The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« You Say You’re a Homeowner and Not a Renter? Think Again. | Main | The Same-Old, Same-Old Labor Market »

March 19, 2013

Being Ahead of the Curve: Not Always a Good Thing

Our friends at the New York Fed have a nifty interactive graphic that compares the unemployment rate, labor force participation rate, and employment-to-population ratio over the last five business cycles. You can even break these indicators down by gender, by age, or by a particular business cycle. (For a deeper dive, check out this post at Liberty Street Economics by Jonathan McCarthy and Simon Potter.) And though it’s not exactly late-breaking news, no matter which of the three indicators you look at, you can’t help but conclude that the most recent recession is an outlier.

The Beveridge curve is a fourth and particularly useful graphical representation of a steady-state economy showing how, in theory, one might expect the vacancy rate to change, given an unemployment rate. It depicts the relationship between job openings and the unemployment rate. (The Atlanta Fed’s magazine, EconSouth, discussed the Beveridge curve.) It, too, has been standing out over the course of the most recent recovery, so much so that we think it warrants at least a second glance. There are a number of ways to estimate a Beveridge curve (see, for example, methods described by Gadi Barlevy of the Chicago Fed here and by the Richmond Fed’s Thomas Lubik here).

We use the method described by Barnichon et al. (2012) to estimate the solid curve used in the first chart below. The square plots represent actual vacancy rate (y-axis) and unemployment rate (x-axis) combinations by month from December 2000, when the Job Openings and Labor Turnover Statistics (JOLTS) data series from the U.S. Bureau of Labor Statistics (BLS) series begins, to January 2013, the most recent month of data available for both series.

Blue squares represent data from December 2000 to December 2009, when the “errors” between actual plots and the curve estimation were below 2 percentage points, and red squares represent data since January 2010, where data suddenly seem to jump higher than the predicted Beveridge curve to the tune of 2 percentage points or greater (see the chart below).

But why?

In June 2012, Regis Barnichon and his coauthors concluded that the unemployment rate’s lackluster performance so far in the recovery was attributable to a shortfall in hires per vacancy. Since then, the vacancy rate has climbed its way back to its June 2008 level of 2.7 percent. However, the unemployment rate has clearly not returned to either its June 2008 level (5.6 percent) or where the Beveridge curve says it should be given this vacancy rate, which one might predict to be 5.5 percent using the methodology of Barnichon et al.

This “ahead of the curve” phenomenon has not gone unnoticed and has prompted some explanations. In a March 6, 2013, article in The New York Times (which also has some cool charts), Catherine Rampell posits that available positions are staying unfilled longer, while interview processes have become lengthier.

The next day, Rampell went into more detail about why we’re going “off the curve” in a New York Times Economix post. She cites skills mismatch and a skills atrophy effect of the long-term unemployed affecting the ability of employers to fill positions (which she explains aren’t full explanations, yet we would expect to see wages for highly coveted positions rise significantly).

Rampell goes back to the explanation many of us continue to hear from business contacts: employers are unwilling to fill vacant positions because of economic and fiscal policy uncertainty. She quotes Stephen Davis of Chicago’s Booth School: “They’re taking longer to fill vacancies because they just feel less need to fill jobs now,” Davis said. “They recognize that in a slack labor market, there is an abundance of viable candidates. If something happens, and if they need to hire quickly, they know they can do that. That’s harder in a tight labor market.”

So maybe as labor markets “tighten up,” or perhaps if the speed by which they tighten up quickens, we’ll get back on the Beveridge curve. Only time, and several BLS releases, will tell.

Photo of Patrick HigginsBy Patrick Higgins, an economist at the Atlanta Fed, and

Photo of Mark CarterMark Carter, a senior economic analyst at the Atlanta Fed

March 19, 2013 in Economics , Employment , Labor Markets | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Being Ahead of the Curve: Not Always a Good Thing :


This post de-emphasizes Rampell's point that vacancies remain open longer because of economic uncertainty and growth expectations. Instead, it stresses the effect of the labor market supply.

Firms' hiring decisions likely are impacted by the number of qualified candidates available. However, most firms probably also hire because of the amount of work on hand, and the amount of work anticipated. The latter point is obscured by the quote by Stephen Davis, which is unfortunate, since it is probably quite important. In fact, macroblog referenced this relationship in a previous post, but unfortunately, decided to ignore it here.

Posted by: Carl | March 21, 2013 at 10:49 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad