The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.
Comments are moderated and will not appear until the moderator has approved them.
Please submit appropriate comments. Inappropriate comments include content that is abusive, harassing, or threatening; obscene, vulgar, or profane; an attack of a personal nature; or overtly political.
In addition, no off-topic remarks or spam is permitted.
January 16, 2020
Do Higher Wages Mean Higher Standards of Living?
Editor's note: We have updated macroblog's location on our website, although archival posts will remain at their original location. Readers who use RSS should update their feed's URL to https://www.frbatlanta.org/rss/macroblog.aspx. Also, we are implementing a new commenting system for posts. In the comments section at the end of this post, you'll be able to create an account to leave comments.
A recent macroblog post used Atlanta Fed Wage Growth Tracker data to observe that the hourly wage of the lowest-paid workers has rebounded in recent years after declining for a decade. The chart below depicts this finding, showing the median hourly wage of the lowest-paid 25 percent of workers in the Tracker sample relative to the median for all workers.
Moreover, the post showed that this recovery was not just a story about states and localities increasing their minimum wages. It also appears that there has been a significant tightening in the labor market for unskilled or low-skilled jobs.
Taken at face value, this is good news for workers employed in low-wage jobs. But here's the rub: the median wage in the first quartile is still low—$11.50 in 2019, or 55 percent of the overall median wage. Moreover, these are hourly wages before taxes and transfers (we'll get back to this shortly). They don't represent what is happening to these workers' ability to make ends meet, which depends crucially on income after taxes and transfers.
For households at the bottom of the income distribution, means-tested transfers can play an especially important role. Means-tested transfers—cash payments and in-kind benefits from federal, state, and local governments designed to assist individuals and families with low incomes and few assets to meet their basic living needs—represent about 70 percent of income before taxes and transfers for households in the bottom quintile of the income distribution, according to a recent report by the Congressional Budget Office. However, the size of the transfers tends to decrease as earnings increase, and they stop altogether when a worker exceeds income- and asset-eligibility thresholds.
The interaction between changes in earnings and various means-tested public assistance programs is an important public policy issue, and it is one that staff at the Atlanta Fed are studying. In a March 2019 macroblog post, David Altig and Laurence Kotlikoff reported that this interaction results in low-income households facing a higher median effective marginal tax rate than high-income households. For low-income households with children, this effect can be especially severe because the presence of children increases the value of programs such as the Supplemental Nutrition Assistance Program (or SNAP, formerly known as the food stamp program) and the likelihood of enrollment in additional programs such as federally subsidized child care. (You can read further research on the effective or implicit marginal tax rates of low-income households at Congressional Budget Office (2016), Romich and Hill (2018), and Chien and Macartney (2019).)
To illustrate the point, the Atlanta Fed team studied the case of a hypothetical single mother with two young children who works in a near-minimum-wage, full-time job and whose basic living expenses are helped by various transfer programs. One avenue to improving her family's standard of living is if she were to return to school and pursue a higher-paying career as a nurse. Over the long term, the net gains from education and career advancement are unambiguous. However, the Atlanta Fed's analysis shows that as long as her children still require care, the reduction in payments from various benefit programs could partially or even completely offset the gains. Look for an Atlanta Fed paper discussing this very real dilemma coming soon on the Bank's Economic Mobility and Resilience webpage.
What do findings like this mean for interpreting the Wage Growth Tracker's evidence that people in the bottom part of the wage distribution are experiencing relatively larger wage gains? Perhaps there is a bit less to celebrate than meets the eye. Around 46 percent of these individuals are in households with children. To the extent that they also participate in means-tested public assistance programs, the relative increase in their family's standard of living could be much less than the size of their pay raise would suggest.
December 16, 2019
Faster Wage Growth for the Lowest-Paid Workers
On November 25, Fed chair Jay Powell gave a speech titled "Building on the Gains from the Long Expansion," in which he observed that
Recent years' data paint a hopeful picture of more people in their prime years in the workforce and wages rising for low- and middle-income workers.
In making this point, Chair Powell used a cut of the Atlanta Fed's Wage Growth Tracker that looks at the median annual wage growth of workers in the lowest 25 percent of the wage distribution. As the following chart shows, the lowest-paid workers have been experiencing higher median wage growth (the blue line) in the last few years than workers overall (the green line). This reverses the pattern seen in the wake of the Great Recession, when median wage growth for lower-paid workers slowed by more than for workers overall.
The faster median wage growth for lower-wage workers shown in chart above has also translated into an increase in the relative median wage level of these workers. To see this, the following chart shows the median wage level for those in the lowest wage quartile relative to the median for all workers in the Wage Growth Tracker dataset.
The chart shows that for workers in lower-wage jobs, their relative median wage over the 2000s has deteriorated, and that erosion has reversed course only in the last few years. This reversal may reflect increasing tightness of the labor market for lower-wage jobs relative to other jobs over the last few years. The challenge of filling jobs requiring few skills is something we have been hearing about a lot recently from the businesses we talk to (for example, see here), and this sort of challenge could be behind higher wages for those workers. However, several state and local governments have increased the minimum wage in recent years, which would also push up the relative pay for those in the lowest-paid jobs.
Are the observations in the previous chart solely attributable to minimum wage increases? To get some idea, the next chart contrasts the relative median wage in states that increased their minimum wage at some point between 2014 and 2019 to those that did not. The blue line is the relative median wage of the lowest quartile in the 28 states that increased their minimum wage (23 states introduced new minimum wage levels, and five implemented increases legislated before 2014), and the green line is relative median wage for the states that did not increase their minimum wage.
We would expect to see a rise in the relative median wage in the states that raised their minimum wage, and indeed we do. For the group of states that increased minimum wages (the blue line), the relative median wage is now closer to that of states that did not increase their minimum wage (the green line). Interestingly, though, even in the "no increase" states, the relative median wage has improved, suggesting that the increased tightness of labor markets, or some other factor than hikes in state minimum wages, is playing a role in pushing up the pay for those in lower-wage jobs. Consistent with the message of Chair Powell's speech, the good news is that there is scope to continue to build on the gains from the long and ongoing expansion for workers at the bottom end of the wage distribution.
November 25, 2019
Is Job Switching on the Decline?
Here's a puzzle. Unemployment is at a historically low level, yet nominal wage growth is not even back to prerecession levels (see, for example, the Atlanta Fed's own Wage Growth Tracker). Why is wage growth not higher if the labor market is so tight? A recent article in the Wall Street Journal posited that the low rate of job-market churn likely explains slow wage growth. Switching jobs is typically lucrative because it tends to be going to a job that better uses the person's skills and hence offers higher pay. Job switchers can also help improve the bargaining position of job-stayers by inducing employers to pay more to retain them.
But is the job-switching rate really lower? A paper that Shigeru Fujita, Guiseppe Moscarini, and Fabien Postel-Vinay presented at the Atlanta Fed's 10th annual employment conference looked at a commonly used measure of employer-to-employer transitions. That measure, developed by Fed economists Bruce Fallick and Charles Fleischman in 2004, uses data from the Current Population Survey (CPS) on whether a person says that he or she has the same employer this month as last month. Job switchers are those reporting having a different employer. As the following chart shows, the Fallick and Fleischman data (the yellow line) support the Wall Street Journal story that the rate of job switching is much lower than it used to be.
Source Fujita et al. (2019)
However, Fujita and his coauthors discovered a potential problem with these data, noting that the CPS doesn't ask the same-employer question of all surveyed people who were employed in the prior month. Importantly, the incidence of missing answers has increased dramatically since the 2006, as the following chart shows.
Source Fujita et al. (2019)
If these missing answers were merely randomly distributed among job switchers and job stayers, then it wouldn't matter much for the Fallick-Fleischman measure. But Fujita et al. found that the missing answers were disproportionately from people who look more like job switchers in terms of observable characteristics such as age, marital status, education, industry, and occupation—making it likely that the Fallick-Fleischman measure undercounts job switchers.
The researchers developed a statistical adjustment to the Fallick-Fleischman measure to account for this bias (the blue line label our series in the following chart—this is the same as the first chart) that tells a somewhat different story than the original measure (yellow line).
Source Fujita et al. (2019)
In particular, the adjusted job-switching rate is only moderately lower than it was 20 years ago and has fully recovered from the decline experienced during the Great Recession. Although a decline in job switching might be a factor in the story behind low wage growth, based on this adjusted measure it doesn't seem like the dominant factor. The low wage growth puzzle remains a puzzle.
September 26, 2019
Digging into Older Americans' Flat Participation Rate
The rate of labor force participation (LFP) by people age 55 and over had been rising during the decade leading up to the Great Recession. But more recently, as the following chart shows, the share of older individuals engaged in the labor market has barely budged. (We should note that for all the charts in this post, the data are from the Current Population Survey from the U.S. Bureau of Labor Statistics and the authors' calculations.) What behavioral and demographic factors could be underlying these trends?
We can use the Atlanta Fed's Labor Force Dynamics web page to explore why the 55-and-over population has a relatively flat LFP rate of late. One factor working to depress older Americans' LFP rate is the increase in the share of those 55 and older who are retired: from 46.9 percent in the second quarter of 2014 to 47.8 percent in the second quarter of 2019 (see the chart).
This change in the overall retirement rate of the 55-plus cohort is actually the result of the union of two opposite forcesone demographic and one behavioral. From the perspective of demographics, a greater share of the population is reaching the typical retirement age threshold of 65. For instance, the share of people aged 65 and older has increased from 18 percent in the second quarter of 2014 to 20.3 percent in the second quarter of 2019 (see the chart).
This demographic shift is important because the retirement rate is much higher for those 65 and older compared to those from 60 to 64 years old. For example, the retirement rate in 2019 for those 60–64 is around 25 percent, but it's 56 percent for those aged 65–69, and it's 81 percent for those 70 and older (see the chart).
However, the retirement rate among those 60–64 and those 65–69 has also declined in recent years. This change in behavior within older age groups is partly offsetting the downward pressure on participation coming from having a larger share of population over 65.
A second factor that has worked against the overall retirement effect and helped push up the LFP rate of the 55-plus population has been a decline in the share of older individuals saying they are not participating in the labor force because of disability or illness. This rate has decreased from 8.6 percent in the second quarter of 2014 to 8.1 percent in the second quarter of 2019 (see the chart).
As the chart below shows, this shift is largely a demographic effect. Nonparticipation because of disability/illness drops off significantly as people turn 65, so the fact that a greater share of people 55 and older are now 65 and older decreases the overall share of nonparticipation for disability/illness reasons as well. However, it's important to note that a lower disability/illness rate for those 65 and older doesn't mean that this older population is actually less disabled or sick. It is more that individuals in the data fall into only one category, and those 65 and older are more likely to say they are retired than to say they are disabled or ill.
A third force helping push up the LFP rate is the decline in the share of older individuals on the sidelines of the labor market—those who are not participating but nonetheless want a job—from 1.8 percent in the second quarter of 2014 to 1.6 percent in the second quarter of 2019 (see the chart).
This “shadow labor force” effect on participation is mostly the result of changes in behavior (that is, a reduced propensity to remain on the sidelines within age groups) rather than from demographic changes because the rate has declined within age groups, whereas the levels are roughly similar across age groups (see the chart).
To put these various pieces together, the following chart summarizes the overall contributions of demographic and behavioral forces on the LFP rate among the 55-plus population between the second quarter of 2014 and the second quarter of 2019. The chart shows that the contributions stemming from changes in demographics and behavior have largely offset each other.
As I've already described, the biggest demographic effects come from having more people at an age with two specific characteristics: a relatively high rate of nonparticipation because of retirement and a relatively low rate of nonparticipation because of disability or illness. But as the following chart shows, from the second quarter of 2014 to the second quarter of 2109, the retirement demographic dominates, so the overall demographic LFP effect is a negative one.
Conversely, the largest behavioral shifts are, first, a lower propensity within older age groups to stay out of the labor force because of retirement and, second, a lower share of older people wanting a job but not looking for one. As the following chart shows, from the second quarter of 2014 to the second quarter of 2019, these behavioral changes combine to push up LFP by enough to nearly offset the demographic shifts.
It seems reasonable to presume that the aging population will continue to be an important source of downward pressure on the LFP rate of older Americans over the next few years. What will be telling is whether or not the behavioral shifts we have seen will persist as strongly as they have up to this point and continue to provide a countervailing positive influence on participation.
The tools on the Atlanta Fed's Labor Force Dynamics web page are very useful for understanding what's behind changes in LFP for different demographic groups. In addition to cuts for different age groups, you can look at differences between men and women and among different racial and ethnic groups and levels of educational attainment. You can download the chart data—the charts in this blog are downloaded images from the web page—and even download the underlying Current Population Survey microdata from the Kansas City Fed's CADRE web page if you want to create your own cuts. Check it out!
- 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 AmericaSouth America
- Monetary Policy
- Money Markets
- Real Estate
- Saving Capital and Investment
- Small Business
- Social Security
- This That and the Other
- Trade Deficit
- Wage Growth