The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.
- BLS Handbook of Methods
- Bureau of Economic Analysis
- Bureau of Labor Statistics
- Congressional Budget Office
- Economic Data - FRED® II, St. Louis Fed
- Office of Management and Budget
- Statistics: Releases and Historical Data, Board of Governors
- U.S. Census Bureau Economic Programs
- White House Economic Statistics Briefing Room
October 19, 2017
How Ill a Wind? Hurricanes' Impacts on Employment and Earnings
According to the Current Employment Statistics payroll survey, seasonally adjusted nonfarm payroll employment declined 33,000 in September. This decline was the first drop in employment since 2010 and followed a 169,000 gain in August. At the same time, seasonally adjusted average hourly earnings in the private sector increased 2.9 percent year over year in September. This increase in average wages was the largest since the end of the Great Recession in 2009. However, it seems likely that the decline in employment contributed to the rise in average hourly earnings. Why would a decline in employment contribute to an increase in average hourly earnings? We're glad you asked!
As noted by the U.S. Bureau of Labor Statistics, Hurricanes Harvey and Irma reduced employment in the payroll survey, whose reference period is the pay period that includes the 12th of the month. Hurricane Harvey first made landfall in east Texas on August 25 and again in Louisiana on August 30, and Hurricane Irma made landfall in south Florida on September 10. The storms forced large-scale evacuations and severely damaged many homes and businesses. For workers who are not paid when they miss work, being unable to work during the surveyed pay period means they are not counted in September payrolls.
To measure the size of Harvey and Irma's effect on payroll employment, we first looked at data from the Current Population Survey (CPS). We found that the bad weather forced about 1.5 million nonfarm workers who had a job during the September reference week to miss work. Of those, about 1.2 million were wage and salary earners, and about 760,000 of those were unpaid during their absence from work.
Our analysis indicates that September saw a shortfall in seasonally adjusted payroll employment between 200,000 and 300,000 jobs, suggesting that workers returning to work could result in a large rebound in payroll employment. (Not to get too far into the weeds, but our analysis involved regressing payroll employment growth on its lagged values as well as current and lagged seasonally adjusted changes in shares of workers who were not at work because of bad weather.)
What about average hourly earnings? Changes in average hourly earnings over time reflect both the effect of people getting pay raises and changes in who is working this month versus last month or last year. This latter effect can be large during recessions, when workers in lower-wage jobs are disproportionately more likely to be laid off. The absence of these workers from payrolls increases the average wage among the remaining employed workers, even if those remaining workers are not getting much of a pay increase (see this macroblog post for more discussion).
The September payroll survey depicted a particularly large decline in employment in the leisure and hospitality sector, which is significant because average hourly earnings in that sector are typically about 40 percent lower than overall average hourly earnings. In addition, from the CPS we see that the usual hourly earnings of workers not at work because of bad weather is much lower than for other workers. These data suggest that temporary absences from work because of bad weather likely put upward pressure on average hourly earnings, and some of that upward pressure could reverse itself as these workers return to their jobs. If the pace of average hourly earnings doesn't relax, however, then that would suggest more workers getting larger pay raises due to a tightening labor market.
July 31, 2017
Behind the Increase in Prime-Age Labor Force Participation
Prime-age labor force participation has been on a tear recently. Over the last eight quarters, it is up by about 65 basis points (bps) and more than 40 bps in just the last year. When combined with declines in the rate of unemployment, this increase has helped lift the employment-to-population (EPOP) ratio for this key population group by around 120 bps during the last two years.
Placed in the context of an almost 260 bp decline in the prime-age EPOP ratio between 2007 and 2015, this development is significant. Although the unemployment rate is close to what most economists consider full employment, rising labor force participation can indicate that the labor market might still have some room to run before the employment gap is fully closed. (The Congressional Budget Office offers some analysis consistent with this idea.)
So what's behind the increase in prime-age (defined as people between 25 and 54) participation in the last year? Changes in the labor force participation rate (LFPR) either can be the result of changes in the mix of demographic groups in the population with different average rates of participation (for example, across education and race/ethnicity), or they can result from changes in average participation rates within demographic groups. It turns out that most of the increase in the prime-age LFPR has been because of increased LFPR within demographic groups—in particular, prime-age women and especially women without a college degree. Prime-age men have not contributed much to the rise in participation beyond the increased participation associated with a more educated population.
The following chart shows the contribution to the change in the prime-age LFPR over the last year as a result of changes in the relative mix of age-education-race groups (the blue bars) and changes in participation rates within age-education-race groups (the orange bars). It shows the contribution from both sexes combined and from prime-age women and men separately.
Note that the we computed the contributions using six five-year age groups, three education groups (less than high school, high school but no college degree, and college degree), three race/ethnicity groups (Hispanic, non-Hispanic black, and non-Hispanic white/other), and two sexes.
Of the total increase in the prime-age LFPR, most of that was the result of changes in labor force participation behavior within female demographic groups. In fact, changes in LFPR behavior from prime-age men served as a drag on the overall prime-age LFPR. The modestly positive demographic effect on the LFPR for both men and women reflects the higher LFPR for those with a college degree and the relative increase in the share of both prime-age men and women with a college degree.
This development stands in contrast to the drivers of the change in the prime-age LFPR between 2015 and 2016. Of the 24 bp increase in prime-age LFPR between the second quarters of 2015 and 2016, changes in the demographic composition of the population (primarily increased education levels) accounted for all of it rather than changes in average participation rates within demographic groups.
The next chart shows the contribution to the change in the prime-age LFPR between 2016 and 2017 due to changes in the LFPR behavior of women for specific education-race groups.
As the chart shows, the bulk of the demographically adjusted contribution from female labor force participation came from women without a college degree, and the largest contribution across female education-race groups was from Hispanics without a college degree. The increase in labor force participation among women with less education is consistent with evidence of recent improvement in the wage gains for relatively low-wage earners.
Although this simple decomposition doesn't explain why nondegreed women are increasingly finding the labor force to be an attractive option, we can infer some clues by looking at changes in the reasons people give for not participating. In particular, the largest contribution from changes in behavior among prime-age women over the last year came from a decrease in the propensity to be out of the labor force because of poor health or being in the shadow labor force (wanting a job but not looking).
Recently, former Minneapolis Fed President Narayana Kocherlakota has argued that macroeconomists should take more seriously the differences in behavior across demographic groups. The Atlanta Fed's Labor Force Dynamics web page contains more information on the behavioral trends in the reasons people give for not participating in the labor force across demographic groups, and the page was just updated to include data for the second quarter of 2017. Check it out, and we'll keep reporting here on the relative contributions to the labor force of behavioral versus demographic changes—and whether the winning streak for prime-age labor force participation continues.
July 12, 2017
An Update on Labor Force Participation
With the unemployment rate essentially back to prerecession levels, economists have been paying increased attention to the labor force participation rate (LFPR). Many economists, including those at the Congressional Budget Office , believe untapped resources remain on the sidelines of the labor market.
What exactly does "on the sidelines" entail? Discouraged workers are only a small part of the story. To help unravel the rest of the mystery behind the elevated share of people not participating, we at the Atlanta Fed use the microdata from the Current Population Survey to code the activities of persons not in the labor force. We then calculate how changes in each activity contribute to the total change in the LFPR.
The chart below depicts the drivers of the change in the LFPR from the first quarter of 2016 to the first quarter of 2017. (The interactive tool on our website allows you to make comparisons across gender, age group, and time.) The LFPR rose just slightly (about 0.06 percentage points). However, that small change was the net result of much larger countervailing forces. Other things equal, demographic changes during the year would have lowered the LFPR by around 0.14 percentage points. The aging of the population put significant downward pressure on the LFPR (pushing it down 0.24 percentage points), but a more educated workforce helped push up the LFPR (0.10 percentage points). If the age and education mix of the population had not changed, the LFP rate would have risen by about 0.19 percentage points (see the chart).
The following chart further breaks down the behavioral and cyclical components at work. After controlling for shifts in the demographic mix of the population during the year, the largest contributing factor was a decline in the rate of nonparticipation because of family responsibilities.
This is a particularly important explanation for prime-age women (defined as women between 25 and 54 years of age). A smaller share of prime-age women who say they are busy with home and/or family responsibilities accounts for about half of the 0.62 percentage point increase in LFPR that occurred between the first quarter of 2016 and the first quarter of 2017 (see the chart).
To examine factors affecting prime-age men's participation or to learn more about the cyclical and structural factors behind each reason, visit our website.
July 11, 2017
Another Look at the Wage Growth Tracker's Cyclicality
Though Friday's employment report showed that payroll employment rose by a robust 222,000 jobs in June—much higher than most forecasts—enthusiasm for the news was tempered somewhat by average hourly wages coming in below expectations. Is the (ongoing) relatively tepid pace of wage growth a cause for concern? Perhaps, but the ups and downs of average wages over the course of the business cycle—the pattern of expansion-recession-expansion that typifies modern economies—are a bit more complicated than they may seem.
The year-over-the-year growth in the average wage level that we see in the official employment conditions report is influenced by wages paid to people who were employed either today or a year earlier. That is, the wages of those who remained employed (EE) as well as those who entered employment (NE) and those who exited employment (EN). Because the individuals in these groups may command different wages on average—due to experience, for example—the usual wage growth measures confound the effects of changes in the average wage of people with particular types of year-over-year employment histories. In that sense, the usual wage growth statistic may not exactly be comparing apples to apples.
Research by, for example, Solon, Barsky, and Parker 1992 and Daly and Hobjin 2016 explores the effect of the changing composition of workers over time using microdata on individuals with known employment histories. They show that people who enter and exit employment have a lower average wage than those who stay employed over the year and that the net exit/entry flow increases when the labor market is weak—more people leave employment, and fewer people enter it. As a result, the disproportionate increase in the net flow of workers with a lower-than-average wage serves to boost the overall average wage level during recessions.
One approach to making a more apples-to-apples comparison of average wages over time is to strip out the effect that comes from the change in the share of workers who stay employed and who entered or exited employment. Technically speaking, the composition-adjusted wage growth series is determined by adding the change in average log hourly wage within the EE group and the same change within the EN/NE group, while holding constant the respective average population shares in each group. The chart below illustrates the result of this adjustment.
I should note that the change in the average wage uses data only for people who have a known employment status a year earlier, which results in a wage growth series that is somewhat higher than the change in the average wage of all employed people, some of whom have an unknown employment history.
As the chart shows, relative to the adjusted series (the green line), growth in overall average wages (the orange line) stayed up longer during the last recession, then fell by less, and was slower to adjust to improving labor market conditions (falling unemployment) after the recession ended. The correlation between the overall growth in average wages and the inverse of the unemployment rate is 0.72, and this correlation rises to 0.79 using the adjusted wage growth series.
An alternative approach to making a more apples-to-apples comparison of average wages is to ignore the entry/exit margin and only look at people who are employed both today and a year earlier (EE). The Wage Growth Tracker (computed here as the difference in average log hourly wage) does that for the subset of EE people who have an actual wage record in both periods (no earnings information is collected for self-employed workers in the Current Population Survey). The following chart compares this version of the Wage Growth Tracker with the growth in overall average wages.
The Atlanta Fed's Wage Growth Tracker uses the median change in wages rather than the average change, but it displays very similar dynamics.
As the chart shows, the growth in average wages for those who remain in wage and salary jobs (the red line) is a bit smoother than growth in overall average wages (the orange line) and moves more in sync with the inverse of the unemployment rate (the correlation is 0.85). However, its level is quite a bit higher than growth in overall average wages. This disparity is because the average wage for those entering employment is less than for those exiting, so the change in average wages along the entry/exit margin is always negative.
But enough math—let's put this all together. If you want a measure of wage growth that reflects relative labor market strength, then looking at wage growth after controlling for entry/exit composition effects is probably a good idea. The Wage Growth Tracker seems to do that job reasonably well. However, the Wage Growth Tracker almost certainly overstates the growth in per hour wage costs that employers are facing. Most importantly, it ignores the employment exit/entry margin. Hence, one should avoid interpreting the Wage Growth Tracker as a direct measure of growth in labor costs—a point also discussed in this recent Atlanta Fed podcast episode . The next reading from the Wage Growth Tracker will be available when the Census Bureau releases the Current Population Survey microdata, usually within a couple of weeks of the national employment report. Given that the unemployment rate has remained relatively low recently, I would expect the Wage Growth Tracker to stay at a relatively high level. Check back here then and we'll see what we learn.
May 05, 2017
Slide into the Economic Driver's Seat with the Labor Market Sliders
The Atlanta Fed has just launched the Labor Market Sliders, a tool to help explore simple "what if" questions using actual data on employment, the unemployment rate, labor force participation, gross domestic product (GDP) growth, and labor productivity (GDP per worker).
We modeled the Labor Market Sliders after the popular Atlanta Fed Jobs Calculator. In particular, the sliders take the rate of labor productivity growth and the rate of labor force participation as given (not a function of GDP or employment growth) and then asks questions about GDP growth and labor market outcomes. Like the Jobs Calculator, the sliders require that things add up, a very useful feature for all those backyard economic prognosticators (we know you're out there).
Let's look at an example of using the sliders. The Congressional Budget Office (CBO) projects that the labor force participation rate (LFPR) will maintain roughly its current level of 62.9 percent during the next couple of years, as the downward pressure of retiring baby boomers and the upward pressure from robust hiring hold the rate stable. The CBO also projects that labor productivity growth will gradually increase to almost 1 percent over roughly the same period.
Suppose we want to know what GDP growth would be over the next couple of years (other things equal) if labor productivity, which has been sluggish lately, returned to 1 percent, as projected by the CBO. By moving the Labor Productivity slider in the tool to 1 percent and the Months slider to 24, you will see how productivity alone affects GDP growth: it increases to about 2 percent (see the image below). In this experiment, the unemployment rate, average job growth, and LFPR are constrained to current levels.
However, there's more than one way to achieve GDP growth of 2 percent over the next two years. Let's take a look.
Hit the reset button, and productivity, GDP growth, and months revert to their starting values. Then move the Months slider to 24 and the GDP Growth slider to 2 percent. You then see that—at current levels of labor force participation and labor productivity growth—achieving 2 percent GDP growth over the next two years would require the economy to create about 200,000 jobs per months (see the image below), which would push the unemployment rate down to 3.1 percent (a rate not seen since the early 1950s).
Hit the reset button again. Achieving 2 percent GDP growth over the next two years is also realistic with a higher LFPR, some other things equal. First, move the Months slider to 24, then move the Labor Force Participation Rate slider to 63.7 percent. The higher LFPR is consistent with about 2 percent growth in GDP and roughly 200,000 additional jobs added each month (see the image below). (This scenario constrains the unemployment rate and labor productivity growth rate to their current levels.) Of course, we haven't seen the LFPR at 63.7 percent since 2012, but that's another discussion.
What if we wanted something a bit more ambitious, such as averaging 3 percent GDP growth over the next couple of years? Hit the reset button again, and try this scenario. Keep Labor Force Participation Rate at its current level (consistent with the CBO's projection), set Labor Productivity growth to 1 percent (also using the CBO projection as a guide), move the Months slider to 24, and the GDP Growth slider to 3 percent. The Labor Market Sliders allow us to see that the economy would need to add an average of about 240,000 jobs each month for those two years. This scenario, the tight-labor-market method of achieving 3 percent GDP growth, would bring the unemployment rate down to 2.6 percent.
However, suppose the United States were somehow able to recapture productivity growth of around 2 percent, which we experienced in the late 1990s and early 2000s. In that case, 3 percent GDP could be achieved at the current employment growth and unemployment rate.
I encourage you to play around and devise your own "what if" scenarios—and use the Labor Market Sliders to make sure they add up.
April 11, 2017
Going to School on Labor Force Participation
In the aftermath of the Great Recession, labor force attachment declined. However, that pattern has been reversing itself lately. In particular, the labor force participation rate (LFPR) of the prime-age (25 to 54 years old) population, the core segment of the workforce, has been moving higher since late 2015. While this is good news, the prime-age LFPR remains well below prerecession levels, meaning that there are more than two million fewer prime-age people participating in the labor force. What factors have contributed to that decline? Where did those people go?
The Atlanta Fed LFP dynamics web page has an interactive tool that allows users to drill down into the drivers of the change in LFPR. The tool breaks the change in LFPR into two parts. The first part is the effect of shifts in the share of the population in different age groups (we use five-year age groups). The second part is the change attributable to shifts in the rate of nonparticipation. Using a methodology described here, we can drill deeper into the second part to learn more about the reasons for not participating in the labor force.
The U.S. Census Bureau will make the first quarter 2017 microdata on the reasons for nonparticipation available in a few weeks, so the following chart shows a decomposition of the 1.8 percentage point decline in the prime-age LFPR (not seasonally adjusted) between the fourth quarters of 2007 and 2016.
In this chart, "residual" pertains to the part of the total change in the LFPR that is attributable to the simultaneous shifts in both age-group population shares and age-group participation rates. In the present case, the residual is zero.
Because we are examining changes in prime-age participation, and all age groups within prime-age have reasonably similar participation rates, a change in the composition of ages tends to have little impact on the overall prime-age LFPR. Instead, the decline is due to shifts in the nonparticipation rate within age groups (the orange bar). In particular, the decline could indicate an increased likelihood of being in school, having family responsibilities that prevent participation, being in the shadow labor force (wanting a job but not actively looking), and a disability or poor health.
Although all these factors put downward pressure on participation, an important countervailing influence is that the education level of the population has been rising over time, and participation tends to increase with more education. In 2007, 41.0 percent of the prime-age population had a college degree, and they had an 88.3 percent participation rate versus 79.5 percent participation for those without a degree. By the end of 2016, the fraction with a degree had increased to 47.3 percent, and that cohort's participation rate had declined 1 percentage point, to 87.3 percent, versus a drop of 3.5 percentage points, to 76.0 percent, for those without a degree.
To see the importance of rising education on participation, the following chart shows the decomposition of the 1.8 percentage point decline in prime-age LFPR based on education-group population shares (degree and nondegree) instead of age-group shares.
In this chart, "residual" indicates the part of the total change in LFPR due to the simultaneous shifts in both education-group population shares and education-group participation rates.
As the chart shows, the shift in the education distribution of the prime-age population from 2007 to 2016 by itself would have increased the prime-age participation rate by about 0.7 percentage points (the green bar). Conversely, if education levels had not increased then the participation rate would have decreased by even more than it actually did. The nonparticipation effect would be larger for most nonparticipation reasons and especially for reasons of disability or poor health (−0.8 percentage points versus −0.5 percentage points). See the charts and analysis in the "health problems" section of the Labor Force Dynamics web page for more information on health-related nonparticipation by education.
Despite some partial reversal over the last year and a half, the prime-age LFPR is still lower than it had been prior to the recession. However, the decline in participation could have been even larger if the education level of the population had not also increased. Rising education is associated with a lower incidence of nonparticipation than otherwise would be the case, and it's principally associated with less nonparticipation attributable to disability or poor health. While researchers agree on the positive association between education and health, pinning down the specific reasons for this remains somewhat elusive. Factors such as income, informational, and occupational differences—as well as public policy choices—all play a role. Recent research by Nobel laureate Angus Deaton and Anne Case suggests that both education and racial differences are important considerations—emphasizing the sharply rising incidence of health problems among middle-age, white families with lower levels of education—and this Washington Post article highlights rising disability rates in rural America.
March 20, 2017
Working for Yourself, Some of the Time
Self-employment as a person's primary labor market activity has become much less commonplace in the United States (for example, see the analysis here and here ). This is a potentially important development, as less self-employment may indicate a decline in overall labor market mobility, business dynamism, and entrepreneurial activity (for example, see the evidence and arguments outlined here ).
Recessions can be particularly bad for self-employment, with reduced opportunities for potential business entrants as well as greater difficulty in keeping an existing business going (see here for some evidence on this). However, the rate of self-employment has been drifting lower over a long period, suggesting other factors are also playing a role in the decision to enter and exit self-employment.
One especially troubling development is the decline in the rate of self-employment for those in high-skill service providing jobs (management, professional, and technical services)—the people you might expect to be particularly entrepreneurial. For example, for workers aged 25 to 54 years old, the self-employment rate has declined from 13 percent in 1996 to 9 percent in 2016, and for those 55 years of age or older, the rate has dropped from 27 percent to 19 percent (using data from the Current Population Survey).
Not only are people in high-skill service jobs less likely to be self-employed than in the past, those who are self-employed are also less likely to be working full-time. The fraction usually working full-time has decreased from about 79 percent in 1996 to 74 percent in 2016. (The full-time rate for comparable private sector wage and salary earners has remained relatively stable at around 88 percent.) One possible explanation for the decline in hours worked is the last recession's lingering effects, which made it harder to generate enough work to maintain full-time hours. Another possibility is that more of the self-employed are choosing to work part-time.
It turns out that both explanations have played a role. The following chart shows the percent of part-time self-employment in high-skill service jobs. The blue lines are for unincorporated businesses and the green lines are for incorporated businesses. In order to distinguish cyclical and noncyclical effects, the chart shows the part-time rate for those who want to work full-time but aren't because of slack business conditions or their inability to find more work (part-time for economic reasons, or PTER), and those who work part-time for other reasons (part-time for noneconomic reasons, or PTNER).
In the chart, I classify someone as self-employed when that person's main job is working for profit or fees in his or her own business (and hence it does not capture people whose primary employment is a wage and salary job but are also working for themselves on the side). The self-employed could be sole proprietors or own their business in partnership with others, and the business may assume any of several legal forms, including incorporation. The chart pertains to the private sector, excluding agriculture, and part-time is usually working less than 35 hours a week.
On the cyclical side, the PTER rates (the dotted lines) rose during the last recession and have been slowly moving back toward prerecession levels as the economy has strengthened. In contrast, the PTNER rates (the solid lines) have moved higher since the end of the recession, continuing a longer-term trend. Choosing to work part-time has been playing an increasingly important role in reducing full-time self-employment in high-skill jobs. Note that there is not an obvious long-term trend toward greater PTNER for those self-employed in middle- or low-skill jobs (not shown).
Shifting demographics is one important factor contributing to the decline in average hours worked. In particular, the PTNER rate for older self-employed is much higher than for younger self-employed, and older workers are a growing share of part-time self-employed, a fact that reflects the aging of the workforce overall. (For more on the self-employment of older individuals, see here .) The net result is a rise in the fraction of self-employed choosing to work part-time. The higher rate of PTNER for the older self-employed appears to be mostly because of issues specific to retirement, such as working fewer hours to avoid exceeding social security limits on earnings.
The last recession and a relatively tepid economic recovery reduced the hours that some self-employed people have been able to work because of economic conditions. However, there has also been a longer-term reduction in how many hours other self-employed people (especially those in occupations requiring greater education and generating greater hourly earnings) choose to work. This increased propensity to work only part-time in their business is another factor weighing on overall entrepreneurial activity.
February 28, 2017
Can Tight Labor Markets Inhibit Investment Growth?
One of the most vexing developments of the current expansion has been the long and persistent reduction in the pace of business fixed investment (see chart 1).
The slide in investment spending evident in this chart has had a substantial impact on the pace of gross domestic product (GDP) growth in recent years and is also behind the slow pace of capital accumulation that has been a major factor in the slow labor productivity growth postrecession .
The other notable aspect of chart 1 is that employment growth has been robust during most of the recovery, and that growth remains robust. That sustained performance has taken the economy to the point where measures of labor market performance can be reasonably described as "close to a state of full employment."
Continued strong employment growth could sensibly support a relatively bullish story on investment going forward. As the table below shows, "high-pressure" labor markets—defined as periods when the official unemployment rate falls below the Congressional Budget Office's estimate of the "natural unemployment rate"—tend to be associated with strong levels of business fixed investment spending.
That said, we are taking note of some cautionary sounding from a special question about investment constraints on the most recent Federal Reserve Small Business Credit Survey, whose full results will be released in April. (The Small Business Credit Survey is a collaboration among Federal Reserve Banks and collects information from small businesses throughout the country. The 2016 survey was open from mid-September to mid-December, and generated more than 16,000 responses—about 10,000 of which were from employer firms.)
One of the survey's special questions was the following: What factors constrained your investment decisions over the past 12 months? The respondents were allowed to check as many factors as they deemed relevant and, perhaps not surprisingly, the collective answer was "a lot of things," as chart 2 shows.
Though there are a lot of contenders in that chart, it was interesting to us that the modal response (though admittedly by a hair) was an inability to find or retain qualified staff. It gets even more interesting when you focus on stable, growing firms—those that were profitable in 2016, are increasing payrolls and revenues, and have been in business for at least six years (see chart 3).
For this group—by definition, the most dynamic firms in the sample—perceived constraints on talent acquisition and retention is easily the largest issue when it comes to investment spending headwinds, independent of the size of the firm (measured by annual revenues). Indeed, more than 50 percent of the businesses with revenue in excess of $10 million identified the labor market as a problem.
We want to be sufficiently modest about interpreting these survey results. (The survey's full results will be released in April.) We have only asked this question once and therefore have no ability to compare with historical data. We also don't know for sure if firms truly are being constrained by their ability to find or retain qualified staff, or if respondents were simply identifying with that option as an issue with their business in general. But the idea that business investment could be constrained by access to talent is important for thinking about the growth potential of the economy. The possibility that education and workforce development efforts could have spillover effects into investment growth is intriguing.
February 23, 2017
More Ways to Watch Wages
The Atlanta Fed's Wage Growth Tracker slipped to 3.2 percent in January from 3.5 percent in December. The Wage Growth Tracker for women was 3.1 percent in January, down significantly from what we saw in late 2016, when gains topped 4 percent. For men, the January reading was 3.4 percent, very close to its average for the past 12 months. As I noted last month, I did not think the unusually high female wage growth was sustainable, and that proved to be the case. Since 2009, the Wage Growth Tracker for women has averaged about 0.3 percentage points below that for men—the same as the gap in the latest data.
Understanding why the Wage Growth Tracker slowed last month highlights the importance of being able to look beyond the top-line number. To provide Wage Growth Tracker users with more information, we have now added several additional cuts of the data to the Wage Growth Tracker web page. The amount of detail we can provide is limited by sample size considerations, and as a result, the additional data are reported as 12-month moving averages. The new data provide more detailed age, race, education, and geographic comparisons, as well as comparisons across broad categories of occupation, industry, and hours worked. As an example, here is a look at the (12-month average) median wage growth data for those who usually work full-time versus those who usually work part-time.
Have fun with these new tools, and we encourage you to comment and let us know what you think.
February 21, 2017
Unemployment versus Underemployment: Assessing Labor Market Slack
The U-3 unemployment rate has returned to prerecession levels and is close to estimates of its longer-run sustainable level. Yet other indicators of slack, such as the U-6 statistic, which includes people working part-time but wanting to work full-time (often referred to as part-time for economic reasons, or PTER), has not declined as quickly or by as much as the U-3 unemployment rate.
If unemployment and PTER reflect the same business-cycle effects, then they should move pretty much in lockstep. But as the following chart shows, such uniformity hasn't generally been the case. In the most recent recovery, unemployment started declining in 2010, but PTER started to move substantially lower beginning only in 2013. The upshot is that for each unemployed worker, there are now many more involuntary part-time workers than in the past.
Regarding the above chart, I should note that I adjusted the pre-1994 data to be consistent with the 1994 redesign of the Current Population Survey from the U.S. Bureau of Labor Statistics (see, for example, research from Rob Valletta and Leila Bengali and Anne Polivka and Stephen Miller ). This adjustment amounts to reducing the pre-1994 number of PTER workers by about 20 percent.
The elevated level of PTER workers has been most pronounced for workers in low-skill occupations. As shown in the next chart, PTER workers in low-skill jobs now outnumber unemployed workers who left low-skill jobs. Prior to the most recent recession, low-skill unemployment was always higher than low-skill PTER.
The increase in PTER workers is also mostly in the retail trade industry, as well as the leisure and hospitality industry, where low-skill occupations are concentrated. The PTER-to-unemployment ratio for the goods-producing sector (manufacturing, construction, and mining) has remained essentially unchanged. In those industries, unemployment and PTER move together.
Some researchers, such as our colleagues at the San Francisco Fed Rob Valletta and Catherine van der List, have argued that the increase in the prevalence of involuntary part-time work relative to unemployment suggests the importance of factors other than overall demand for labor. Among these factors are shifting demographics (a greater number of older workers who are less willing to do part-time work) and industry mix (more employment in industries with higher concentrations of part-time jobs). Such factors are almost certainly playing a role.
Recent analysis by Jon Willis at the Kansas City Fed suggests that the elevated levels of PTER in low-skill occupations may reflect that during the last recession, firms reduced the hours of workers in low-skill jobs more than they cut the number of low-skill jobs. In other words, firms still had some work that needed to get done, probably with peak demand at certain times of the day, and those tasks couldn't readily be outsourced or automated.
As the following chart from Willis's research shows, between 2007 and 2010, low-skill (non-PTER) employment actually increased slightly overall, but the mix of employment shifted dramatically toward part-time.
Since the recession, the pace of (non-PTER) low-skill job creation has been modest (about 20,000 jobs per month compared with 60,000 jobs per month in the years preceding the recession). Initially, this trend helped reduce low-skill unemployment more than the incidence of PTER—one reason why the ratio of PTER to unemployment continued to increase.
But the number of PTER workers in low-skill jobs has since been declining as more people have been able to find full-time jobs. At the current pace of job creation and (net) transition rates out of PTER, Willis estimates it would take until 2020 to return to prerecession levels of low-skill PTER. That seems a reasonable guess to me.
- GDPNow's Forecast: Why Did It Spike Recently?
- How Low Is the Unemployment Rate, Really?
- What Businesses Said about Tax Reform
- Financial Regulation: Fit for New Technologies?
- Is Macroprudential Supervision Ready for the Future?
- Labor Supply Constraints and Health Problems in Rural America
- Building a Better Model: Introducing Changes to GDPNow
- How Ill a Wind? Hurricanes' Impacts on Employment and Earnings
- When Health Insurance and Its Financial Cushion Disappear
- What Is the "Right" Policy Rate?
- February 2018
- January 2018
- November 2017
- October 2017
- September 2017
- August 2017
- July 2017
- May 2017
- April 2017
- March 2017
- Business Cycles
- Business Inflation Expectations
- Capital and Investment
- Capital Markets
- Data Releases
- Economic conditions
- Economic Growth and Development
- Exchange Rates and the Dollar
- Fed Funds Futures
- Federal Debt and Deficits
- Federal Reserve and Monetary Policy
- Financial System
- Fiscal Policy
- Health Care
- Inflation Expectations
- Interest Rates
- Labor Markets
- Latin America/South America
- Monetary Policy
- Money Markets
- Real Estate
- Saving, Capital, and Investment
- Small Business
- Social Security
- This, That, and the Other
- Trade Deficit
- Wage Growth