October 15, 2014

What's behind Declining Labor Force Participation? Test Your Hypothesis with Our New Data Tool

The share of people (age 16 and over) participating in the labor market—that is, either working or looking for work—declined significantly during the recession. As many researchers have noted (see our list of supplemental reading under "More Information"), there is clearly a cyclical component to the decline. When labor market opportunities dry up, it influences decisions to pursue activities other than work, such as schooling, taking care of family, or retiring. However, much of the decrease in the overall labor force participation rate (LFPR) could be the result of the continuation of longer-term behavioral and demographic trends.

While the U.S. Bureau of Labor Statistics (BLS) produces aggregate statistics on LFPR by various demographic measures, the published data tables don't detail the reasons people give for not participating in the labor market. However, we have cut and coded the micro monthly data from BLS's Current Population Survey so that you can explore your own questions.

For example: Are millennials less likely to participate in the labor market than earlier cohorts? Are people retiring sooner? Are women less likely to stay at home than in the past? You can answer these and other types of questions on our new Labor Force Participation Dynamics page (click on "Interact and Download Data").

In addition to allowing you to create your own charts and download the chart data, the website also guides you through some of the major factors we found that contributed to the decline in LFPR from 2007 to mid-2014, as well as a picture of the trend in those factors before the recession began.

The chart below (also in the Executive Summary) provides an overview of the major factors that we noted in our analysis of the data. Each bar shows the contribution to the 3 percentage point change in the overall LFPR from 2007 to mid-2014.

What the chart doesn't show is whether the trends were occurring before the Great Recession. For a deeper dive into any of the factors in the chart, see the "Long-Term Behavioral and Demographic Trends" section. The most influential factor has been the changing distribution of the population (see "Aging Population"). The fact that a greater portion of Americans are retirement age now than in 2007 accounts for about 1.7 percentage points of the decline. At the same time, older Americans are more likely to be working than in the past, a trend that has been putting upward pressure on LFPR for some time. All else being equal, if those older than 60 were just as likely to retire as they were in 2007, LFPR would be about 1.0 percentage point lower than it is today.

Other factors bringing down the overall LFPR include an increased incidence of people saying they are unable to work as a result of disability or illness (click on "Health Problems"), increased school attendance among the young (click on "Rising Education"), and decreased participation among individuals 25–54—the age group with the greatest attachment to the labor force (click on "Focus on Prime Working-Age Individuals").

These are the factors we found to be the most significant drivers of changes in LFPR, but you can also explore many other questions with these data. Check out the interactive data tools and read our take on the data and let us know what you think.

By Melinda Pitts, director, Center for Human Capital Studies,

John Robertson, a vice president and senior economist in the Atlanta Fed's research department, and

Ellyn Terry, an economic policy analysis specialist in the Atlanta Fed's research department

October 15, 2014 in Economic conditions, Employment, Labor Markets | Permalink | Comments (3) | TrackBack (0)

September 29, 2014

On Bogs and Dots

Consider this scenario. You travel out of town to meet up with an old friend. Your hotel is walking distance to the appointed meeting place, across a large grassy field with which you are unfamiliar.

With good conditions, the walk is about 30 minutes but, to you, the quality of the terrain is not so certain. Though nobody seems to be able to tell you for sure, you believe that there is a 50-50 chance that the field is a bog, intermittently dotted with somewhat treacherous swampy traps. Though you believe you can reach your destination in about 30 minutes, the better part of wisdom is to go it slow. You accordingly allot double the time for traversing the field to your destination.

During your travels, of course, you will learn something about the nature of the field, and this discovery may alter your calculation about your arrival time. If you discover that you are indeed crossing a bog, you will correspondingly slow your gait and increase the estimated time to the other side. Or you may find that you are in fact on quite solid ground and consequently move up your estimated arrival time. Knowing all of this, you tell your friend to keep his cellphone on, as your final meeting time is going to be data dependent.

Which brings us to the infamous “dots,” ably described by several of our colleagues writing on the New York Fed’s Liberty Street Economics blog:

In January 2012, the FOMC began reporting participants’ FFR [federal funds rate] projections in the Summary of Economic Projections (SEP). Market participants colloquially refer to these projections as “the dots” (see the second chart on page 3 of the September 2014 SEP for an example). In particular, the dispersion of the dots represents disagreement among FOMC [Federal Open Market Committee] members about the future path of the policy rate.

The Liberty Street discussion focuses on why the policy rate paths differ among FOMC participants and across a central tendency of the SEPs and market participants. Quite correctly, in my view, the blog post’s authors draw attention to differences of opinion about the likely course of future economic conditions:

The most apparent reason is that each participant can have a different assessment of economic conditions that might call for different prescriptions for current and future monetary policy.

The Liberty Street post is a good piece, and I endorse every word of it. But there is another type of dispersion in the dots that seems to be the source of some confusion. This question, for example, is from Howard Schneider of Reuters, posed at the press conference held by Chair Yellen following the last FOMC meeting:

So if you would help us, I mean, square the circle a little bit—because having kept the guidance the same, having referred to significant underutilization of labor, having actually pushed GDP projections down a little bit, yet the rate path gets steeper and seems to be consolidating higher—so if it’s data dependent, what accounts for the faster projections on rate increases if the data aren’t moving in that direction?

The Chair’s response emphasized the modest nature of the changes, and how they might reflect modest improvements in certain aspects of the data. That response is certainly correct, but there is another point worth emphasizing: It is completely possible, and completely coherent, for the same individual to submit a “dot” with an earlier (or later) liftoff date of the policy rate, or a steeper (or flatter) path of the rate after liftoff, even though their submitted forecasts for GDP growth, inflation, and the unemployment rate have not changed at all.

This claim goes beyond the mere possibility that GDP, inflation, and unemployment (as officially defined) may not be sufficiently complete summaries of the economic conditions a policymaker might be concerned with.

The explanation lies in the metaphor of the bog. The estimated time of arrival to a destination—policy liftoff, for example—depends critically on the certainty with which the policymaker can assess the economic landscape. An adjustment to policy can, and should, proceed more quickly if the ground underfoot feels relatively solid. But if the terrain remains unfamiliar, and the possibility of falling into the swamp can’t be ruled out with any degree of confidence...well, a wise person moves just a bit more slowly.

Of course, as noted, once you begin to travel across the field and gain confidence that you are actually on terra firma, you can pick up the pace and adjust the estimated time of arrival accordingly.

To put all of this a bit more formally, an individual FOMC participant’s “reaction function”—the implicit rule that connects policy decisions to economic conditions—may not depend on just the numbers that that individual writes down for inflation, unemployment, or whatever. It might well—and in the case of our thinking here at the Atlanta Fed, it does—depend on the confidence with which those numbers are held.

For us, anyway, that confidence is growing. Don’t take that from me. Take it from Atlanta Fed President Lockhart, who said in a recent speech:

I'll close with this thought: there are always risks around a projection of any path forward. There is always considerable uncertainty. Given what I see today, I'm pretty confident in a medium-term outlook of continued moderate growth around 3 percent per annum accompanied by a substantial closing of the employment and inflation gaps. In general, I'm more confident today than a year ago.

Viewed in this light, the puzzle of moving dots without moving point estimates for economic conditions really shouldn’t be much of a puzzle at all.

By Dave Altig, executive vice president and research director of the Atlanta Fed

September 29, 2014 in Economic conditions, Economics, Federal Reserve and Monetary Policy, Forecasts | Permalink | Comments (0) | TrackBack (0)

September 15, 2014

The Changing State of States' Economies

Timely data on the economic health of individual states recently came from the U.S. Bureau of Economic Analysis (BEA). The new quarterly state-level gross domestic product (GDP) series begins in 2005 and runs through the fourth quarter of 2013. The map below offers a look at how states have fared since 2005 relative to the economic performance of the nation as a whole.

It’s interesting to see the map depict an uneven expansion between the second quarter of 2005 and the peak of the cycle in the fourth quarter of 2007. By the fourth quarter of 2008, most parts of the country were experiencing declines in GDP.

The U.S. economy hit a trough during the second quarter of 2009, according to the National Bureau of Economic Research, but 20 states and the District of Columbia recovered more quickly than the rest. The continued progress is easy to see, as is the far-reaching impact of the tsunami that hit Japan on March 11, 2011, which disrupted economic activity in many U.S. states. By the fourth quarter of 2013, only two states—Mississippi and Minnesota—experienced negative GDP.

The map shows that not all states are growing even when overall GDP is growing, and not all states are shrinking even when overall GDP is shrinking. But if we want to know more about which states are driving the change in overall GDP growth, then the geographic size of the state might not be so important.

Depicting states scaled to the size of their respective economies provides another perspective, because it’s the relative size of a state’s economy that matters when considering the contribution of state-level GDP growth to the national economy. The following chart uses bubbles (sized by the size of the state’s economy) to depict changes in states’ real GDP from the second quarter of 2005 through the fourth quarter of 2013.

This chart shows how the economies of larger states such as California, New York, Texas, Florida, and Illinois have an outsize influence on the national economy, despite some having a smaller geographic footprint. (Conversely, changes in the relatively small economy of a geographically large state like Montana have a correspondingly small impact on changes in the national economy.)

Overall GDP is now well above its prerecession peak. But have all states also fully recovered their GDP losses? The chart below depicts the cumulative GDP growth in each state from the end of 2007 to the end of 2013. The size of the circle represents the magnitude of the change in the level of real GDP between the end of 2007 and 2013. Most states have fully recovered in terms of GDP. (North Dakota’s spectacular growth stands out, thanks to its boom in the oil and gas industry.) However, Florida, Nevada, Connecticut, Arizona, New Jersey, and Michigan had not returned to their prerecession spending levels as of the end of 2013. For Florida, Nevada, and Arizona, the depth of the collapse in those states’ booming housing sectors is almost certainly responsible for the relative shortfall in performance since 2007.

The next release of the state-level GDP data, scheduled for September 26, will provide insight into the relative performance of state economies during the first quarter of 2014 at a time when overall GDP shrank by more than 2 percent (annualized rate). Some analysts have suggested that weather disruptions were a leading cause for that decline. The state-level GDP data will help tell the story.

By Whitney Mancuso, a senior economic analyst in the the Atlanta Fed's research department

September 15, 2014 in Economic conditions, Economic Growth and Development, Economics, GDP | Permalink | Comments (0) | TrackBack (0)

August 25, 2014

What Kind of Job for Part-Time Pat?

As anyone who follows macroblog knows, we have been devoting a lot of attention recently to the issue of people working part-time for economic reasons (PTER), which means people who want full-time work but have not yet been able to find it. As of July 2014, the number of people working PTER stood at around 7.5 million. This level is down from a peak of almost 9 million in 2011 but is still more than 3 million higher than before the Great Recession. That doesn’t mean they won’t ever find full-time work in the future, but their chances are a lot lower than in the past.

Consider Pat, for example. Pat was working PTER at some point during a given year and was also employed 12 months later. At the later date, Pat is either working full-time, still working PTER, or is working part-time but is OK with it (which means Pat is part-time for noneconomic reasons). How much luck has Pat had in finding full-time work?

As the chart below shows, there is a reasonable chance that after a year, Pat is happily working full-time. But it has become much less likely than it was before the recession. In 2007, an average of 61 percent of the 2006 Pats transitioned into full-time work. The situation got a lot worse during the recession, and has not improved. In 2013, only 49 percent of the 2012 cohort of Pats had found a full-time job. The decline in finding full-time work is largely accounted for by the rise in the share of Pats who are stuck working PTER. In 2007, 18 percent of the Pats were still PTER after a year, rising to around 30 percent by 2011, where it has essentially remained.

Now, our hypothetical Pats are a pretty heterogeneous bunch. For example, they are different ages, different genders, different educational backgrounds, and in different industries. Do such differences matter when it comes to the chances of Pat finding a full-time job? For example, let’s look at Pats working in goods-producing industries versus services-producing ones. In goods-producing industries, the chance is greater that Pat will find full-time work (more jobs in goods-producing industries are full-time), and there is a bit more of a recovery in full-time job finding for goods-producing industries than for services-producing ones. But overall, the dynamics are similar across the broad industry types, as the charts below show:

As another example, the next four charts show the average 12-month full-time and PTER job-finding rates for all of our hypothetical Pats by gender and education. The full-time/PTER finding rates display broadly similar patterns across gender and education, albeit at different levels. (The same holds true across age groups but is not shown.)

People who find themselves working part-time involuntarily are having more difficulty getting full-time work than in the past, even if they stay employed. But it doesn’t seem that much of this can be attributed to any particular demographic or industry characteristic of the worker. The phenomenon is pretty widespread, suggesting that the problem is a general shortage of full-time jobs rather than a change in the characteristics of workers looking for full-time jobs.

By John Robertson, a vice president and senior economist, and

 

Ellyn Terry, an economic policy analysis specialist, both of the Atlanta Fed's research department

August 25, 2014 in Economic conditions, Employment, Labor Markets | Permalink | Comments (2) | TrackBack (0)