Letter from Germany - Excellence and incarceration

In this essay Ray Rees, Professor of Economics at the University of Munich, provides an update on the attempt to introduce quality-related funding to Germany’s universities and tells a cautionary tale about the potential pitfalls of experimental economics.

In previous letters I have written about the strong and growing sense in Germany that the universities here are in serious need of reform. The outgoing Schröder government at last pushed through a programme that could seriously contribute to improving the situation. I know that university teachers in the UK have been going through the stress and exertions of the latest RAE. The Excellence Initiative introduced here in Germany about three years ago aims at a similar goal, of raising the quality of research in and the international reputation of the universities, but has gone about things in a way that is less demanding of university teachers, and, as it has turned out, has had much more dramatic results.

The basic diagnosis underlying the Initiative was that as a result of huge growth in student numbers and serious underfunding of an almost entirely publicly-funded university system, Germany lacks centres of excellence. Simply to pour money into the system as a whole according to the ‘watering can principle’, under which funds are pretty evenly spread around the country, would not be an effective solution. It was decided to carry out a competitive evaluation, in which not only was the quality of research in every department in every university that chose to participate to be evaluated by committees of international experts, but the university administrations had to put up plans showing how they would use the substantially increased funding they would receive if they qualified as ‘elite universities’. Both achieved-quality and the plans played a part in the evaluation. The idea was to produce a group of perhaps as many as ten elite universities, the centres of excellence that would restore Germany’s standing in the academic world.

The results of the exercise, announced last October, caused something of a furore. The Committee making the final decisions was composed of academics, many based outside Germany, and politicians, with the academics having a small majority. An important point to note is that under Germany’s federal system, universities are run by the states (Länder), and any one politician would see himself as concerned with the position of the universities in his state, first, and in Germany as a whole, second. You can imagine then the reactions from these politicians when the academics, voting as a solid bloc, pushed through a list of only three universities that could be considered as excellent and qualifying for elite status and, more importantly, funding. These were my own university, the Ludwig-Maximilians-University, Munich, its sister university, The Technical University, Munich, and the Technical University of Karlsruhe. Traditionally famous universities such as Heidelberg, Göttingen, and the Humboldt University, Berlin, and the large universities in major cities such as Hamburg, Frankfurt, Stuttgart and Cologne, were considered by the academics to be just not good enough, or at least not having good enough plans.

The story is however not yet over. A second round is taking place, and it will be interesting to see the extent to which the affronted politicians will be able to reinstitute the watering can principle. I plan to report on all this at greater length in my next year’s letter.

In the meantime, I thought that readers of this letter may be interested in the latest developments in experimental economics in Germany. This has always been a strong area of interest, and I read recently of an interesting experiment in contract theory. A well-known problem in experimental economics is the question of whether the results obtained, usually with student-subjects in a laboratory setting, really carry over to the real world. Here, a group of experimental economists managed to persuade the government of the state of Holwig-Schlestein (let us say) to let them take over a small, white collar, low security prison and run it according to the principles of modern contract theory.

As is well-known, prisons are very labour intensive – about three-quarters of the costs are labour costs, primarily of guards. This leads to a problem when funding is constrained — there is a strong incentive to reduce costs by employing low-quality personnel as guards. The experimenters thought of a brilliant solution to this problem — a prison with no guards at all. They introduced the self-regulating prison, in which a price mechanism solves the central dilemma of such a system, what could be called the prison-owners’ dilemma: how to ensure that in a prison without guards nobody escapes. The solution was to use a properly formulated contract to construct the appropriate incentives to ensure that a prisoner would prefer to stay in prison rather than to escape. The only way to do this of course is to ensure that the utility level enjoyed by a prisoner is at least as great in prison as it is on the outside, so that it is a dominant strategy not to escape. They referred to this as the Incarceration Constraint (IC).

Their first step was to use the money saved by not having to pay guards, construct elaborate security and surveillance systems and so on, to equip the prison with first class hotel facilities. But there was here of course a hold-up problem. It only pays to invest in this way if they could be sure that the prisoners would not escape. This is where the contract came in. On entering the prison the new inmate signs a contract promising him an annual income at least as high as that he makes by his life of crime outside, but which is payable, after deduction of hotel costs, but with the addition of interest, only at the end of a completed sentence. If he escapes, he loses the entire income, including that owing to him up to the time at which he escapes. Clearly, once the prisoner has begun to serve his sentence, this contract is self-enforcing.

The problem was that the funds available to the experimenters were nowhere near large enough to finance these contracts. They solved this by exploiting another insight. Prisons are an enormous waste of human capital. Prisoners spend much of their time doing nothing, or engaging in tasks, such as breaking stones or sewing mailbags, that yield average revenue well below the average earning power of the white-collar prisoner, who is often a very skilled person. Just think of forgers, computer criminals, financial swindlers, CEO’s of major companies, and so on. The experimenters therefore formed a consulting firm, which hired out the prisoners as consultants to police forces, security companies, banks, insurance companies, or indeed anyone prepared to pay the market price. This was so successful that the revenue was sufficient to meet all claims falling due under the contracts signed with the prisoners. Note, incidentally, that for the first year or so, the business brought in income without any corresponding payments to discharged prisoners, so this provided a buffer stock of capital that more than covered the start-up costs and, in the steady state, provides a permanent reserve.

At the time of writing, the experiment is facing two problems. The first is the penalty of success. There has been a small but significant increase in white collar crime in the catchment area of the prison, with the offenders making no attempt to evade arrest. The proposed solution is to make committing a crime no longer a necessary condition for admission to the jail, so that it will be open to non-offenders, provided they can prove they have the necessary skills.

The second problem is perhaps more fundamental. The prison is of course serving its purpose of removing criminals from society, at least to some extent, but some people are complaining that it is unethical that the prisoners are not being punished, which is of course implicit in the IC constraint. The experimental economists, not unusually, are very uncomfortable with ethical discussion, and are trying to avoid having to confront this issue. I will be happy to pass on to the experimenters (who bear no resemblance to any person living or dead) any proposals readers of this letter may have for solution of that problem.

Page Options