Tuesday, November 9, 2010

Structural Unemployment and the Beveridge Curve

One of the more interesting aspects of unemployment is how policy-makers choose to address it. But what many don't know is that the remedy needs to match the unemployment. That means you can't effectively address structural unemployment with programs mean to address cyclical unemployment.

This brings us to an interesting article from the Federal Reserve Bank of San Francisco. While it introduces a topic you probably don't cover, even in AP Macroeconomics - the Beveridge Curve - it provides a lot of information to help you through the section on unemployment.


Anonymous said...

If you know what Beveridge Curves are and you understood the article, congratulations, you are an academic wonk.

More importantly, If you believe there is any merit in the conclusion, you are in trouble!

Common sense suggests that if you are going to model data and make predictions based on previous events, current conditions need to be the same as, or at least similar to prior conditions.

Every one of those curves represents a typical economic slowdown, not a consumer-led credit and housing implosion the likes of which has not been seen since the great depression.

Thus, the San Francisco Fed's conclusion is as likely to be correct as concluding a robust housing rebound is likely to start any time based on analysis of housing charts from prior recessions.

Well a housing rebound is not likely, and if it's not, neither is a hiring spree, nor a rebound in commercial real estate nor any other meaningful increase in jobs.

In short, the study is invalid because conditions have changed. No one has modeled a credit crunch or a housing bust because we have not seen one since the Great Depression.

The only reason unemployment is not above 10% right now is the participation rate has been plunging like a rock (people have been dropping out of the workforce).

Heaven help us if the economy slows further, even if it does not technically double-dip.

The bottom line is simple: There is still no driver for jobs, there is little likelihood of a driver for jobs, and reliance on academic wonks for economic forecasts is fraught with error.

Tim Schilling said...


You are correct in every respect. However, the point of my bringing the article the attention of my readers is that there is a difference in types of unemployment. And the different types of unemployment need different approaches to solve them. Cyclical unemployment is different from frictional unemployment and structural unemployment. A failure to recognize that fact can result in faulty policy recommendations?

Thanks for commenting.

Christina said...

Does this Beveridge Curve measure all types of unemployment or just cyclical? And if it is just cyclical, what are the other ways we can lower unemployment overall? We learned in my high school economics class that if there is inflation the unemployment rate goes down because inflation and the employment rate are inverses of each other. Could that be a possible solution?

Tim Schilling said...


The curve doesn't differentiate. It just examines the various levels of unemployment and the ability of the economy to match job openings.

In this article, the SF Fed economists don't find anything that would indicate that current unemployment is largely structural as opposed to cyclical.

But that raises the question you asked. What is to be done? The trade-off between inflation and unemployment you mentioned is the Phillips curve. But many doubt that it's a relationship that can be taken advantage of. In the long run, the trade-off doesn't seem to hold. And there are periods where it just doesn't seem to hold up at all.

The best remedy may be to remove uncertainty about the future by adopting consistent policies that encourage hiring, investment and growth.