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
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 22, 2017
GDPNow's Second Quarter Forecast: Is It Too High?
Real gross domestic product (GDP) growth slowed from a 2 percent pace in 2016 to an annual rate of 0.7 percent in the first quarter of 2017. The Federal Open Market Committee viewed this slowdown in growth "as likely to be transitory," according to its last statement.
Indeed, current quarter GDP forecasting models maintained by the Federal Reserve Banks of New York, St. Louis, and Atlanta have been pointing toward stronger second quarter growth (2.3 percent, 2.6 percent and 4.1 percent, as reported on their respective websites on May 19, 2017).
The Atlanta Fed's model—GDPNow—is at the high end of this range and is also high relative to other professional forecasts. The median forecast for second quarter real GDP growth in the May Survey of Professional Forecasters (SPF) was 3.1 percent, for instance, and recent forecasts from Blue Chip Publication surveys displayed on our GDPNow page show some divergence from our model as well.
We encourage—and frequently receive—feedback on our GDPNow tool, and some users have suggested that our forecast for second quarter growth is too high. In fact, some empirical evidence supports that view. The evidence considered here correlates differences between consensus Blue Chip Economic Indicators Survey and GDPNow forecasts for growth about 80 days before the first GDP release with the GDPNow forecast errors (see the chart below).
A note about the chart: The horizontal axis shows the difference between the Blue Chip consensus forecasts and GDPNow's forecast. The vertical axis measures the 80-day-ahead GDPNow forecast error, defined as the difference between the first published estimate of real GDP growth and the GDPNow forecast at the time of the mid-quarter Blue Chip survey.
As the chart shows, there is a positive relationship between the Blue Chip-GDPNow discrepancy and the GDPNow forecast error. A simple linear regression would predict that the GDPNow forecast of 3.7 percent growth on May 5 was too high by nearly 1.0 percentage point. Moreover, the chart suggests that there has been a bias in GDPNow forecasts since the fourth quarter of 2015 of between 0.9 and 2.0 percentage points at the time of these mid-quarter Blue Chip surveys. If you are inclined to think the GDPNow forecast for second quarter growth is a bit too high, then this evidence will not change your mind.
Given this evidence, you might think that putting relatively little stock in the GDPNow forecast at this point in the quarter would be prudent. Indeed, if we calculate the weighted average of the historical Blue Chip consensus and GDPNow forecasts that produced the most accurate forecast of the first estimate of real GDP growth, then the optimal weight of the GDPNow forecast lies somewhere between 0.34 and 0.55 (see the chart below). The weight depends on the number of days until the first GDP release.
For example, the optimal weight of 0.55 on GDPNow about 54 days before the first GDP release means that 0.55 times the GDPNow forecast plus 0.45 times Blue Chip consensus survey forecast has been more accurate, on average, than any other weighted average of the two forecasts. The lowest weight on GDPNow corresponds to forecasts made about 83 days before the first GDP release—the time when GDPNow's bean-counting algorithms have the least amount of source data to work with.
A weighted average of the Blue Chip consensus and GDPNow forecasts at that time would put the GDP forecast about 0.6 to 0.7 percentage points below the current GDPNow forecast. However, the confidence bands around these estimates are wide, so the positive weight placed on GDPNow early in the quarter could just be the result of chance.
Let's cut to the chase—why, exactly, is the GDPNow forecast for second quarter GDP growth so high? The details of the GDPNow forecast provide some clues. We can compare the GDPNow forecasts of GDP components with those from the SPF. (The Blue Chip forecast does not provide detail on all the GDP components.) The following table translates the median SPF forecasts into contributions to second quarter real GDP growth. These contributions are shown alongside GDPNow's forecasted contributions as well as the average contributions to real GDP growth over the prior four quarters.
Clearly, more than half of the difference between the GDP growth forecasts from GDPNow and the SPF is due to inventories. For both forecasts, inventory investment also accounts for over half of the pickup in second quarter growth from the trailing four-quarter average.
A macroblog post I wrote last year showed that the growth-forecast contribution of mid-quarter inventory investment produced roughly equivalent accuracy in the SPF and GDPNow models, but it was much less accurate than the contribution forecasts of the other GDP components. Based on experience, we can't be confident that either forecast of inventory investment is likely to be very accurate or that one is likely to be much more accurate than another.
With very little hard data in hand for the second quarter for most of the GDP components—and for inventories in particular—we will continue to closely monitor if the data are as strong as GDPNow is anticipating or if they hew more closely to other forecasts. Check back with us to see.
- Behind the Increase in Prime-Age Labor Force Participation
- An Update on Labor Force Participation
- Another Look at the Wage Growth Tracker's Cyclicality
- GDPNow's Second Quarter Forecast: Is It Too High?
- Are Small Loans Hard to Find? Evidence from the Federal Reserve Banks' Small Business Survey
- Slide into the Economic Driver's Seat with the Labor Market Sliders
- The Fed’s Inflation Goal: What Does the Public Know?
- Going to School on Labor Force Participation
- Bad Debt Is Bad for Your Health
- Working for Yourself, Some of the Time
- July 2017
- May 2017
- April 2017
- March 2017
- February 2017
- January 2017
- December 2016
- November 2016
- October 2016
- September 2016
- 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