March 2010

Moderator: Economists have tracked the labor market closely since the government began compiling data on it, so when the U.S. Bureau of Labor Statistics (BLS) introduced the Job Opening and Labor Turnover Survey, commonly called JOLTS, economists had a new tool for viewing employment trends: They could see a snapshot of workers moving into and out of jobs as well as jobs that go unfilled.

Julie Hotchkiss, a research economist and policy adviser at the Atlanta Fed, and Menbere Shiferaw, an analyst in the Atlanta Fed's research department, have written an article about the survey for EconSouth, and they're here to discuss the region's changing employment landscape based on the JOLTS survey and how those changes can be viewed.

Julie and Menbere, thanks for joining us today.

Menbere Shiferaw: Thanks for having us.

Julie Hotchkiss: Thanks.

Moderator: Julie, let me start by asking, how does JOLTS's perspective on the job market differ from other data reports?

Photo of Julie Hotchkiss
Julie Hotchkiss
Photo of Menbere Shiferaw
Menbere Shiferaw

Hotchkiss: Well, the survey conducted by the BLS that most people are familiar with, or at least familiar with the report, is the employer survey; this is also referred to as the establishment survey. Now, the purpose of that survey is to primarily find out how many jobs there are in the economy. This is sort of a stock measure of employment, or rather, kind of a static picture of the labor market at any given point in time.

Now, the JOLTS is a survey that was designed to get a closer look at the dynamics of the labor market. The number of jobs in the economy at any given point in time, of course, is the result of workers being hired by firms and separating from firms.

And the survey asks firms about the number of people that they've hired, the number of people that have separated from their payrolls during the survey month, as well as the number of jobs that go unfilled at the end of every month. Now, job openings, or vacancies, are a useful indicator of the tightness of the labor market. A tighter labor market is one that is typically characterized by plenty of vacancies and low unemployment.

Moderator: How did the need for this type of survey become apparent to those who study the economy? In other words, what need did it fill?

Hotchkiss: Well, information on the turnover of workers, or the flows of workers in and out of firms, have been useful for private industry to gauge their own performance in maintaining a stable workforce. For example, a firm with excessive turnover is typically incurring a great deal of start-up costs with each new hire, so a firm with low turnover is likely able to operate at lower cost.

So economists at the Fed, and other entities concerned with overall economic performance, make use of the information to gauge the strength of the labor market and, in some cases, as an indicator of things to come. For example, if economists observe a significant decline in hiring rates, this suggests that the rise in the unemployment rate is not far behind. Firms' hiring rates are sensitive to changes in consumer demand, of course, and as the demand declines, firms will hire fewer workers. They hire fewer workers in order to replace those that might have left, or just to not expand production because of the lower demand.

One of the uses to which the vacancy data from the JOLTS can be put is the construction of what's called the Beveridge curve, and we describe in some detail what the Beveridge curve is in the EconSouth article. But basically, this curve plots the number of unemployed against the number of job vacancies, and its movement tells us something about both cyclical changes in the labor market, as well as structural changes. The JOLTS provides easy access to one of the important pieces of information that's needed to construct this useful tool.

Moderator: Julie, can you describe how economists assessed the labor market prior to JOLTS's debut in 2001?

Hotchkiss: Well, the most recent survey of labor turnover conducted by the Bureau of Labor Statistics before the JOLTS was actually discontinued all the way back in 1981. Now, while this was the only source of consistent information on turnover, the survey was quite limited in its scope. The survey only covered turnover of workers in the manufacturing, mining, and telephone communications sectors. Now, in spite of this limited coverage, it was referenced often as an indicator of labor market activity. I'll have to tell you, there was about as much excitement as you can hope to see among economists when the BLS started releasing JOLTS data in 2002, after a nearly 20-year hiatus in collecting turnover data.

Now, regarding measures of job vacancies, for roughly 50 years the Conference Board published the Help Wanted Advertising Index, which produced information on job openings found in 51 major newspapers across the country. This index was discontinued in 2008 as job searches went from newspapers to the Internet. The new digital age in job search was acknowledged finally in 2004 with the introduction of the Monster Employment Index. This index was designed to provide a snapshot of nationwide employer online recruitment activity occurring on more than 1,500 Web sites.

The advantage of the JOLTS, however, is that it captures all job openings, regardless of the recruitment mechanism being used by the employer.

Moderator: Menbere, I'd like to ask you, can you describe how the survey is designed in terms of its scope and sample size, etc.?

Shiferaw: Sure. So, the BLS conducts the JOLTS survey each month, and it covers about 16,000 business establishments across the U.S. The data collection center is actually located right here in Atlanta, and it covers most industries. The JOLTS survey covers the private nonagricultural sector and the government sector, but it excludes forestry, fishing, and hunting. It also provides a pretty useful breakdown of information by geographical region—and by that I mean the four Census regions: the South, the Northeast, the West, and the Midwest—and also by industry. However, the turnover data is not broken apart by occupation or establishment size, which right now in the downturn we're in probably would be of particular interest to job seekers and economists and analysts.

Moderator: How is the South's labor market different from other regions of the United States?

Shiferaw: So, the South is pretty representative of the rest of the U.S., regardless of how you define it, whether it's by the Census definition or the Federal Reserve Sixth District definition. As it's defined by the Census, which is what the JOLTS uses, the South is the largest labor market in the country, and thus its characteristics often define those of the U.S. labor market. But not only by size, the diversity of the South also contributes to its generalizibility. Pretty much everything from agriculture to manufacturing to tourism, the South pretty much has it all. As we saw in the vacancy and separation rates in the EconSouth article, it would be pretty hard to say that the South exhibited any sort of employment dynamics during this most recent recession that were unique to the region.

Moderator: Menbere, the EconSouth article mentions the Beveridge curve—a tool that is useful in viewing the balance between labor demand and supply. What does the curve tell us about Southeastern labor markets? Is the regional market considerably different from what you would see in other regions of the United States?

Shiferaw: So the Beveridge curve, like Julie mentioned earlier, charts the relationship between unemployment and job vacancies. The Beveridge curve indicates that the South saw an increase in inefficiencies between 2003 and 2007, as the region was growing so fast that vacancies were increasing faster than unemployment was decreasing, or in other words, jobs couldn't find people. But during the most recent downturn between 2007 and 2009, the curve moved from the Northwest to the Southeast as the economy contracted, and the number of unemployed people grew and vacancies declined. So, in other words, people couldn't find jobs, and that's where we are today—people can't find jobs. So, the curve remains kind of in the southeast quadrant.

The Beveridge curve for the U.S. looks pretty similar to that of the South, although until just recently the mismatch between job vacancies and unemployed was greater in the South than the rest of the nation. The mismatch in the South, or any other region for that matter, is really not the issue of concern any longer. What's more important is that labor markets are creating vacancies for people to get hired into.

Moderator: Thank you for being with us today, Julie and Menbere.

Hotchkiss: You're welcome.

Shiferaw: Thanks for having us.

Moderator: Again, we've been speaking today with Julie Hotchkiss and Menbere Shiferaw of the Atlanta Fed's research department. This concludes our EconSouth Now podcast on the JOLTS survey of employment. For more information, please see the first quarter 2010 edition of EconSouth magazine. On our Web site,, you can read the full article about this topic or subscribe to EconSouth in print.

Thanks for listening, and please return for more podcasts. If you have comments, please e-mail us at