Managing the expectations trade-off: Why flexible agreements often prove counter-productive

» Lire en français: French

A common criticism of economics is that economic theory often fails to predict real world events, spectacularly so in some cases. More recently, however, a more empirical approach to economics has emerged that combines economic theory, psychology, and laboratory experimentation, in order to better understand decision making in real life situations.

7 min read

Christian Zehnder, professor of organizational decision making at HEC Lausanne is part of this new tradition. He spoke to Steve Coomber about his work on contracts conducted alongside Nobel Prize winning economist Oliver Hart and celebrated behavioral economist and neuroeconomist Ernst Fehr. It is work that provides important insights into how we should frame the many agreements that govern working life and organizational behavior.

Christian Zehnder is a professor of Organizational Behavior. His work focuses on behavioral economics, a line of research that combines insights from economics and psychology.

Before we move onto your contracts work – your research involves laboratory experimentation, and this is something relatively new in economics.

Experimental economics began to produce a lot of results that were at odds with the predictions of the standard economic mode

For a long time people thought experiments weren’t possible in economics. If you wanted to understand economic behavior you had to do it in the “real world”, with natural decision making in natural environments, which was difficult. But in the 1980s a new approach emerged; putting people into laboratory settings, where we can completely control the environment and where the theory makes a clear and testable prediction. Theories were tested in the lab, but also developed and adapted from the observed deviations from the expected behavior. Experimental economics began to produce a lot of results that were at odds with the predictions of the standard economic model which assumes that people are completely rational and selfish.

Contracts play a very important role in economic thinking. Have these results, showing that people are not necessarily completely rational, also affected the way that economists think about contracts?

The implicit assumption in traditional microeconomic models is that contracts are complete. For every possible situation that can arise there’s something in the contract that describes what people are supposed to do. However, this is not what we see in reality. As soon as the situation becomes complex, contracts are typically very incomplete. Just think about the typical employment contract. The contract provides a framework, but the details need to be determined on the spot. The reason is that people cannot perfectly foresee the future and it would be immensely costly to write down how all the possible events that one can imagine should be dealt with. The insight that contracts are typically incomplete has triggered a whole new strand of literature in economics.

How did you meet Oliver Hart and why did you start collaborating with each other?

Ernst Fehr and I convinced Oliver Hart that it might be a good idea to put his theory to the test and check whether it is a good description of how people really behave

This was essentially a chance encounter. I believe it was in 2006 when Oliver came to Zurich and gave a talk on a new research project of his. I was a postdoc at the University of Zurich at this time and I was fascinated by his presentation. During the talk he suggested that economists should start taking into account the possibility that contracts may not only define rights and obligations, but may also have important psychological effects. I was quite surprised by these statements, because I had always thought of Oliver Hart as a rather traditional economist. I was eager to talk to him after the presentation and so we ended up having dinner together with Ernst Fehr. It was during this dinner that Ernst and I convinced Oliver that it might be a good idea to put his theory to the test and check whether it is a good description of how people really behave.

So how should we think about contracts in the light of your work? What are these psychological effects that you mentioned?

The basic idea is that when you contract about something you’re going to do in the future, the nature of that contract, the way it’s written, determines what you expect to get – it creates a sense of entitlement. As the contract proceeds, events may occur that are unanticipated and people negotiate until they reach an outcome. If that actual outcome is different to the expected outcome, especially if someone gets less than they expected, the assumption is that people feel aggrieved and are unhappy about the outcome. As a response they have a desire to retaliate and take it out on their trading partner. This leads to counterproductive behavior, conflicts and inefficiency.

You said that the form of the contract is important. How does a contract need to be written to reduce these problems?

The problem with flexibility is that each trading party may hope for a completely different outcome

According to the theory and our findings from the experiment, the central issue is the flexibility of the contract. From a standard economic point of view, flexible contracts are attractive, because they allow for a lot of possible outcomes. This may be very helpful in situations where you don’t know much about the future. Whatever happens, there’s enough flexibility in the contract to respond to events and adjust the terms in a way that improves the outcome. However, the problem with flexibility is that each trading party may hope for a completely different outcome. If people have a tendency to be a bit self-serving, everybody will hope for an outcome which is particularly good for themselves. In most cases, this implies that not everyone’s expectations can be met so that at least someone is always going to be unhappy with the outcome.

One possible way of avoiding this problem is to write a very rigid contract

One possible way of avoiding this problem is to write a very rigid contract, pinning down as many aspects as possible from the outset. The advantage of such a contract is that everybody knows exactly what to expect and expectations are therefore likely to be aligned. This will reduce conflicts and counterproductive behaviors, because nobody is going to be particularly unhappy about the outcome. They got what they expected. However, the downside of such a contract is that if unanticipated events occur there is no flexibility to modify the contract and the parties risk ending up with a sub-optimal outcome that could have been avoided in a more flexible arrangement.

So the crucial thing here seems to be the trade-off between flexible and rigid contracts, in terms of less or more retaliation and associated costs, and less or more flexibility. And in the lab you tested to see if this trade-off exists as predicted?

We constructed an experiment with buyers and sellers, who could choose a flexible contract or rigid contract. We observe that rigid contract had problems in some situations because the initial contact terms did not fit the evolving situation, but at the same time we also observed that rigid contracts created much less counterproductive behaviors than the more flexible contracts.

After publication of your first paper, there were a couple of points in particular that were picked up on, and which you addressed in a more recent paper.

Managing expectations through communication helps a bit, but not very much

One thing people said in response to the first paper’s findings was that having a rigid contract wasn’t the only way to avoid counterproductive behavior arising from unmet expectations. In reality, expectations could be managed between trading partners through communication. When we tested this we found that managing expectations through communication helps a bit, but not very much. However, in the meantime other people have worked on this question as well and found more positive effects from communication. We are now working on a new project that will hopefully help us to better understand what exactly is the role of communication in such contracting situations.

Another point raised was that we didn’t allow for renegotiation in the case of the rigid contract, whereas in reality you can change the agreement. So rigid contracts don’t really exist in that sense and in real life would lead to more counterproductive behaviors. But even when we allowed for renegotiation, rigid contracts still helped.

In everyday working life, organizations and individuals adopt policies and practices that are governed to a lesser or greater extent by rules. Effectively these are agreements (many unwritten) between parties, which while not having the force of law, are similar to contracts in intent. And these agreements range on a continuum from very flexible to very rigid. Is the reference point theory, the trade-offs etc. equally applicable to these types of agreement. Are we all making and encountering the flexibility / rigidity trade-off constantly in organizational life?

Yes, I think you can fully apply the theory to agreements as well. In fact, in Oliver Hart and John Moore’s original paper on this topic, they refer to the contracts as “agreements to agree”, because they look at a situation in which the trading parties can always decide to walk away if they do not want to trade. The assumption of voluntary trade is not essential, but it shows that the theory also directly refers to the type of agreements that you mention.

Given your findings, what would be the main messages for practitioners?

We often see quite rigid contracts even in quite uncertain environments

For a start, our findings go against the prediction of standard economic theory, which says maximum flexibility is typically best in situations with a lot of uncertainty about the future. However, in reality we often see quite rigid contracts even in quite uncertain environments. Our findings help to understand why this is the case.

Sometimes it may make sense to have very rigid rules – Flexibility may lead to even bigger problems

Sometimes it may make sense to have very rigid rules, even if these rules sometimes lead to bad outcomes that could have been avoided with more flexibility. The reason is that flexibility may lead to even bigger problems, especially if there is the potential for very costly counterproductive behavior.

Take a simple, real life example. In a department with seven or eight professors and some 35 assistants I implemented a rule about assigning speakers to seminars that is, on the face of it, inefficient. Every professor can pick one speaker per semester. This might not seem to make much sense on examination. There may be semesters where someone might want two speakers, whereas another person might not need any – so more flexibility would allow more efficient seminar speaker arrangements.

However, the rule is there for a reason; with a lot of flexibility, comes a lot of conflict, endless discussions that are very costly, people with unmet expectations. So instead we decided to live with an apparently inefficient rigid agreement, in order to avoid all the long discussions we had before.

So think about the trade-off. If you have an agreement that causes (or potentially causes) a lot of conflict – consider what you would lose if you were to make the agreement more rigid and lay down some simple rules to avoid all the conflicts. If the loss from lack of flexibility is minimal, you should change.