Expert knowledge and policy reform: Stephan von Cramon-Taubadel's rejoinder to Jean-Marc Boussard
Stephan von Cramon-Taubadel, Professor at the University of Göttingen, says that models are like maps: they are never entirely realistic but help us find the way. In the case of the post-2013 CAP, the reform path does not even depend on the traditional models because the main issue is not greater market orientation.
Jean-Marc Boussard has posted an insightful and stimulating comment on Valentin Zahrnt’s „Guide to the CAP reform policies”. Very few agricultural economics can draw on the breadth of historical overview and wealth of experience that Jean-Marc brings to bear on agricultural policy issues, and his comment is a trenchant reminder that there is a great deal that we still do not know about how people and policies interact to influence market outcomes.
While I sympathise with many of the arguments that Jean-Marc makes, I would like to demur to several points. The first of these concerns the ‘reality’ of economics models and the assumptions that go into them. Jean-Marc argues that general equilibrium (GE) theory is an inappropriate basis for policy recommendations in agriculture (and presumably any other sector) because it is based on numerous questionable and unrealistic assumptions. There is a tradition of debate among economists on how realistic models must be; back in the 1950s two titans and Nobel Laureates, Milton Friedman and Paul Samuelson locked horns on this issue. As an agricultural economist who is in no great danger of winning a Nobel Prize, there is little that I can contribute to this debate. But one point bears repeating: models are by definition never entirely realistic; they are either useful or not useful. A map is a model that grossly abstracts from reality – everything is much too small and flat, the rivers that maps depict are not wet and their streets are never congested. Nevertheless, maps are very useful models precisely because they assume away most aspects of reality and focus on those aspects that travellers need, such as location, distance and connection.
This is not to say that GE theory is above reproach. For example, the fact – alluded to by Jean-Marc – that much human need for food on this planet cannot express itself as economic demand, is indeed not “satisfying”. The standard GE model fails to account for the distinction between basic human needs and demand, and it therefore fails to adequately capture the biggest single global food policy issue. Nevertheless, some readers may interpret Jean-Marc’s comments as stating that any model that is not entirely realistic (i.e. all models, until we get around to duplicating our world down to the last detail) is therefore never useful. This would amount to throwing out the baby with the bath water.
Second, Jean-Marc, while calling for more “modesty” on the part of agricultural economists, doesn’t pull punches when it comes to the putative record of liberal economic policies and the apparent successes of “dirigisme”. We read that “the validity of the liberal approach to agricultural problems has never been confirmed by experience”, and that the fact that global famine did not occur in the 1990s is “largely a consequence” of policies that “disconnect agriculture from the market”. As someone who has spent a considerable amount of time in the countries of the former Soviet Union, where policies did indeed completely disconnect agriculture from the market for many decades, with disastrous consequences for production, the environment and human welfare (including a man-made famine of apocalyptic proportions), I would be leery of such blanket statements. To be sure; rough-hewn, one-size-fits-all attempts to liberalise agriculture in these countries since the early 1990s have not produced stellar results and warn us that markets can only function when key public and private institutions are in place. This warning notwithstanding, over time and across countries it is difficult to deny a correlation between broadly democratic institutions, predominantly liberal economic policies and improvements in human welfare. Jean-Marc is sceptical of the ability of liberal economic recommendations to lead to improvements in welfare because the real world is much more complex than our models, and involves many processes and much information that these models do not and cannot capture. Fair enough. But the same concerns about complexity, information deficits and “multiple and inefficient equilibria” constrain attempts to design and implement dirigistic policies as well, and are compounded by political economic concerns about policy makers’ underlying incentives and goals.
Finally, even if we swallow Jean-Marc’s arguments about liberal economics and GE models hook, line, and sinker, I submit that they largely miss the point of the ongoing debate on the CAP after 2013. This debate is not about liberalisation. It is at best only peripherally about reducing state interference on agricultural commodity markets. This reduction has already occurred to a great extent in the EU since 1993 in the course of reforms initiated by Agricultural Commissioners MacSharry, Fischler and Fischer-Boel. To what extent these reforms have been successful is an interesting and important question, but it is not the issue currently at stake. Even if Jean-Marc is right and GE theory is woefully inadequate, is the current CAP (or any of its earlier incarnations for that matter) in any way a response to this inadequacy? Exactly what shortcomings of GE theory justify providing farmers today, in the absence of any means-testing whatsoever, with income transfers that are largely based on the amount of agricultural products they (or their ancestors) produced back in 1993? In what sense do payments of over 600 Euro/hectare in Greece and under 100 Euro/hectare in Latvia reflect a superior understanding of the economy than that embodied in the GE framework? The current debate is not about liberalising agriculture; it is about designing more effective and efficient forms of intervention than those that have made the CAP a permanent construction site since its inception. The case that current CAP can be improved upon hinges neither on the validity of GE theory, nor on the question whether GE theory can help to design improved policies.