December 17, 2008
Presenter: Ryan Booth
Respondent: Robb Rutledge
On Political Economy
This session centered around two questions: First, what does an applied economic theorist do? And second, what is the relevance of economic theory? The discussion of the first point began with an examination of the theoretical model in a classic paper from the political economy literature, "Protection for Sale" by Gene Grossman and Elhanan Helpman. The talk then segued into my own research, which examines the political economy behind public finance decisions within cities. At the conclusion of the evening, we then discussed Ariel Rubinstein's "Dilemmas of an Economic Theorist," in which the author proclaims his own views concerning several alleged shortcomings of economic theory.
Ryan gave an interesting talk describing a model for special-interest group contributions developed by Grossman & Helpman in their article “Protection for Sale”. He also discussed his own work extending the work of Grossman & Helpman to consider circumstances that are more realistic. Ryan also discussed an article by Rubinstein entitled “Dilemmas of an Economic Theorist” that touched on questions about the relevance of economic models to the real world.
Economists have long argued in favor of free trade, that total welfare is reduced by protectionism (tariffs). However, special-interest groups want tariff protection for their industries because then they can charge higher prices. For example, cars are produced domestically and imported from abroad. The car industry lobbies politicians to raise tariffs on imported cars. With tariffs, imported cars cost more so domestic car companies can raise their prices and increase profits. Because prices are higher, consumers buy fewer cars. The owners of domestic car companies benefit from the tariff but society is hurt by reduced access to cheap cars, and economists have shown that this results in net social loss. The Grossman & Helpman model assumes that politicians set domestic prices by maximizing their own welfare, which depends on both the total bribes received and the country’s welfare, adjusted by the weight they give to social welfare. The model implies that only industries with lobbies receive protection, the larger the industry, the larger the tariff, and the more the politician cares about social welfare, the lower the tariffs.
Ryan’s own work extends this model to one in which the politician and the lobby play a repeated game where the lobby’s capacity to bribe depends to the current policy and union size, and the politician is subject to reelection. Ryan’s model accounts for more of the variables that our intuition suggests are likely to be releveant for lobby and politician behavior and policy decisions. To what extent are these models useful for describing real world phenomena, or for influencing policy decisions? Rubinstein asks the question in his article: Has the world learned something useful?
This question is related to one we often ask in the forum concerned with a field’s identity. What is economics? What is the goal of the field of economics? This sort of question is as important for an old field like economics as it is to new interdisciplinary fields. Should economics be relevant to reality and should it influence policy? Rubinstein seems to believe that economics tells useful fables that can influence culture, like film does (“Yes, I do think we are simply the tellers of fables, but is that not wonderful?”). To me, it seems a shame if economics is just fables. Our government must make big economic policy decisions and it would be a shame if economic models could tell them nothing about how those policy decisions might affect future social welfare.
When are economic models useful? Grossman & Helpman’s model made predictions about how the sensitivity of imports to prices influences the size of tariffs. Real world data were shown to be consistent with these predictions. Ryan’s model also makes predictions about how policy persistence depends on how forward looking or myopic voters are. Clearly there are situations where economic models and economic reality align, but there are also situations where they do not (like the current economic crisis). Economic models often seem unable to describe rare and often catastrophic phenomena. This is unfortunate because these are just the phenomena we would most like to understand and avoid. This failing is not altogether surprising or uncommon to researchers in other areas. Catastrophic natural phenomena (earthquakes, volcanic eruptions) are also difficult to predict. Like in economics, scientific models break down in certain situations. For example, Newtonian mechanics describes the motion of objects at low velocities. Near the speed of light, these laws are no longer accurate. Newtonian physics is still taught in school, as it should be, although it they are “wrong”. I get the feeling that some economists would discard the models when encountering conflicting evidence. For a wide range of circumstances, Newton’s equations accurately describe natural phenomena. Furthermore, they are far simpler to understand and apply than their relativistic counterparts. What is important is not that we find the one model for everything (although that would be nice), but that we understand the limitations of the different models that we have available. In the same way, rather than rejecting them, economists might attempt to identify the conditions over which their models usefully describe behavior. Where the model systematically deviates from reality, an entirely different sort of model may be necessary. It is important to try to develop models for those rare cases, rather than ignoring them because our models stumble there.
I return to the question of how a field should define itself. For economics this depends on how much it would like to explain actual market behavior. It would be more amusing if it wasn’t so unfortunate that economic models best explain the behavior of economists. The behavior of the people that might actually influence and make policy (for example, business and law students) systematically deviates from the predictions of economic models. If it wants to be more relevant to reality, it should maintain a closer relationship between theory and data. Ryan argues for this with his research, in which his models are closely related to political and economic realities. Economists seem to write models, and then look for data to reject or revise those models. Scientists seem to collect data and then look for a vague theory that fits the data. There are few theorists in science, and a difference in theoretical and experimental language limits communication. Data and theory should repeatedly interact to achieve the best fit, ideally in individuals with expertise in both areas, but at the very least with better communication between theorists and experimentalists. The fact that the models so often break down under rare circumstances suggests that perhaps different sorts of models are necessary to explain these phenomena. One possibility is that psychologists and neuroscientists will have insights into how people make decisions under these unusual circumstances. If economics wishes to remain relevant to reality, it might benefit from a more interdisciplinary approach with better communication with all the fields with expertise that might be relevant to understanding economic reality.