At Vox,
Paul DeGrawe describes contemporary macro models as "top down" because agents are assumed to know the structure of the world they live in. He writes:
There is a general perception today that the financial crisis came about as a result of inefficiencies in the financial markets and economic actors’ poor understanding of the nature of risks. Yet mainstream macroeconomic models, as exemplified by the dynamic stochastic general equilibrium (DSGE) models, are populated by agents who are maximising their utilities in an intertemporal framework using all available information including the structure of the model – see Smets and Wouters (2003), Woodford (2003), Christiano et al. (2005), and Adjemian, et al. (2007), for example. In other words, agents in these models have incredible cognitive abilities. They are able to understand the complexities of the world, and they can figure out the probability distributions of all the shocks that can hit the economy. These are extraordinary assumptions that leave the outside world perplexed about what macroeconomists have been doing during the last decades.
In its place, he advocates the modeling the economy as a "bottom up" system:
Bottom-up systems are very different in nature. These are systems in which no individual understands the whole picture. Each individual understands only a very small part of the whole. These systems function as a result of the application of simple rules by the individuals populating the system.
While I do think this is worth exploring (his paper will be on my holiday break reading list), his description of contemporary macroeconomics is not quite fair. The models do not imply that macroeconomists believe people really have all this knowledge (
we most of us know better than
that). Rather, we believe it makes sense to model them
as if they do. Milton Friedman famously made this point by noting that it makes sense to model the shots of a champion billiard player
as if he has a sophisticated understanding of physics.
It is worth recalling that macroeconomics adopted the paradigm of rational expectations and dynamic optimization because simple and apparently realistic assumptions like adaptive expectations led to situations where people could systematically and repeatedly be fooled, and this tendency could be exploited by policymakers. While rational expectations may be unrealistic, in a world where people are learning and updating their beliefs, one would expect tendencies that lead to significantly suboptimal outcomes to be driven out (and in the context of markets, with the help of competition). So, while real people are groping around in the fog trying to figure out a rough idea of how the world works, we would expect them to arrive at behaviors consistent with those of rational forward looking agents just as a billiard player, by trial-and-error, learns to make shots consistent with the laws of physics.
So the "unrealism" of the assumptions is not, by itself, a problem. The trouble arises if that unrealism leads us to incorrect predictions. Macroeconomics may indeed have gone astray in some ways, and the lack of realism in models may indeed vex non-economists, but the assumption of "incredible cognitive abilities" is not
inherently a problem.
Update (11/22): I had a chance to a look at
DeGrawe's paper (pdf). Although I disagree with his characterization of modern macroeconomics, I do think his paper is interesting, though I suspect it may be vulnerable to some of the same criticisms as adaptive expectations models. (via
Mark Thoma's link to the "Whats Wrong with Modern Macroeconomics" conference papers).