Monday, September 5, 2011

Two examples of gaps in Macroeconomics

Much has been said about the failure of Macroeconomics (and macroeconomists) to explain, predict or find solutions for the last financial crisis and its aftermath. The topic came back to my mind after reading two recent articles about current macroeconomic events, articles that capture well some of the gaps of the academic economics literature. 

The first article is by Robert Shiller in The New York Times. Robert Shiller is well known for his work around excesses in financial markets. He has produced many interesting pieces of academic work on this subject but he is better known to the public for his book Irrational Exuberance about the asset price bubble during the 90s as well as his warnings on the real estate bubble prior to the collapse in 2007. His article discusses recent volatility in the stock market in terms of the "beauty contest" analogy suggested by Keynes. His view of markets is one of volatility and behavior which is more about predicting what others do than understanding the fundamental value of assets. The article presents a view of stock markets which might not be far from what many (investors, those involved in financial markets, the general public) believe about those markets. But it is not the view that most macroeconomic or financial models have of how investors behave. Yes, we have models that try to capture some of the anomalies and volatile behavior described by Robert Shiller, but these models have very limited influence when it comes to macroeconomic analysis or advise for policy makers. Erratic and volatile behavior in financial markets and asset prices can be a source of volatility for the economy, lead to unsustainable behavior (bubbles) and make the recovery from those events much harder because of how pessimism spreads through trade and interactions among investors. Ignoring this behavior makes it impossible to talk about the most recent business cycles.

The second article is from Brad DeLong and it talks about what to do with the US labor market. This is another area where macroeconomic models struggle to deal with current events, in particular the very negative labor market that the US economy has had for the last three years. And in this case it is not because the labor market was not part of macroeconomic models. All macroeconomic models include a labor market and they discuss the issue of unemployment. But their discussions are not too helpful in understanding the current US environment. Most macroeconomic models have a view of labor markets clearing around a certain unemployment rate (structural unemployment, natural rate of unemployment,...). Some models only generate movements in unemployment when the equilibrium level changes - but it is hard to argue that this is what we have witnessed since 2007. Other talk about cyclical unemployment caused by frictions. But there is no agreement on the origin of these frictions. In the standard New-Keynesian model, frictions come mostly from price rigidity. These frictions are temporary and as time passes unemployment returns to its natural value. But as Brad DeLong correctly argues:

"Excess supply on the labor market not is registering at all, is not leading to any upward pressure on the rate of matching workers who want jobs with jobs that want workers. We have an awful lot of people who are answering the CPS survey by saying; "No, I don’t have a job, and yes, I have done something to look for one over the past four weeks". Your most basic economic matching model says: multiply the fraction of workers looking for jobs by the fraction of jobs looking for workers and that will b proportional to your job-finding rate. It isn't. The equilibrium-restoring forces in the labor market at the macro level appear to be much much weaker than I thought they were in 2007, appear much weaker than I would have believed back in 2007 that they possibly could be."

As a final point, and to be fair to the academic literature, there are plenty of models that incorporate volatile or irrational behavior in financial markets and there are many models that include a variety of frictions in the labor market that could account for some of what we see today. But these features are not at the core of the standard analysis of the current macroeconomic environment and they are not articulated in a way that allows for concrete and specific policy advise. And those who dare using some of these features in their analysis and policy recommendations seem to do it mostly when talking to "outsiders" and less so in conversations with other academics. 

Antonio Fatás