2013 Business Hiring Plans: Employment, Effort, Hours, and Fiscal Uncertainty

How much is fiscal uncertainty holding back hiring? The answer seems to depend on whom you ask. Early in January, the Atlanta Fed spoke to 670 businesses in the Southeast about employment. Conditional on the respondents’ 2013 hiring plans (expand, hold steady, or contract), the following set of charts summarizes the results for how the businesses viewed activity relative to their own interpretation of “normal” along three dimensions: their current employment level, the amount of effort required from their staff per hour, and the average hours worked per employee. These questions were modeled on questions asked in the Atlanta Fed’s December 2012 Business Inflation Expectations Survey. In the following three charts, the green bars represent firms that said they planned to expand employment in 2013. The grey bars represent firms that said they did not plan to change their employment level in 2013, and the red bars represent firms that planned to reduce employment in 2013.

The first chart shows the results for current employment. Regardless of hiring plans over the next 12 months, most firms said they were currently at or below normal employment levels. Those planning on increasing employment over the next 12 months were a bit more likely to say they have already surpassed normal levels of employment than other firms, while those looking to shed employees were very likely to say their employment level is below normal employment levels.

Chart 2 shows that businesses are generally pushing hard along the effort dimension. Firms were quite likely to say that their staff’s effort per hour worked was currently at or above normal,  whether or not they were planning to change employment in 2013.

Chart 3 shows that firms planning to expand were very likely to say that average hours worked were at or above normal (28 percent said hours were above normal, 60 percent about normal), whereas firms planning to contract were more likely to say that hours were at or below normal (48 percent about normal, 39 percent below normal).

Taken together, these results suggest that some firms are approaching the limit of how far they can go along the intensive margins of effort and hours before they have to hire more workers. With effort elevated, as more firms increase average hours worked to above-normal levels, one might expect more hiring to follow.

Each business was also asked how uncertainty about future fiscal policy was affecting its hiring plans. Firms planning to reduce employment tended to cite fiscal uncertainty as having a negative impact on their hiring plans. However, for those firms, hours also tended to be well below normal, so it is unlikely that removing fiscal uncertainty would move many of those firms into expansion mode (although it may help stabilize their outlook).

In contrast, fiscal uncertainty was generally viewed as having less impact by those planning to expand employment and those planning to hold employment levels steady. Presumably, reducing fiscal uncertainty would move some of the firms planning to hold steady into expansion mode, and those planning to expand would do so a bit more. To get some idea of this potential, Chart 4 shows the responses by firms who reported above-normal effort per hour and above-normal average hours worked. About 40 percent of those businesses said that fiscal uncertainty had caused them to scale back their hiring plans.

It is unclear whether eliminating fiscal uncertainty would have a big impact on the hiring plans of these firms. But these results suggest that it sure couldn’t hurt.

By John Robertson, vice president and senior economist, and

Ellyn Terry, a senior economic analyst in the Atlanta Fed's research department

Reading Labor Markets

When the September employment report was released on October 5, the top-line payroll employment gain for the month, as reported in the U.S. Bureau of Labor Statistics' (BLS) establishment survey, logged in at 114,000. Under standard assumptions, a number of this magnitude would be barely enough to absorb the growth of the labor force and keep the unemployment rate constant. In contrast, in that same October 5 report we learned from the BLS household survey that the measured unemployment rate fell from 8.1 percent in August to 7.8 percent in September.

According to Friday's BLS report on the employment situation for October, the top-line payroll employment gain for the month from the establishment survey was 171,000. At that pace—which is also the current average gain for the past three months—the Atlanta Fed jobs calculator suggests the unemployment rate should fall another one-half of a percentage point over the next year. At the same time, according to the BLS household survey, the unemployment rate rose from 7.8 percent in September to 7.9 percent in October.

This is as good an illustration as any to explain why, on November 1, Atlanta Fed President Dennis Lockhart said the following in a speech to the Chattanooga Tennessee Downtown Rotary Club:

In its post-meeting statement on September 13, the FOMC said, "If the outlook for the labor market does not improve substantially, the Committee will continue its purchases of agency mortgage-backed securities, undertake additional asset purchases, and employ its other policy tools as appropriate until such improvement is achieved in a context of price stability."...

For policy purposes, I think it's appropriate to be cautious about relying on a single indicator of labor market trends—for example, the unemployment rate—to determine whether the condition of "substantial improvement" has been met.

As the FOMC went into its September meeting, the official BLS statistics indicated that net U.S. job creation in August was a mere 92,000. That number is below the “all else equal” threshold of about 100,000 jobs required to keep the unemployment rate from rising, and that information is what Fed policymakers had in hand when they met on September 12–13 and decided on the policy action described by President Lockhart.

On Friday, after two revisions, the BLS told us jobs expanded by 192,000 in August, well above the average for the jobs recovery that started in early 2010, 100,000 jobs (more than double) above the initial estimate.

Looking through month-to-month variations is not a lot of help in real-time tea-leaf reading. Here is the 12-month moving average of employment gains, the blue line indicating the way things looked in September, the red line showing the way they look today:

(enlarge)

Over time, it remains the case that monthly employment gains are pretty consistently coming in at 150,000 to 160,000 jobs created per month, and that rate has been enough to generate relatively steady declines in the unemployment rate:

(enlarge)

That could change, of course, and the last four months of data have generally shown an acceleration in the job-growth trend. But the data definitely were not indicating that trend as it was happening, an unfortunate reality that isn't likely to change. One way to soften the blow of that problem, emphasized in the Lockhart speech, is to keep an eye on as a broad a set of signals as possible:

... let me share a qualitative framework for defining "substantial improvement."

The starting point certainly should be the headline unemployment rate and the payroll jobs number. The interpretation of movements in these two statistics would be enriched and reinforced by a review of additional data elements.

I added the emphasis there, as I think the point bears highlighting. President Lockhart goes on to give examples of what he would look for in determining whether the substantial improvement threshold has been met. Things like reductions in the numbers of marginally attached and discouraged workers, growing labor force participation rates, declining numbers of people who want full-time work but have to settle for part-time, and positive forward indicators like falling initial claims for unemployment insurance.

The Calculated Risk blog continues to be a one-stop shop for a lot of this information—here and here, for example—and overall nothing much overturns the picture of steady, but slow, progress. That would suggest the acceleration of the past several months is probably not a new trend, but a continuation of the same-old same-old. But then again, the track record painfully demonstrates how hard that is to know in real time.

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

 

Divergent Jobs Reports: Will the Real State of the Labor Market Please Stand Up?

The September employment report from the U.S. Bureau of Labor Statistics (BLS) was predestined to create a significant amount of buzz. But the confluence of headline jobs growth at a modest clip of 114,000 and a surprisingly large 0.3 percentage point reduction in the unemployment rate has made the report more buzz-worthy than we (here at macroblog) expected. Although some of the commentary has been more heat than light, there have been some particularly good reminders of the difference between the establishment survey data, from which the headline jobs figure is derived, and the household survey data, from which come the unemployment statistics. The discussions on Greg Mankiw's blog and by Catherine Rampell (at The New York Times's Economix blog) are especially useful. Or, perhaps even better, you can go to the source at the BLS.

It's important to remember that both surveys are subject to error and, because of its much smaller sample size, the household survey can be subject to particularly sizeable swings. Specifically, the standard error of the household survey's monthly change in employment is 436,000(!). Based on the most extreme assumptions about flows in and out of unemployment and in and out of the labor force, understating or overstating actual employment by 436,000 would imply a measured unemployment rate ranging from 7.5 percent to 8.1 percent. (The BLS estimate of the standard error for unemployment puts a range on September's number of 7.6 percent to 8 percent.)

In his post, Greg Mankiw makes reference to a Brookings Institution paper by George Perry from a few years back that offers what is probably good advice: since both the payroll and household surveys are subject to error, and since the errors in each are likely unrelated to one another, the clearest picture about what is happening to employment in real time can be gleaned by combining information from both.

In fact, in a directional sense, both the household and payroll surveys are giving the same signals. In the table below, we compare the recent trends in monthly job gains measured in both surveys. The coverage in the two reports is slightly different. Unlike the payroll count, the household survey includes the self-employed and counts multiple jobs held by a single person as a single instance of employment. Because of this, the BLS also reports an adjusted version of the household survey, called the payroll concept adjusted employment measure. This payroll-consistent measure is designed to control for definitional differences across the household and establishment reports and also makes statistical adjustments for changes to the population controls in various years. So we include the data from this measure in the last column of the table.

Overall, all three measures suggest a weaker trend over the last six months than over the last nine months. All three measures also indicate that things were somewhat stronger on average in the last three months than in the prior three months. The bottom line in our view is that, though the employment levels can be quite different across the three measures, all suggest that the jobs picture has improved somewhat in the past three months.

The suggestion in George Perry's Brookings paper—combining the household and establishment data—can be implemented by constructing a weighted average of the two surveys. In our variation we put weights in proportion to the inverse of the sampling variability of the payroll and household surveys, which would roughly imply an 80 percent weight on the establishment measure and 20 percent on the payroll-consistent household measure. The estimates using these weights are reported in the last column of the table above. Because the component employment measures display directionally similar trends in recent months, the weighted average does as well. 

In a speech given a few weeks ago, Atlanta Fed President Dennis Lockhart, our boss here, offered the opinion that

Taking a two-year view, the trend rate of gains in employment has been roughly 150,000 per month. This pace would be sufficient, at current levels of participation in the workforce, to sustain a steady, gradual reduction of the unemployment rate.

As the September employment reports show, predicting the unemployment rate month to month can be tricky business, and there may be better ways than just extrapolating from the jobs data. But thus far we are inclined to think that slow but steady progress on the jobs front is still the best story.

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

Examining the Recession’s Effects on Labor Markets

Four years after the onset of the Great Recession, labor market outcomes in the U.S. remain depressed. The fraction of 16- to 64-year-old individuals who are employed fell from above 72 percent in 2007 to less than 67 percent in 2009 and remains stuck there. The unemployment rate rose from 4.5 percent to 10 percent and still hovers above 8 percent. And the fraction of unemployed workers who have been looking for a job for more than six months has increased to a share not seen in the United States in at least 60 years. The Atlanta Fed's Center for Human Capital Studies hosted a conference last weekend, organized by Richard Rogerson (Princeton University), Robert Shimer (University of Chicago) and the Atlanta Fed's Melinda Pitts that explored why the employment losses were so large and why the labor market recovery has been so weak. Examining these questions is important because different hypotheses about the nature of the recession suggest that different policy interventions may help to accelerate the recovery.

The paper "On the Importance of the Participation Margin for Labor Market Fluctuations" by Michael Elsby, Bart Hobijn, and Ayşegül Şahin offered some suggestions on how to think about the disparate behavior of the unemployment rate and labor force participation rate during the last couple of years. While the unemployment rate has steadily fallen back towards its historic levels, labor force participation has fallen, keeping the employment-population ratio constant. At some level, this movement suggests that the decline in labor force participation has acted as a relief valve for the unemployment rate. Using evidence on the gross flows of workers between employment, unemployment, and out-of-the-labor-force, Elsby and his coauthors question that interpretation. Instead, relatively few unemployed workers have dropped out of the labor force during the recovery, reflecting the high desire to work among the current stock of unemployed individuals.

A number of papers offered specific hypotheses about the reason for the large and persistent deterioration in labor market outcomes and tested those hypotheses using a variety of methodologies and datasets. For example, the paper "What Explains High Unemployment? The Aggregate Demand Channel" by Atif Mian and Amir Sufi explored the implications of the negative shock to household balance sheets that followed the collapse in house prices. They document that employment in the nonconstruction, nontraded sector declined most in U.S. counties that experienced the largest adverse shock to house prices, while the decline in the traded goods sector occurred equally nationwide. If wages and prices were flexible, we would expect the balance sheet shock to reduce the demand for nontraded goods and raise the supply of labor and hence employment in the traded good sector. The fact that this did not happen is evidence that wages and prices have not adjusted. They infer that roughly two-thirds of the total employment losses can be attributed to the balance sheet shock, in combination with wage and price rigidities.

A second hypothesis is that the recovery has been so weak because of underlying adverse trends in the U.S. labor market. "Manufacturing Busts, Housing Booms, and Declining Employment: A Structural Explanation" by Erik Hurst, Matt Notowidigdo, and Kerwin Charles shows how the ongoing decline in the demand for less educated men in manufacturing has generated a negative trend in labor market outcomes for these workers for three decades. This trend continued unabated during the years after the 2001 recession but was masked by the housing boom, which lifted employment for less-skilled workers for another five years. This observation is relevant for how one interprets the time series changes in labor market outcomes. If we view the housing boom as an aberration that is unlikely to resume, it is inappropriate to compare current labor market outcomes with those just preceding the onset of the Great Recession.

The paper "The Trend is the Cycle: Job Polarization and Jobless Recoveries" by Nir Jaimovich and Henry Siu focuses on a related but distinct long-term phenomenon in the U.S. labor market: job polarization. This refers to the fact that the U.S. labor market increasingly consists of low- and high-paying jobs with relatively few middle-income jobs. While this ongoing change has been noted by other researchers, Jaimovich and Siu show that this long-term evolution has not been occurring at a slow and steady rate but rather has been concentrated during aggregate downturns. They argue that the recent phenomenon of jobless recoveries is simply a reflection of the fact that these are the periods in which middle income jobs are disappearing, never to be brought back.

On the other hand, "The Labor Market Four Years Into the Crisis: Assessing Structural Explanations" by Jesse Rothstein explores and finds little direct evidence for a number of specific structural channels that might explain the weak recovery. For example, there are no identifiable sectors of the U.S. economy with strong wage growth, which suggests that the shortage of suitable workers is probably not a large constraint on employment growth.

A third hypothesis is that the weak recovery reflects an increase in economic uncertainty, which induces firms to wait rather than hire and invest. "Measuring Economic Policy Uncertainty" by Scott Baker, Nicholas Bloom, and Steve Davis proposes a novel methodology for quantifying the overall level of economic uncertainty and the portion of uncertainty that is induced by economic policy. They show that both measures of uncertainty have been elevated since the onset of the Great Recession and have scarcely recovered during recent years. "Uncertainty, Productivity and Unemployment in the Great Recession" by Edouard Schaal examines how an increase in uncertainty affects labor market outcomes in the context of a job search model. He focuses on one measure of uncertainty, the cross-sectional variability of sales growth rates across business establishments, which increased sharply in 2008 but has since subsided. Because of this finding, Schaal finds that the model can account for a large deterioration in labor market outcomes at the time of the shock but that it cannot explain why the deterioration has been so persistent.

A final hypothesis is that the weak recovery reflects disincentive effects of new tax and transfer programs that have been introduced since the onset of the recession. One aspect of this that has attracted particular attention is the extension of unemployment benefits. "The Effect of Unemployment Insurance Extensions on Reemployment Wages" by Johannes Schmieder, Till von Wachter, and Stefan Bender uses evidence from Germany to explore this hypothesis. They show that extending unemployment benefits by six months causes approximately a one-month increase in the amount of time it takes an individual to return to work. This extension has two effects on the wage of workers when they return to work. On the one hand, the additional time to look for a job allows workers to find better jobs. On the other hand, workers' skills tend to decline during an unemployment spell. On net, these effects roughly cancel so extended benefit programs do not have a large impact on average wages.

The framework that most economists use to study the behavior of unemployed workers is search theory. Robert Hall's paper "Viewing the Observed Acceptance Decisions of Job-Seekers through the Lens of Search Theory" analyzes detailed data on the job finding process for a sample of unemployed workers in New Jersey from 2009 in the context of this theory to assess how well the theory can provide a consistent explanation for observed behavior. Previous work had suggested that this framework has problems in accounting for observed job acceptance decisions, but Hall shows that with a few simple modifications, the framework offers a consistent explanation of how workers behave given labor market conditions.

The discussions at the conference questioned the usefulness of labels like deficient demand, structural unemployment, and cyclical unemployment. These terms mean different things in different contexts and do not clarify the key causal factors. Explanations such as "employment is slow because uncertainty is high" could easily fit under any of these banners. Instead, isolating the key changes that have taken place in the U.S. economy, and then scrutinizing the factors that have influenced how those changes have affected the labor market, would be more conducive to arriving at answers.

By Richard Rogerson of Princeton University and Robert Shimer of the University of Chicago, both advisers to the Atlanta Fed’s Center for Human Capital Studies, and Melinda Pitts, a research economist and associate policy adviser in the Atlanta Fed's research department

 

The skills gap: Still trying to separate myth from fact

Peter Capelli has looked at the skills gap explanation for labor market weakness and sees more myth than fact:

"Indeed, some of the most puzzling stories to come out of the Great Recession are the many claims by employers that they cannot find qualified applicants to fill their jobs, despite the millions of unemployed who are seeking work. Beyond the anecdotes themselves is survey evidence, most recently from Manpower, which finds roughly half of employers reporting trouble filling their vacancies.

"The first thing that makes me wonder about the supposed 'skill gap' is that, when pressed for more evidence, roughly 10% of employers admit that the problem is really that the candidates they want won't accept the positions at the wage level being offered. That's not a skill shortage, it's simply being unwilling to pay the going price."

To some extent, the issue is semantic:

"But the heart of the real story about employer difficulties in hiring can be seen in the Manpower data showing that only 15% of employers who say they see a skill shortage say that the issue is a lack of candidate knowledge, which is what we'd normally think of as skill. Instead, by far the most important shortfall they see in candidates is a lack of experience doing similar jobs. Employers are not looking to hire entry-level applicants right out of school. They want experienced candidates who can contribute immediately with no training or start-up time..."

In the language of economists, Capelli is defining skill as the possession of generalized human capital, while businesses are defining skill as the possession of firm- or job-specific human capital. In more familiar language, Capelli appears to be focused on innate skill levels and education, while businesses are looking for the types of skills that would be attained through past on-the-job training. In even more colloquial language, Capelli wants businesses to appreciate book-learning, and businesses prefer those who have already survived the school of hard knocks.

We have recently completed our own version of the Manpower survey Capelli references. Our results are based on the responses of about 100 businesses in the Sixth Federal Reserve District represented by the Atlanta Fed, and we do not claim that they are conclusive. But we do think they are instructive.

Of those firms that said they experienced an increase in hiring difficulty over the last year, our poll respondents confirm the notion that businesses are looking for candidates with specific skills:

(enlarge)

The lack of technical skills is the only factor that really jumps out as an issue that businesses have with the pool of job applicants. We often hear anecdotal complaints about job seekers' lack of "soft skills," or the difficulty in finding applicants who can pass required background checks. But only 14 percent of all selections indicated too few applicants with required interpersonal skills, and only 7 percent indicated a problem with applicants passing screening requirements like drug-use or credit checks.

On the other hand, our poll found scant support for Capelli's claim that businesses are "unwilling to pay the going price." Only 9 percent of respondents reported that too few applicants would accept the offered compensation package.

Despite the fact that we see some evidence consistent with skill mismatch, it is far from clear that this issue is the smoking gun that explains the current anemic state of job growth. When asked if a dearth of skilled applicants is a persistent problem, our survey respondents overwhelmingly answer "yes." But when asked if they have had more difficulty hiring over the past 12 months, the overwhelming majority answered "no":

(enlarge)

Even among the minority of businesses that report recent hiring difficulties, only half indicate that this difficulty is restraining growth:

(enlarge)

We infer a couple of lessons from all of this information. First, it does appear that there is a long-term skill level problem in the U.S. economy. Adopting Capelli's definition of skill does not mean the existence of skill mismatch is a myth.

But turning to the short run, we've been pretty sympathetic to structural explanations for the slow pace of the recovery. Nonetheless, we have yet to find much evidence that problems with skill-mismatch are more important postrecession than they were prerecession. We'll keep looking, but—as our colleagues at the Chicago Fed conclude in their most recent Chicago Fed Letter—so far the facts just don't support skill gaps as the major source of our current labor market woes.

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

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

Will labor force participation continue to rise?

The labor force participation rate ticked up in May, as did the rate of unemployment. As we have noted in the past, the near-term trajectory of the unemployment rate depends critically on what happens to the participation rate. So the question is, can we expect further upward changes in the participation rate? The answer depends a lot on the labor market attachment of those that are currently out of the labor force.

A few weeks ago, my frequent coauthor, Julie Hotchkiss, wrote about what we can gain from detailed labor market data about the activities of people who have exited the labor force. In her posting, she discussed the overall increase in exits from the labor force, with a focus on 25–54 year olds. Her work concluded that while people identified "Household Care" as the dominant activity for those not in the labor force, there has been a significant upward shift since the recession in those indicating "School" or "Other" as their primary reason for not being in the labor force. A supposition is that at least those that indicated they were in school would reenter the labor force at some point, doing so with a higher level of skills or, at least, with skills that are better aligned with labor demand. However, because we know little about those in the Other category, the future labor market attachment for them is less clear.

This post explores data on transitions into the labor force, primarily for those in the Other category. As in the earlier blog, the focus is on individuals aged 25–54, as retirement dominates the activity of older individuals not in the labor force and schooling dominates the activity of younger individuals not in the labor force.

One indicator of whether those in the Other group are planning to reenter the labor force is whether the individuals in this group are classified as marginally attached to the labor force. A nonparticipant who is marginally attached indicates they want employment or are available for employment. Also, they indicate having looked for a job in the previous year but not actively looking for a job at present. Using monthly data from the Current Population Survey (CPS) that are matched year over year, we see that the marginally attached workers do transition back into the labor force at twice the rate of all individuals who are not in the labor force, as chart 1 illustrates. These rates are relatively stable over time.

(enlarge)

As chart 2 shows, a much higher proportion of individuals in the Other category are marginally attached to the labor force, compared to other types of nonparticipants. Moreover, the percentage of these marginally attached nonparticipants has increased from around 20 percent to 30 percent over the last three years.

(enlarge)

This higher probability of marginally attached workers returning to the labor force combined with the significantly increased share of marginally attached workers in the Other category suggests that we should expect to find a higher share of those in the Other group returning to the labor force than we've seen in the past. But it turns out that this expected development is not what has happened. The Other group also includes individuals who are not marginally attached to the labor market, and their transition rates into the labor market have declined. On net, while the transition rate to employment is highest for the Other category (reflecting the large of share of marginally attached), the transition rate into the labor force does not fully reflect the increased level of marginal attachment to the labor force.

(enlarge)

The group with the next highest transition rate to employment is in the School category, which reflects the inherent transitory nature of that activity. However, it is noteworthy that the school transition rate is lower than it was before the recession. This development reflects an increase in the share of individuals continuing to indicate that school is their primary reason for not participating in the labor force from one year to the next. And it suggests that the lower opportunity cost of attending school is influencing the decision to remain in school longer.

While these trends suggest that we could expect to see higher rates of return to the labor force going forward, this potential development will likely require a much better showing of jobs numbers than were seen today before kicking in.

By Melinda Pitts, research economist and associate policy adviser

Labor force nonparticipants: So what are they doing?

As Dave Altig, Atlanta Fed research director, pointed out earlier this week in this blog post, there is a great deal of interest these days in the labor force participation rate—particularly its level and the direction it's going. The question that seems to be on everyone's mind is how many of the nonparticipants in the labor force can we expect to return to the market. The answer to this question has immediate implications for the unemployment rate (especially if all these nonparticipants were to return to unemployment rolls), and longer-term implications for economic growth—our economy needs workers to fuel production.

The analyses that I can find to date are all primarily focused on a statistical detangling of demographic versus behavioral changes, structural versus cyclical changes, and employment trend versus employment gap debates. But all of this discussion begs the question that my colleague, Melinda Pitts, and I have been investigating: What are these labor force nonparticipants doing? Perhaps an answer to that question will help us get a better handle on which nonparticipants are likely to return to the labor force in the near future.

The Current Population Survey (CPS), administered by the U.S. Bureau of Labor Statistics (BLS), asks labor force nonparticipants about their reason for absence (details of the CPS questionnaire are available from the NBER). The reason given by nonparticipants that gets most of the attention is "discouraged over job prospects." In April 2012, these people accounted for only 1.1 percent of all nonparticipants (41 percent of the marginally attached—those who want a job, are available to work, and searched in the previous year). The vast majority of nonparticipants are absent because of retirement, disability, going to school, caring for household members, or other reasons.

Using the latest survey data we have available (November 2011), we find that most nonparticipants are retired (48 percent); the share who are in school, disabled, or taking care of household members are 18 percent, 16 percent, and 15 percent, respectively; and the share in the category termed "Other" comes in at about 2 percent.

For purposes of better understanding the decline in labor force participation, however, we look at the reasons for absence given by people who leave the labor force. Those who have left the labor force are arguably more likely to return (depending on the reason, of course) than those who have never been in the labor force. A feature of the CPS allows us to track certain individuals from one year to the next, so we are able to identify people who leave the labor force. Chart 1 illustrates how individuals who are not in the labor force—but who were employed or unemployed the previous year—are distributed across the reasons for nonparticipation. The raw data are not seasonally adjusted, of course, so we plot the numbers as a 12-month moving average—this approach does not affect the overall observed trends in the data. In addition, we restrict our analysis here to those between the ages of 25 and 54, since retirement overwhelmingly dominates the nonparticipation decisions of older workers, and schooling dominates the nonparticipation decisions of younger workers.

(enlarge)

Chart 1 illustrates what the labor force participation rates have been telling us. For every reason given for absence, except perhaps "Retired," the number of people leaving the labor force has increased during or after the recession of 2008. The most dramatic increases are seen among those people giving "School" and "Other" as a reason. However, since we are in search of changes in reasons that might be out of the ordinary, especially any significant upward shifts in nonparticipants giving a particular reason during and after the recession, we also look at how these folks leaving the labor force are distributed across the different reasons. This information will tell us whether the number of people giving one particular reason increased disproportionately compared with the other reasons.

Chart 2 plots the shares of all of those leaving the labor force (ages 25–54) giving each reason for their absence. Since the beginning of the recession, there has been a significant shift toward the reasons of "School" and "Other" among nonparticipants who have left the labor force within the previous year. The share levels attained by the reasons of "School" and "Other" are historically unprecedented by the end of the data series. These shifts also appear to have come mostly from a decline in the share of people leaving the workforce to take care of household members (HHcare). This is evidenced through the dramatic drop in the share giving the "HHcare" reason at the same time.

(enlarge)

It is difficult to interpret the implications of the rise in share of "Other" as a reason for nonparticipation among those leaving the labor force, although this category may be capturing some of the discouraged workers. The implication for the rise in "School" is unmistakable, however. With reasonable expectations, these individuals should re-enter the labor force with enhanced—or at least better-aligned—skills that will be able to make a positive contribution to overall economic growth.

By Julie Hotchkiss, research economist and policy adviser in the Atlanta Fed's research department

 

A take on labor force participation and the unemployment rate

By now, if you've been paying attention to the coverage following the April employment report, you know the following:

• The March to April decline in the unemployment rate from 8.2 percent to 8.1 percent was arithmetically driven by yet another decline in the labor force participation rate (LFPR).
• The decline in the LFPR, now at its lowest level since the early 1980s, is itself being influenced by a confounding mix of demographic change and other behavioral changes that nobody seems to understand—a point emphasized by a gaggle of blogs and bloggers such as Brad DeLong, Carpe Diem, Conversable Economist, Free Exchange, and Rortybomb, to name a few.

With respect to the first observation, in a previous post my colleague Julie Hotchkiss described how to use our Jobs Calculator to get a ballpark sense of what the unemployment rate would have been had the LFPR not changed. If you follow those procedures and assume that the LFPR had stayed at the March level of 63.8 percent instead of falling to 63.6 percent, the unemployment rate would have risen to 8.4 percent instead of falling to 8.1 percent.

It is clear that interpreting this sort of counterfactual experiment depends critically on how you think about the decline in the LFPR. The aforementioned post at Rortybomb cites two Federal Reserve studies—from the Chicago Fed and the Kansas City Fed—that attempt to disentangle the change in the LFPR that can be explained by trends in the age and composition of the labor force. These changes are presumably permanent and have little to do with questions of whether the labor market is performing up to snuff.

The following chart, which throws our own estimates into the mix, illustrates the evolution of the actual LFPR along with an estimate of the LFPR adjusted for demographic changes:

enlarge

As the header on the chart indicates, our estimates suggest that roughly 40 percent of the change in the LFPR since 2000 can be accounted for by changes in age and composition of the population—in essentially the same range as the Chicago and Kansas City Fed studies. (If you are interested in the technical details you can find a description of the methodology used to generate the chart above, based on work by the University of Chicago's Rob Shimer.

In other words, 0.9 percentage points of the decline in the LFPR since the beginning of the past recession can be explained by demographic trends (as the baby boomers age, the labor force will grow more slowly than the total population [ages 16 and up]). Subtracting the demographic trends still leaves 1.5 percentage points to be explained, a number right in line with Brad DeLong's back-of-the-envelope calculation of "cyclical" LFPR change.

As DeLong's comments make clear, the interpretation of the nondemographic piece of the LFPR change requires, well, interpretation. And the consequences of connecting the dots between changes in the unemployment rate and broader labor market performance are enormous.

In the recently released Summary of Economic Projections following the last meeting of the Federal Reserve's Federal Open Market Committee, the midpoint of the projections for the unemployment rate at the end of 2013 is 7.5 percent. Turning again to our Jobs Calculator, we can get a sense of what sort of job creation over the next 20 months will be required given different values of the LFPR. For these estimates, I consider three alternatives: The LFPR stays at its April level, the LFPR reverts to our current estimate of the demographically adjusted level (that is, increases by 1.5 percentage points), and an intermediate case in which the LFPR increases by 0.7 percentage points—the lower end of DeLong's estimate of "people who really ought to be in the labor force right now, but who are not."

enlarge

DeLong asks:

"Are [people who really ought to be in the labor force right now, but who are not] now part of the 'structurally' non-employed who we will never see back at work, barring a high-pressure economy of a kind we see at most once in a generation?"

As you can see, the answer to that question matters a lot to how we should think about progress on the unemployment rate going forward.

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

 

Is the composition of job growth behind slow income growth?

Harold Meyerson, Washington Post opinion writer, channels a Bloomberg report (via The Big Picture), and thinks he finds a smoking gun:

"Why is this recovery different from all other recoveries?

"... what really sets the current recovery apart from all its predecessors is this: Almost three years after economic growth resumed, the real value of Americans' paychecks is stubbornly still shrinking. According to Friday's Bloomberg Economics Brief, ‘the pace of income gains is well below that of the past two jobless recoveries and real average hourly earnings continue to decline.'

"The Bloomberg report cites one reason for this anomaly: Most of the jobs being created are in low-wage sectors. According to Bloomberg, fully 70 percent of all job gains in the past six months were concentrated in restaurants and hotels, health care and home health care, retail trade, and temporary employment agencies. These four sectors employ just 29 percent of the country's workforce but account for the vast majority of the jobs being created."

Meyerson accurately repeats the Bloomberg story, but that story itself is somewhat misleading. To begin with, the 70 percent figure appears to include the entire category of professional and business services, of which temporary help services are only a part. The types of jobs that fall under the professional and business service label are broadly described by the U.S. Bureau of Labor Statistics and include employment in scientific and technical services, management jobs as well as administrative and support type jobs. In particular, the professional scientific and technical services sector is described as follows...

"The Professional, Scientific, and Technical Services sector comprises establishments that specialize in performing professional, scientific, and technical activities for others. These activities require a high degree of expertise and training. The establishments in this sector specialize according to expertise and provide these services to clients in a variety of industries and, in some cases, to households. Activities performed include: legal advice and representation; accounting, bookkeeping, and payroll services; architectural, engineering, and specialized design services; computer services; consulting services; research services; advertising services; photographic services; translation and interpretation services; veterinary services; and other professional, scientific, and technical services."

... and here is the description of management of companies and enterprises sector:

"The Management of Companies and Enterprises sector comprises (1) establishments that hold the securities of (or other equity interests in) companies and enterprises for the purpose of owning a controlling interest or influencing management decisions or (2) establishments (except government establishments) that administer, oversee, and manage establishments of the company or enterprise and that normally undertake the strategic or organizational planning and decision making role of the company or enterprise. Establishments that administer, oversee, and manage may hold the securities of the company or enterprise."

These parts of the economy are hardly made up of the prototypical low-wage jobs and, according to my calculations, you don't get to Bloomberg's 70 percent number without including them.

If you focus strictly on "restaurants and hotels" (or, more precisely, the leisure and hospitality sector), health care, retail, and temporary employment services, your conclusion would be that these sectors accounted for about 50 percent of total job growth/change over the past six months, a share that may still strike you as pretty significant. But is it really? A little historical context might help:

(enlarge)

It is true that this expansion, which began in July 2009, has been unusually concentrated in the four sectors identified by Bloomberg and highlighted in the Meyerson piece. However, a closer look reveals that the only one of the four that looks unusual is employment in temporary help services, the share of which in this recovery has been five times the post-1990 level as a whole. (We reach the same conclusion even if we compare where we are today in this recovery—roughly three years out—with that same period following the recoveries from the 1991 and 2001 recessions.)

On the other hand, it is also true that the share of temp services in total jobs gains has been much lower over the past six months than it was earlier in this recovery. I don't know if that share will eventually fall to the (remarkably stable) level that characterized the (almost) two decades before the past recession. But even if that share remains near 12 percent, as opposed the more historical 6 percent level, I think the story remains the broad-based nature of the relatively tepid growth (in incomes and jobs) that has characterized this recovery.

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

 

The structure of the structural unemployment question

In the middle of its thorough analysis of U.S. labor markets, the New York Fed tucked in a direct look at whether persistently high unemployment can be plausibly ascribed to mismatches between the skill sets of unemployed workers and those skill sets required by available jobs. The operating hypothesis goes something like this: structural unemployment arises when the skills that are appropriate for declining sectors are not easily transferable to the jobs available in expanding sectors. In the current context, we can think, for example, about the challenge of turning construction workers into nurses (a metaphor offered a while back by Philadelphia Fed President Charles Plosser). If skill mismatch is an important source of postcrisis unemployment, it stands to reason that we would find its markers in the construction sector.

In fact, the authors (Richard Crump and Ayşegül Şahin) of a New York Fed study find no evidence that construction workers are "experiencing relatively worse labor market outcomes." Though this observation comes with its caveats—in this space my colleagues Lei Fang and Pedro Silos noted that construction workers who are finding employment in nonconstruction businesses apparently have suffered unusually large wage reductions—the Crump-Şahin results generally conform to other research questioning the proposition that skill mismatch looks to be a larger-than-normal problem in the current recovery.

The idea that inter-sectoral flows of employment, or the lack thereof, is a source of structural unemployment has a venerable history in macroeconomics. But it is increasingly clear to me that the bigger story is not about skill mismatches as workers flow across sectors but about mismatches as workers are faced with changing skill requirements within sectors. In other words, the issue is not changing construction workers into nurses, but changing both construction workers and nurses from old-style workers to new-style workers.

"Old style" and "new style" here refer to jobs defined by the performance of routine tasks versus those that require the performance of nonroutine tasks. The labor market outcomes associated with this shift from old style to new style has come to be known as "job polarization." Job polarization is the subject of a new paper by Nir Jaimovich and Henry Siu, described last week by David Andolfatto at MacroMania:

"Job polarization refers to the recent disappearance of employment in occupations in the middle of the skill distribution...

"Evidently, these classifications correspond to rankings in the occupational income distribution. Non-routine cognitive occupations tend to be high-skill jobs, and non-routine manual occupations tend to be low-skill jobs. Routine occupations—both cognitive and non-cognitive—tend to be middle-skill occupations.

"... across three decades, the share of employment in the middle of the skill distribution appears to be disappearing. Prime suspect: routine biased technological change (e.g., think of ATMs replacing bank tellers)."

The post-1980s job polarization trend has received a lot of attentions over the past decade—notable studies by MIT economist David Autor (here and here), for example—but the essential message of the Jaimovich-Siu study is the observation that trend changes are not smooth, but concentrated around downturns in the economy. Jaimovich and Siu explain:

"... job polarization is not a gradual phenomenon: the loss of middle-skill, routine jobs is concentrated in economic downturns. Specifically, 92% of the job loss in these occupations since the mid-1980s occurs within a 12 month window of NBER [National Bureau of Economic Research] dated recessions (that have all been characterized by jobless recoveries). In this sense, the job polarization 'trend' is a business 'cycle' phenomenon... Our first point is that polarization happens almost entirely in recessions.

"Our second point is that jobless recoveries are due to job polarization... jobless recoveries are observed only in... disappearing, middle-skill jobs. The high- and low-skill occupations to which employment is polarizing either do not experience contractions, or if they do, rebound soon after the turning point in aggregate output. Hence, jobless recoveries are due to the disappearance of middle-skill, routine occupations in recessions."

A few posts back, I posed this question:

"[The pace of improvement in employment, overall and by sector,] draw a clear picture of labor markets that are underperforming by historical standards—a position that I take to be the conventional wisdom. An argument against following that conventional wisdom centers on the question of whether historical standards represent the appropriate yardstick today. In other words, is the correct reference point the level of employment or the pace of improvement in the labor market from a permanently lower level?"

The Jaimovich-Siu results really do suggest that the answer could well be the latter. That said, the levels of employment in the broad nonroutine job categories identified in Jaimovich and Siu's paper remain below the peak levels associated with the 2001 recession—something that was not apparently true at this point in the recoveries after the 1990–91 and 2001 recessions.

Furthermore, not everyone agrees that the Jaimovich-Siu case is persuasive. Mark Thoma, for example:

"There is plenty of evidence pointing in the other direction, i.e. plenty of evidence indicating the problem is cyclical and we are nowhere near full recovery.

"With so much uncertainty remaining, the advice from Stevenson and Wolfers in a post... about how policymakers should react when they are unsure of how strong the recovery will be is appropriate:

'... the cost of too little growth far outweighs the cost of too much. If we readily bear the burden of carrying an umbrella when there's a reasonable chance of getting wet, we should certainly be willing to stimulate the economy when there's a reasonable risk that doing nothing could yield a jobless generation.'

"The fact that the costs are asymmetric and what this means for policy—it should lean against the more costly outcome—seems strangely absent from policy discussions and decisions."

It is worth noting that asymmetric costs referenced here are a matter of judgment, not theory. In fact, if the employment losses suffered through the recession are structural, stimulating the economy is exactly the wrong thing to do. (The classic exposition of this point, in math terms, was provided years ago by Michael Woodford.) In this sense, Thoma's argument just begs the question.

And though there may be "plenty of evidence" pointing in the direction of labor market slack, there is also developing evidence of tightness directly related to the job-polarization phenomenon. From Kathleen Madigan, at The Wall Street Journal:

"The U.S. labor market is struggling with a paradox: despite an 8.3% unemployment rate, many jobs go begging.

"The Institute for Supply Management-New York said this week that 20% of its members say the shortage of skilled labor is an obstacle to business. On Thursday, the National Federation of Independent Business [NFIB] reported a rising share of small business owners who say they have jobs that are hard to fill."

Care should be taken not to over-interpret these types of observations. Though the degree of skill shortages reported in the NFIB surveys was higher in 2011 than 2010, it is still well below prerecession levels. As I indicated in my earlier post, in the end the truth is likely to seen in the behavior of inflation. The asymmetry to which Thoma, and Stevenson and Wolfers, appeal is implicitly based on their belief that the risks of inflation are very low. With that in mind, this summary at Angry Bear of the March employment report warrants some notice:

"Recently, unit labor cost has been rising faster than prices, implying margin pressure and very weak profits. To sustain profits growth, firms have to reestablish stronger productivity growth. The weakness in March employment is a strong indicator that business is trying to rebuild productivity growth and profits growth."

The other possibility, of course, is that businesses will try to rebuild profit growth by raising prices.

The story continues to develop. Watch this space.

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