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
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 forces—one 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!
August 05, 2019
What the Wage Growth of Hourly Workers Is Telling Us
The Atlanta Fed's Wage Growth Tracker has shown an uptick during the past several months. The 12-month average reached 3.7 percent in June, up from 3.2 percent last year. But in 2016, it depicted acceleration that eventually reversed course. So is this recent increase real or illusory?
Although using a 12-month average quiets much of the noise in the monthly data, it is possible that the smoothed series still may exhibit some unwanted variation due to the way the Wage Growth Tracker is constructed. For example, how the monthly Current Population Survey reports individual earnings might be a factor introducing unwanted noise into the Tracker. Specifically, some people directly report their hourly rate of pay, and some report their earnings in terms of an amount per week, per month, or per year.
Relative to those paid an hourly rate, there are at least a couple of reasons why using the earnings of nonhourly workers might introduce additional variability into the Wage Growth Tracker's overall estimate of wage growth. First, reported nonhourly earnings include base pay as well as any overtime pay, tips, and commissions earned, and hence can vary over time even if the base rate of pay didn't change. For a worker paid at an hourly rate, reported earnings exclude overtime pay, tips, and commissions and so are not subject to this source of variation. Second, the method we use to convert nonhourly earnings to an hourly rate is likely subject to some margin of error since it involves using the person's recollection of how many hours they usually work. These two factors suggest that the earnings of workers paid at an hourly rate might be a somewhat cleaner measure of hourly earnings.
To investigate whether this distinction actually matters in practice, we created the following chart comparing the 12-month average Wage Growth Tracker since 2015 (depicted in the green line) with a version that uses only the earnings of those paid at an hourly rate (blue line).
As the chart shows, the 12-month average of median wage growth for hourly workers generally tracks the overall series—both series are about a percentage point higher than at the beginning of 2015. However, the hourly series is a bit less variable, making the recent uptick in wage growth more noticeable in the hourly series than in the overall series. This observation suggests that as we monitor shifts in wage pressure, the hourly series could complement the overall series nicely. Versions of the Wage Growth Tracker series for both hourly and nonhourly workers are now available on the Wage Growth Tracker page of the Atlanta Fed's website.
If you would like to use the Wage Growth Tracker's underlying microdata to create your own versions (or to conduct other analysis), follow this link to explore the data on the Atlanta Fed's website. See this macroblog post, "Making Analysis of the Current Population Survey Easier," from my colleague Ellyn Terry to learn more about using this dataset.
May 06, 2019
Improving Labor Force Participation
Without question, the U.S. labor market has tightened a lot over the last few years. But a shifting trend in labor force participation—and especially a rise in the propensity to seek employment by those in their prime working years—seems to be relieving some labor market pressure.
From the first quarter of 2015 to the first quarter of 2019, the labor force participation (LFP) rate among prime-age workers (those between 25 and 54 years old) increased by about 1.5 percentage points (see the chart below), adding about 2 million workers more than if the participation rate had not increased.
Changes in the distribution of the prime-age population in terms of age, education, and race/ethnicity toward groups with higher participation rates and away from groups with lower rates accounts for about a third of the rise in the overall prime-age LFP rate. The other two thirds can be pinned on an increase in LFP rates within demographic groups—what we call "behavioral" effects.
Of the increased participation behavior within demographic groups, there has been a decline in the share of the prime-age population that say they want a job but are not actively looking for work at the moment. We refer to these individuals as the "shadow labor force" because even though they are not in the labor force this month, they have a relatively high propensity to have a job next month. Second, there's been a decline in the share of the prime-age population that are not participating because they are too sick or disabled to work. The contribution of the change in behavior in these two categories (as well as several others from the first quarter of 2015 to the first quarter of 2019) are shown in the following chart, which is taken from the Atlanta Fed's Labor Force Participation Dynamics tool.
In contrast, consider the period from the first quarter of 2008 through the first quarter of 2015, a time when the rate of prime-age LFP declined by almost 2 percentage points. During that period, even though slow-moving demographic changes were putting modest upward pressure on the prime-age participation rate, that support was more than swamped by negative changes in participation rates within demographic groups. The following chart shows the relative contributions of these behavioral changes.
Within demographic groups, the increased incidence of being too sick or disabled to work stands out as the largest contributor to the decline in prime-age labor force participation between 2008 and 2014.
Since 2014, prime-age LFP has benefited from the movement of both demographics and participation behavior. But so far, less than half of the overall behavioral decline between 2008 and 2014 has been reversed.
Though demographic trends are likely to remain positive, how much more participation behavior—especially as it is related to disability and illness—can shift as the labor market tightens remains unclear. The share of the prime-age population too sick or disabled to work had been on a rising trend for the decade prior to the last recession, suggesting that there may be some deeper and structural health-related issues that could keep the disability/illness rate elevated despite an increasingly tight labor market.
March 26, 2019
Young Hispanic Women Investing More in Education: Good News for Labor Force Participation
In a recent recent macroblog post, my colleague John Robertson found that the recent rise in female prime-age (ages 25 to 54 years) labor force participation (LFP) over the last few years has been driven in large part by increased participation among Hispanic women. (Hispanic refers to people of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race.) Much of the LFP improvement among Hispanic women has come as they've shifted away from household duties.
To understand this development and determine whether it's a trend likely to continue, we look at trends in the activities of younger Hispanic women. In particular, we look at the so-called NEETs rate among women ages 16 to 24. The NEETs rate is the share of the youth population that is "Not Employed or pursing Education or Training." This group is sometimes referred to as "disconnected youth" or "opportunity youth" because they are generally less likely to be attached to the labor force as they move into their prime working years and are at higher risk of experiencing long-term unemployment, persistent poverty, poor health, and criminal behavior.
A look at the next chart shows substantial improvement in the NEETs rate among young Hispanic women over the last two decades. The gap has narrowed considerably and in recent years has tracked much more closely with black non-Hispanic women.
The declining NEETs rate for young Hispanic women primarily reflects shifting preferences toward more education and away from household responsibilities. As you can see in the next chart, the share of young Hispanic women who are in education or training has risen over the last two decades, up nearly 19 percentage points since 2000. Their share now more closely matches that of young black and white non-Hispanic women.
Mirroring the rise in educational activities has been a shrinking share of young Hispanic women who are not in the labor force because they are taking care of home or family, as the following chart shows.
Young Hispanic women have invested increased time in their education over the last two decades and as a result have higher average levels of educational attainment than earlier cohorts moving into their prime working years. To see this, the next chart shows the distribution of educational attainment over time for Hispanic women aged 25.
The higher levels of LFP in recent years among prime-age Hispanic women partly reflects the greater investment in education by younger Hispanic women. If this trend continues—and there is no obvious reason why it wouldn't—then it will help drive even higher labor force attachment for prime-age Hispanic women in the years to come.
March 22, 2019
A Different Type of Tax Reform
Two interesting, and important, documents crossed our desk last week. The first was the 2019 edition of the Economic Report of the President. What particularly grabbed our attention was the following statement from Chapter 3:
Fundamentally, when people opt to neither work nor look for work it is an indication that the after-tax income they expect to receive in the workforce is below their "reservation wage"—that is, the minimum value they give to time spent on activities outside the formal labor market.
That does not strike us as a controversial proposition, which makes the second of last week's documents—actually a set of documents from the U.S. Department of Health and Human Services (HHS)—especially interesting.
In that series of documents, HHS's Nina Chien and Suzanne Macartney point out a couple of things that are particularly important when thinking about the effect of tax rates on after-tax income and the incentive to work. The first, which is generally appreciated, is that the tax rates that matter with respect to incentives to work are marginal tax rates—the amount that is ceded to the government on the next $1 of income received. The second, and less often explicitly recognized, is that the amount ceded to the government includes not only payments to the government (in the form of, for example, income taxes) but also losses in benefits received from the government (in the form of, for example, Medicaid or child care assistance payments).
The fact that effective marginal tax rates are all about the sum of explicit tax payments to the government and lost transfer payments from the government applies to us all. But it is especially true for those at the lower end of the income distribution. These are the folks (of working age, anyway) who disproportionately receive means-tested benefit payments. For low-wage workers, or individuals contemplating entering the workforce into low-wage jobs, the reduction of public support payments is by far the most significant factor in effective marginal tax rates and the consequent incentive to work and acquire skills.
The implication of losing benefits for an individual's effective marginal tax rate can be eye-popping. From Chien and Macartney (Brief #2 in the series):
Among households with children just above poverty, the median marginal tax rate is high (51 percent); rates remain high (never dipping below 45 percent) as incomes approach 200 percent of poverty.
Our own work confirms the essence of this message. Consider a representative set of households, with household heads aged 30–39, living in Florida. (Because both state and local taxes and certain transfer programs vary by state, geography matters.) Now think of calculating the wealth for each household—wealth being the sum of their lifetime earnings from working and the value of their assets net of liabilities—and grouping the households into wealth quintiles. (In other words, the first quintile would the 20 percent of households with the lowest wealth, the fifth quintile would be the 20 percent of households with the highest wealth.)
What follows are the median effective marginal tax rates that we calculate from this experiment:
Median Effective Marginal Tax Rate
Consistent with Chien and Macartney, the median effective marginal tax rates for the least wealthy are quite high. Perhaps more troubling, underlying this pattern of effective tax rates is one especially daunting challenge. The source of the relatively high effective rates for low-wealth individuals is the phase-out of transfer payments, some of which are so abrupt that they are referred to as benefits, or fiscal, cliffs. Because these payments differ widely across family structure, income levels within a quintile, and state law, the marginal tax rates faced by individuals in the lower quintiles are very disparate.
The upshot of all of this is that "tax reform" aimed at reducing the disincentives to work at the lower end of the income scale is not straightforward. Without such reform, however, it is difficult to imagine a fully successful approach to (in the words of the Economic Report) "[increasing] the after-tax return to formal work, thereby increasing work incentives for potential entrants into the labor market."
March 06, 2019
X Factor: Hispanic Women Drive the Labor-Force Comeback
The share of the prime-age population engaged in the U.S. labor market is on the rise, led by a sharp rebound in the labor force participation (LFP) rate by prime-age female workers (those ages 25–54). This point was highlighted in a recent Wall Street Journal article.
Since 2015 the LFP rate for prime-age women has increased by about 1.8 percentage points, reversing an almost 16-year slide. Using the data underlying the Atlanta Fed’s new Labor Force Participation Dynamics tool, some of the factors behind this increase become apparent. Of particular note is that Hispanics (people of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race) account for a bit less than one-fifth of the prime-age female population, but they accounted for almost two-thirds of the increase in female LFP during the last three years. This increase was the result of both a rising share of the population that is Hispanic and the rising LFP rate among Hispanic women. The share of the prime-age, Latina population increased by 1 percentage point between 2015 and 2018, and the LFP rate for this group increased 3 percentage points. (The reason why rising Hispanic LFP didn’t result in the overall female LFP rate increasing by more than 1.8 percentage points is because Hispanic women are still 8 percentage points less likely to be in the labor force than non-Hispanic women. But this participation gap is closing rapidly.)
The Atlanta Fed’s web tool also allows us to further explore what is behind the 3 percentage point LFP rate increase for prime-age Latinas in the last three years. (My Atlanta Fed colleague Ellyn Terry provides a longer-term view on Hispanic female labor force dynamics in this related macroblog post.) It’s particularly noteworthy that almost two-thirds of the recent increase is the result of a decline in family or household responsibilities keeping people out of the labor force (see the chart).
This shift away from household duties is attributable to a combination of the shifting demographics of the Hispanic population (such as being more likely to have a college degree and thus obtaining a higher-wage job and being better able to afford child care) and a lower propensity to not participate for family reasons within Hispanic age and education groups.
The rebound in female LFP in the last three years is good news, with rising wages, particularly at the low end, and higher demand in traditionally female-dominated occupations contributing to the increase. But making the labor market a truly viable option for women still poses a number of challenges. The LFP rate of U.S. women has fallen behind that of many other countries, many of which have enacted family-friendly policies to help support women in the workplace.
February 14, 2019
Trends in Hispanic Labor Force Participation
Although the labor force participation (LFP) rate has fallen significantly for the overall population during the past two decades, the trends can differ a great deal depending on which demographic group you examine. One way to view these varied, ever-changing patterns is to use the Atlanta Fed's Labor Force Participation Dynamics tool. We recently redesigned the tool, adding a new interface and more options for understanding specific demographic groups' LFP. The tool also allows users to see what factors (such as disability/illness, being in school, retirement, or family responsibilities) influence changes in the LFP rate for different groups.
While the tool shows us many stories, a particularly interesting one is the experience of people of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race (henceforth referred to as Hispanic). Hispanics are a growing share of the U.S. population (the U.S. Bureau of Labor Statistics [BLS] projects that nearly a fifth of the people in the labor force will be Hispanic by 2024, up from a tenth in 1994), and therefore Hispanics' differences in attitudes and preferences for work will exert an increasingly great effect on headline LFP numbers.
Let's parse the LFP rate among Hispanics. First, the Hispanic population is more likely to engage in the labor force than non-Hispanics. (In 2017, the LFP rate of Hispanics was 66.1 percent, compared to 62.2 percent for non-Hispanics). Second, their LFP has fallen by less during the past two decades.
The pictures below come from the redesigned tool. The first two charts compare the decline in the LFP rate for all ethnicities (Chart 1) versus Hispanic (Chart 2) as the combination of six nonparticipation categories. Each colored bar represents how much a particular category of nonpartipation has changed since the fourth quarter of 1998. The red line shows the summation of the change in each nonparticipation category, or the net change in the LFP rate. For example, the LFP rate overall has declined 4.2 percentage points (ppts) during the past two decades. However, among Hispanics, it has fallen significantly less—just 1.0 ppt. A comparison of the size and direction of each of the nonparticipation categories between the two charts shows many differences in the factors affecting the decline in the LFP rate of each group.
Because differences across ethnicity could reflect differences in their age distributions—Hispanics are younger on average than the population as a whole—it is important to control for this difference. Using the tool, it's easy to narrow this comparison to look specifically at 26–55 year olds.
In particular, the LFP rate for women of all ethnicities from 26 to 55 years old has declined by 1.0 ppt since 1998. In sharp contrast, the LFP rate for Hispanic women 26 to 55 years old has actually increased by 3.8 ppts. Compared to 20 years ago, this group is less likely to say they don't want a job because of disability/illness (1.1 ppt) and family responsbilities (1.4 ppt). This group is also less likely to be part of the shadow labor force (1.4 ppt) compared to two decades ago. (The shadow labor force, as we define it, is made up of individuals who say they want a job but are not considered unemployed by the BLS.) This article from the BLS delves into more detail about Hispanics in the labor force.
The LFP tool allows you to explore many other labor force stories. Users can cut the LFP data by three education categories (less than a high school degree, high school or some college, or associate's degree or higher), two age groups (26–55 or all ages), three race/ethnicity categories (white non-Hispanic, black non-Hispanic, and Hispanic) and for men and women. One thing that the tool makes clear is that the factors that influence individual decisions to work, look for work, or to pursue other activities vary across demographic groups, and each group's experience contributes to our understanding of movements in the overall LFP rate.
January 11, 2019
Are Employers Focusing More on Staff Retention?
Many people are quitting their current job. According to data from the Job Openings, Layoffs, and Turnover Survey (JOLTS) from the U.S. Bureau of Labor Statistics, workers are voluntarily leaving their current workplace (for reasons other than retirement or internal company transfers) at rates not seen since the late 1990s.
A high rate of quits is consistent with a tightening labor market. In such a market, the promise of more pay elsewhere lures an increasing share of workers away from their current employer. The potential benefit of changing jobs is also evident in the Atlanta Fed's Wage Growth Tracker, which typically exhibits a positive differential between the median wage growth of people changing jobs versus those who don't.
However, this Wage Growth Tracker pattern has changed in recent months. Since mid-2018, the median wage growth of job stayers has risen sharply and is at the highest level since 2008. As the following chart shows, the gap between the wage growth of job switchers and stayers has narrowed.
Is it possible that this new pattern is just noise and will disappear in coming months? After all, the wage growth for job stayers has spiked before. But taken at face value, it suggests that the cost of losing staff (recruiting cost, cost of training new employees, productivity loss from having a less experienced workforce, etc.) might be forcing some employers to boost current employees' pay in an effort to retain staff. Interestingly, the pattern is most evident among prime-age workers in relatively high-skill professional occupations.
Firms are always focused on staff retention and employ various strategies to reduce turnover in key positions. One of those strategies—larger pay increases—might now be taking a more prominent role.
We will closely monitor the Tracker for further evidence on this emerging trend. However, that might be later rather than sooner because the source data is provided by the U.S. Census Bureau, which has cut most services because of the current lapse in congressional funding. But when information becomes available, we'll be ready to parse it—so stay tuned!
December 04, 2018
Defining Job Switchers in the Wage Growth Tracker
Among the questions we receive about the Atlanta Fed's Wage Growth Tracker, one of the most frequent is about the construction of the job switcher and job stayer series. These series are derived from data in the Current Population Survey (CPS) and are intended to show how median wage growth differs for those who change their job from last year versus those who are in the same job. However, the monthly CPS does not actually ask if the person has the same job as a year ago.
So how to proceed? The CPS does contain information about the person's industry and occupation that we aggregate into consistent categories that can be compared to the person's industry/occupation reported a year earlier. If someone is in a different occupation or industry category, then we can reasonably infer that the person has changed jobs. To illustrate, for 2017, 18.8 percent of people in the Wage Growth Tracker data are in a different industry category than they were in 2016, and 28.6 percent are in a different occupation category. Of those who remain in the same industry, 23.9 percent changed occupation group, and of those in the same occupation group, 13.5 percent are in a different industry. However, this information doesn't allow us to identify all job switchers because being in the same industry and occupation group as a year earlier does not preclude having changed jobs.
Fortunately, the CPS also has questions based on who a person said their employer was in the prior month. It asks if that person still works there and if the employee's activities and job duties are the same as last month. If someone answers either of these questions in the negative, then it is likely that person is also in a different job than a year earlier. In 2017, 1.5 percent of people in the Wage Growth Tracker data said they have a different employer than in the prior month, and 0.9 percent report having different job responsibilities at the same employer.
Unfortunately, the dynamic structure of the CPS means that the responses to these "same employer/activity" questions can only potentially be matched with an individual's response in the prior two months, and not a year earlier. Moreover, some responses to those questions are blank, even for people whom we identify as being employed in the prior month. For those individuals, we simply don't know if they are in the same job as a month earlier. Of the non-null responses, the vast majority do not change job duties or employer from one month to the next. So if we assume the blank responses are randomly distributed among the employed population, it's reasonable to also treat the blanks as job stayers.
Previously, we had treated the blank "same employer/activity" observations as job switchers, but that approach almost certainly misclassified some actual job stayers as job switchers. Instead, we now define a job switcher as someone in the Wage Growth Tracker data who is in a different occupation or industry group than a year earlier, or someone who says no to either of the "same employer/activity" questions in the current or prior two months. We label everyone else a job stayer.
Does the definition matter for median wage growth? The following chart shows the annual time series of the difference between the median wage growth of job switchers and job stayers based on both the old and new definitions.
As you can see, the results are not qualitatively different. Job switchers have higher median wage growth during strong labor market conditions and lower growth during bad times. Not surprisingly, the gap in median wage growth is generally lower (more negative) using the old definition.
The next chart shows the annual time series of the share of job switchers in the Wage Growth Tracker data based on the new and old definitions. A caveat: we have been unable to construct occupation and industry groupings for 2003 that are completely consistent with the groupings used in 2002. This results in an erroneous spike in measured job switching for 2003.
Notice that the share of job switchers under the new definition peaked prior to the last two recessions, declined during the recessions, and then recovered. That share is now at a cyclical high. A discrete jump in the number of blank responses recorded for the "same employer/activity" questions in the CPS starting in 2009 masked this cyclicality under the old definition.
In about a week from now, the next update of the Wage Growth Tracker data will implement the new and improved definition of job switchers—I hope you'll check it out, and I'll be writing about it here as well.
November 16, 2018
Polarization through the Prism of the Wage Growth Tracker (Take Two)
In a previous macroblog post, I thought I had discovered an interesting differential between the wage growth of middle-wage earners and that of low/high-paid workers. It turns out that what I actually discovered is that my programming skills could be improved upon. The following is an update to the post, written after correcting the coding error. Although there is no obvious wage growth polarization story, the wages of low-wage workers are currently rising at a faster median rate than for other workers.
One of the most frequent questions we receive about the Atlanta Fed's Wage Growth Tracker (the median of year-over-year percent changes in individuals' hourly wage) is about the relationship between wage level and wage growth. For example, do high-wage earners also tend to experience greater wage growth?
When looking at wage growth by wage level, whether you use the prior or current wage level as the reference point matters—a lot. If we looked at wage growth categorized by the prior year's wages, we would find higher median wage growth for low-wage earners than for high-wage earners. This is because some workers who earned low wages last year earn middle or high wages this year, and some of last year's high-wage workers earn middle or low wages this year. If we instead categorized people based on current-year wages, we would see exactly the opposite: lower median wage growth for low-wage workers than for high-wage workers (see here for more discussion).
One way to lessen this wage-level base effect is to categorize an individual's wage growth according to their average wage across the two years. The following chart shows this categorization for the 2016–17 wage growth distribution of all workers in the Wage Growth Tracker data. (Note that since 1997, the annual salary for people whose earnings are only reported on a weekly basis is top-coded at $150,000 a year—these masked observations are excluded from the analysis). In the chart, the first quartile depicts the lowest-paid 25 percent of workers based on their average 2016–17 hourly wage, and so on. The center line of the box for each quartile is the median of that group's wage growth distribution, and the lower and upper boundaries of the box are the 25th and 75th percentiles, respectively. The outer lines are the thresholds for outlier observations (see here for the calculation.)
The chart shows that the wage growth distribution across the average-wage quartiles does, in fact, differ. In particular, the median wage growth for the lowest-paid workers is higher than the median for other types of workers. The median wage growth from 2016 to 2017 for the lowest quartile is 3.8 percent, 3.0 percent for the second quartile, and 3.2 percent for the third and fourth quartiles.
However, the pattern of relatively higher median wage growth for low-wage workers is not uniform over time. This difference is apparent in the following chart, which plots median wage growth over time for each average-wage quartile.
As the chart shows, median wage growth of low-wage workers (the green line, representing the first quartile) currently exceeds that of higher-wage workers, but it was below the median for higher-paid workers in the wake of the Great Recession. This pattern is consistent with the both the severity of the recession and what we have been hearing more recently about emerging shortages of low-skilled workers. It also appears that the median wage growth of the highest-paid workers (the blue line, representing the fourth quartile) slows by a bit less than that of other workers during downturns but is otherwise not much different than for workers in the middle of the wage distribution.
So, relative to the incorrect charts I had in the previous version of this post, there is no obvious wage growth polarization story here. The wages of low-wage workers are currently rising at a faster median rate than for other workers, and these other workers are experiencing broadly similar median wage growth.
- New Evidence Points to Mounting Trade Policy Effects on U.S. Business Activity
- Digging into Older Americans’ Flat Participation Rate
- What the Wage Growth of Hourly Workers Is Telling Us
- Making Analysis of the Current Population Survey Easier
- Mapping the Financial Frontier at the Financial Markets Conference
- The Tax Cut and Jobs Act, SALT, and the Blue State Blues: It's All Relative
- Improving Labor Force Participation
- Young Hispanic Women Investing More in Education: Good News for Labor Force Participation
- A Different Type of Tax Reform
- X Factor: Hispanic Women Drive the Labor-Force Comeback
- November 2019
- September 2019
- August 2019
- July 2019
- June 2019
- May 2019
- March 2019
- February 2019
- January 2019
- December 2018
- 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