both better off.
This reasoning reveals the structure of the commitment problem. The question is not whether the borrower is a good person; it is that at the moment of repayment, repaying no longer does him any good. For lending to happen, his gains and losses at that moment must be changed.
Today's counterpart. Strangers rarely lend money to each other, and when they do, only small sums.
Reasoning back from the moment of repayment
The lender first decides whether to lend; the borrower decides in spring whether to repay. Switch between "Deal only once" and "Deal every year" to see the borrower's gains and losses at the moment of repayment, and the lender's choice beforehand.
Illustrative: figures in brackets are (lender, borrower) gains or losses relative to not lending, in portions. When the two deal every year, a borrower who does not repay can never borrow again and loses the gains from borrowing in every later year, counted here as 5 portions.
3. Reputation: making the borrower care about the future
If the two must deal with each other year after year, the situation changes. Fail to repay this time, and there will be no loan next time. At the moment of repayment, the borrower is no longer weighing only the principal and interest; he is weighing them against the gains from all future borrowing. As long as those future gains are large enough and the borrower cares enough about the future, repaying is worth more than not repaying.
The force of reputation also depends on how far the record travels. If only the person who was cheated knows about a default, the defaulter can simply borrow from someone else. If the whole village knows, the defaulter will never borrow again.
Traces of this mechanism can be found in history. Greif (1993) used the letters of eleventh-century Maghribi traders in the Mediterranean to argue that traders shared information about who had cheated whom, which kept distant agents from cheating. This interpretation is disputed: Edwards and Ogilvie (2012) argue that these traders also relied heavily on courts and formal contracts.
Today's counterpart. Credit records make a single default visible to every lender, and a credit score compresses a history of kept agreements into one number. The easier records are to obtain and the more reliable they are, the stronger the discipline of reputation.
The force of reputation: caring about the future, records that travel
Not repaying once gains 4 portions now. Drag how much the borrower cares about the future, switch how far a record of default travels, and compare what default gains with the future it loses.
Default gains now
Default loses in future
Illustrative: the village has 10 households able to lend, and the borrower gains 1 portion a year from borrowing. "How much the borrower cares about the future" is what one portion next year is worth as a percentage of one portion today. When only the cheated lender knows, the defaulter loses only his dealings with that household and can still borrow from the other 9.
4. Collateral: giving the borrower something to lose
Reputation depends on the future; collateral replaces the promise with something present. The borrower pledges a plot of land, an ox or part of his stored grain to the lender, and if he does not repay, the pledged item goes to the lender. As long as the collateral is worth at least the principal and interest owed, not repaying does not pay, and even if the borrower defaults, the lender recovers the loss. Collateral has a second role. The lender can offer two sets of terms at once: more collateral and lower interest, or less collateral and higher interest. Those confident they will repay are not afraid of losing the collateral and choose the first; those who are not confident choose the second. Which one a borrower picks reveals how confident he is.
Today's counterpart. Mortgages are secured on the house, car loans on the car, and securities-backed lending and margin trading on the securities held. Young people without assets get worse terms on both the amount they can borrow and the interest rate.
Collateral: safer, and it shuts some people out
Each household borrows 4 portions and must repay 4.4 in spring. Drag the share of principal and interest the collateral must be worth, and see who can borrow and how much the lender loses on default.
Can borrowNot enough collateral, cannot borrow
Households that can borrow
Of those, households for whom default pays
Illustrative: the 20 households own collateral worth from 0 to 9 portions; the number in each cell is the value of that household's collateral. When the collateral is worth less than the principal and interest owed, not repaying pays.
5. Third-party enforcement, and the three approaches compared
The third approach is to ask a third party that both sides recognize to enforce the agreement: write it down as a contract, and in case of default let elders, a court or the authorities rule on it and compel the borrower to repay. Third-party enforcement does not depend on the borrower caring about the future, nor does it require him to have something to pledge. But it has its own costs: someone must record contracts, hear disputes and enforce rulings, and the enforcer must itself be trustworthy.
Each of the three approaches has its own range: reputation suits acquaintances who deal repeatedly, collateral suits people with assets, and third-party enforcement suits large transactions between strangers. They can also be used together.
Today's counterpart. Loan contracts, notarization and court judgments are third-party enforcement. Lenders usually check credit records, require collateral or a guarantee, and sign a contract, all at once.
Three ways to make a promise credible
Click an approach to see what it relies on, whom it suits, what it costs and whom it shuts out.
6. Interest rates cannot replace trust
Since some people do not repay, can the lender simply raise the interest rate and use the extra interest to cover losses from default? Work through how the pool of borrowers changes. As the rate rises, prudent borrowers find the loan no longer worthwhile and drop out first. Those still willing to pay high interest are mostly people planning to take big risks: if the venture succeeds they earn a great deal, and if it fails they do not repay. The higher the rate, the larger the share of such people among those who remain. Beyond a certain point, a further rise brings in less extra interest than it adds in defaults, and the lender's expected return actually falls. The lender therefore prefers not to raise the price, and instead limits whom he lends to and how much.
Today's counterpart. Faced with higher-risk applicants, lenders commonly decline the application or lower the credit limit rather than simply raising the interest rate.
The higher the rate, the more likely borrowers are not to repay
70% of the village are prudent borrowers and 30% are risk-takers. Drag the interest rate to see who still comes to borrow, and how much the lender recovers on average for each portion lent.
Prudent borrowersRisk-taking borrowers
Illustrative: prudent borrowers always repay but will not borrow at rates above 30%; only half of risk-taking borrowers can repay, but they will accept rates of up to 100%. The vertical axis is the lender's average net return per portion lent.
7. A village that keeps its word: Civilization Simulator, version 3
The simulator unlocks the rule of "Commitment". In the village, twenty households that harvest in spring borrow grain from households that harvest in autumn. Some of them keep their word; the rest are opportunists, who will not repay as long as not repaying has no consequences. You can choose the approach used to make promises credible, and adjust the share of borrowers who keep their word, how much people care about the future and the share who have collateral, then see whether lending can be sustained over twenty years.
With no approach in place, as long as fewer than 80% keep their word, the defaults of the first year frighten lenders away, lending stops at once and all its gains are lost. With reputation, when borrowers care enough about the future, even opportunists repay on time; when they do not care enough, defaulters are excluded in the first year and those who keep their word go on borrowing. With collateral, the 7 households with enough collateral can all borrow, while the other 13 can never borrow, whether or not they would keep their word.
Civilization Simulator · Chapter 3: a village that keeps its word
Twenty households that harvest in spring borrow grain each year from households that harvest in autumn. Choose an approach to making promises credible, adjust the share of borrowers who keep their word and how much people care about the future, and see whether lending can be sustained over twenty years.
Borrowed and repaid on timeBorrowed but did not repayCould not borrow
Illustrative model: each household borrows 4 portions a year and repays 4.4 in spring; each loan repaid on time brings the two sides 1 portion of gain in total; not repaying gains 4 portions now. Opportunists default only when defaulting pays better. With no approach, they always default. With reputation, they default if the future lost (care for the future ÷ (1 − care for the future) × 1 portion) is less than 4 portions, and are then refused by the whole village. With collateral, they repay if the collateral is worth at least the principal and interest. With no approach, lenders decide whether to keep lending by last year's repayment rate, and stop if it falls below 80%. Of the twenty households, 7 have collateral worth at least the principal and interest.
8. What this step leaves open
Once promises have become credible, several problems remain unsolved:
Finding the right counterparty is still hard. The more people there are, the harder it is to match those with surplus grain to those who lack it, and exchange itself has a cost.
Old age is still unsolved. However credible promises are, the youngest generation cannot find anyone younger to lend to.
Default from inability remains. This chapter assumed harvests always arrive as expected; once a harvest fails, even the most trustworthy cannot repay.
People without assets are shut out. Collateral makes lending safe, and also leaves those who most need to borrow unable to.
What this step leaves open
Four problems still unsolved once promises are credible. Click a problem to see the assumption it hides.
A village that keeps its wordReputation, collateral and enforcement
9. Implications for products
Almost every credit product today is some combination of the three approaches in this chapter. Three implications follow. First, a lending product must begin by answering why anyone should believe the borrower will repay; the only answers are reputation, collateral, third-party enforcement, or a combination of them. Second, the force of reputation depends on the reach and reliability of records: the more complete and the harder to forge the records, the more credible the promise and the cheaper the loan. Third, interest rates cannot replace trust. Covering risk by raising the price drives prudent users away; the better course is to improve identification and constraint.
Chapter summary: making promises credible
Four conclusions, each tied to one mechanism in this chapter.
The commitment problemAt the moment of repayment, repaying no longer pays
In a one-off encounter not repaying is rational, so lending never happens at all.
Three approachesReputation, collateral and third-party enforcement
Make the borrower care about the future, give him something to lose, or let a third party enforce the promise.
CostsEvery approach shuts some people out
Newcomers have no record, people without assets have no collateral, and small loans cannot bear the cost of enforcement.
Interest ratesInterest rates cannot replace trust
The higher the rate, the sooner prudent borrowers leave, and the more likely those who remain are not to repay.
References
Bester, H. (1985). Screening vs. rationing in credit markets with imperfect information. American Economic Review, 75(4), 850–855.
Edwards, J., & Ogilvie, S. (2012). Contract enforcement, institutions, and social capital: The Maghribi traders reappraised. Economic History Review, 65(2), 421–444.
Greif, A. (1993). Contract enforceability and economic institutions in early trade: The Maghribi traders' coalition. American Economic Review, 83(3), 525–548.
Kreps, D. M., Milgrom, P., Roberts, J., & Wilson, R. (1982). Rational cooperation in the finitely repeated prisoners' dilemma. Journal of Economic Theory, 27(2), 245–252.
Nobel Prize Outreach. The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel 2001. NobelPrize.org.
Nobel Prize Outreach. The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel 2005. NobelPrize.org.
Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. American Economic Review, 71(3), 393–410.