Macroblog

February 3, 2022

Rounded Wage Data and the Wage Growth Tracker

The US Census Bureau recently announced some changes it plans to make this year to the Current Population Survey Public Use File (CPS PUF). Here at the Atlanta Fed, we use data from the CPS PUF to construct the Wage Growth Tracker, and one of the planned changes will significantly affect the tracker. Specifically, a person's usual weekly or usual hourly earnings, which are unrounded currently, will be rounded.

The Wage Growth Tracker bases its results on the median, or middle, observation in the distribution of percent wage changes for a sample of individuals linked between the current month and the same month a year earlier. The Wage Growth Tracker time series has yielded useful insight into the rapidly shifting dynamics of the labor market in the wake of COVID-19, especially as compositional effects have distorted wage data. It's also helped economists and policymakers understand which income levels were seeing the greatest growth and that job switchers were finding the most wage gains.

How will the rounding of wage data affect the Wage Growth Tracker? The announced CPS PUF rounding rules vary by wage level and are different if the earnings are reported on an hourly or weekly basis. (You can see more details here .) Most people in the CPS report earnings on an hourly basis, and most wage observations range from $10 to $99.99 an hour. Under the rounding rules those earnings will be rounded to the nearest dollar. So, for example, if someone reports making $14.50 an hour, that wage will be rounded to $15, while workers reporting a wage of $15.40 an hour will also have that wage rounded to $15.

One implication of the rounding rules is that it will make no wage change appear more common than it currently is. To illustrate, suppose someone's pay went from $14.50 to $15.40 an hour. The rounding rules would show no change in the person's wage (both would be recoded as $15) even though that person's actual wage increased by 6.2 percent. Chart 1 shows what happens to the proportion of zero wage changes if the rounding rules were applied to the CPS PUF earnings data used to construct the Wage Growth Tracker from 1998 to 2021.

As you can see, during the Great Recession, when labor demand was especially weak, about 17 percent of wage growth observations based on unrounded earnings data were zero. But if the rounding rules had been applied back then, more than 25 percent of wage growth observations would have been zero.

Obviously, this change is a big deal for the Wage Growth Tracker. When more than a quarter of the observations are zero near the middle of the wage change distribution, it is very likely that the median observation will also be zero. This effect is evident in chart 2, which compares the median Wage Growth Tracker series using unrounded earnings data with what it would have been if the rounded data had been used.

Clearly, if the rounding rules had been in place in the past, the Wage Growth Tracker time-series would be a much less useful indicator of wage growth or labor market trends.

So, what to do? It turns out that the rounding rules don't affect all summary measures of wage growth as much as they affect the median measure. For example, as chart 3 shows, the mean—or average—wage growth comes out of the rounding changes essentially unaffected.

Unfortunately, not only is the average higher than the median, because wage growth varies greatly across individuals (the monthly sample standard deviation is typically around 25 percent), you can also see that both of the average wage growth series are much more variable month to month than the median series using unrounded data. Indeed, the robustness to variability in the underlying wage change data is a primary reason why the Atlanta Fed's Wage Growth Tracker is based on median rather than average wage growth.

But there is potential solution. Borrowing from the research on using trimmed means of price change data to construct measures of inflation that are robust to extreme price changes, I was able to construct a trimmed-mean wage growth series using the rounded data that has broadly similar properties to the (median) Wage Growth Tracker series constructed from unrounded data. Specifically, for each month's sample, I excluded the bottom 20 percent of wage growth observations (that is, the largest percent wage declines) and the top 25 percent (the largest percent increases) and computed the average of the remaining data. (Note that the trimming is asymmetric because more of the large wage changes tend to be increases than decreases, which is also why the average is higher than the median in the previous chart.)

Chart 4 shows the trimmed-mean series constructed using rounded earnings data, along with the (median) Wage Growth Tracker series that uses unrounded data. I would describe this trimmed-mean series as a reasonable (though not perfect) approximation of the Wage Growth Tracker series (something we could have used if we only had rounded earnings data in the past).

When the January 2022 CPS PUF data become available in February, we will produce the trimmed-mean version of the overall Wage Growth Tracker and add it to the Atlanta Fed's Wage Growth Tracker data set. We are currently exploring if a similar approach will produce useful alternatives to the Wage Growth Tracker for other ways to view the data, such as those for job switchers versus job stayers, or by average wage level. Watch this space for updates.

By John Robertson, a senior policy adviser in the Atlanta Fed's Research Department

• February 3, 2022 in
  • Labor Markets
  • Wage Growth
• Permalink
• Comments

November 10, 2021

Compositional Distortions to a Measure of Wage Growth during the Pandemic

Measures of year-over-year growth in wages (or hourly earnings) used in economic analysis often tell a fairly consistent story. For example, chart 1 makes it apparent that wage growth was generally higher heading into the 2007–09 recession than heading out of it and that wage growth stayed low for the first half of the 2010s before trending up moderately over the second half of the decade. However, with the onset of the COVID-19 pandemic, growth in average hourly earnings from the US Bureau of Labor Statistics' (BLS) establishment survey (the blue line in the chart) deviated substantially from the other two series depicted.

The leisure and hospitality industry provides a useful illustration of why the establishment survey measure of hourly earnings growth spiked in March and April of last year. In February 2020, average hourly earnings for production and nonsupervisory workers in leisure and hospitality were 40 percent lower than they were for all private nonfarm payroll workers. And although the leisure and hospitality industry accounted for just under 14 percent of private nonfarm production and nonsupervisory jobs in February 2020, it accounted for nearly 40 percent of the lost private production and nonsupervisory jobs in the subsequent two months. The 4.5 percentage point increase from February 2020 to April 2020 in the blue line in chart 1 falls by 1.9 percentage points if we remove leisure and hospitality from the calculation.

The August 2020 FRBSF Economic Letter—aptly titled "The Illusion of Wage Growth"—by Erin E. Crust, Mary C. Daly, and Bart Hobijn shows that restricting the sample to people employed in the second quarters of 2019 and 2020 reduced growth in median usual weekly earnings over that period by nearly 8 percentage points from the published rate of 10.4 percent. The Atlanta Fed's Wage Growth Tracker, which uses the same type of restriction, and the Employment Cost Index (ECI), which controls for employment share changes among industries and occupations , were not subject to the illusion of wage growth shown by the blue line in chart 1.

Unfortunately, the adjustments used in the Wage Growth Tracker and the ECI are not feasible with the establishment survey measure of hourly earnings because that measure is constructed solely from the information in each month's employment report. As an alternative, Goldman Sachs provides an adjustment for what it terms the composition bias in the establishment survey measure. This adjustment keeps hours worked fixed at their year-ago level in the wage calculation using industry-level data.

I've written an appendix  that provides the details of a related approach for calculating a composition-adjustment term from the monthly establishment survey data. Besides adjusting for industry composition, this approach also adjusts for types of workers: production and nonsupervisory workers versus nonproduction/supervisory employees. The appendix also shows that adjusting for worker type and industry rather than industry alone materially affects the composition-adjusted measure of average hourly earnings for April 2020. It also shows that—unlike measures from the BLS and the San Francisco Fed, which control for educational attainment—the measure of labor composition (sometimes called labor quality ) constructed with only establishment survey data has not trended up much since the mid-2000s.

The basic intuition underlying the approach described in the appendix is that, apart from some trivial rounding error, the BLS measure of aggregate weekly payrolls is equal to the product of average hourly earnings and aggregate weekly hours worked. So, in much the same way that we can express nominal gross domestic product (GDP) as the product of real chain-weighted GDP and a GDP price deflator, aggregate weekly payrolls can also be decomposed as the product of composition-adjusted measures of wages and hours worked. This approach maintains the equality with aggregate payrolls since the composition adjustments to hourly earnings and hours worked offset each other exactly.

Chart 2 shows the results of adjusting for changes in both industry and worker type for measures of average hourly earnings growth and aggregate hours worked during the pandemic. Adjusting for composition makes average hourly earnings growth during the pandemic more like the ECI and Wage Growth Tracker measures, but, nevertheless, some important differences exist. Unlike the composition-adjusted measure of nominal wage growth, the ECI and Wage Growth Tracker measures languished in the second half of 2020 and surged in their most recent readings. Composition-adjusted hourly earnings grew 1.1 percent from March 2020 to April 2020, which is less than the 4.6 percent spike in the unadjusted measure but still strong enough to suggest that the adjustments made here still miss some meaningful changes in worker composition in the earliest months of the pandemic.

As you look at this chart, note that the adjustment is constructed using wage and hours data for 253 industry groups, all but 10 of which are further split into production and nonsupervisory and nonproduction/supervisory employee groups.

The right panel in chart 2 shows private nonfarm payroll employment alongside the standard measures of aggregate hours worked and a measure adjusted for industry and worker-type composition. In October, private nonfarm payroll employment, hours worked, and composition-adjusted hours worked were 2.5, 1.7, and 1.2 percentage points, respectively, below their February 2020 levels.

The composition-adjustment factor (industry by production/supervisory worker employment type) as well as the associated measures of composition-adjusted hours worked and hourly earnings are available here . Future updates of this Excel file will also be available at this link.

By Pat Higgins, a policy adviser and economist in the Atlanta Fed's Research Department

• November 10, 2021 in
  • Labor Markets
  • Wage Growth
• Permalink
• Comments

March 23, 2021

Hourly and Weekly Perspectives on Wage Growth during the Pandemic

Despite record-setting job losses during the COVID-19 pandemic, median growth in the hourly rate of pay for those who stayed employed has held up remarkably well, which we can see in the Atlanta Fed's Wage Growth Tracker (see chart 1).

The Wage Growth Tracker compares individual hourly wages in the current month with what the same individual's hourly wage was 12 months earlier and calculates the change. The fact that the median wage growth has not slowed, despite the increase in unemployment, suggests that the pandemic's impact on the labor market has been quite unusual.

During the Great Recession, the slowing in median hourly wage growth coincided with a large increase in the share of workers reporting that their hourly rate of pay was unchanged from a year earlier. As chart 2 shows, the share of workers reporting zero change in their hourly rate of pay has ticked up a bit during the COVID-19 pandemic, but so far, what we see differs from observations we made during the Great Recession.

Why did the COVID-19 pandemic have a relatively smaller impact on median hourly wage growth compared to the Great Recession? One explanation is that the supply of unemployed job seekers far exceeded job vacancies in the earlier recession. That is, employers typically received many more applicants for each available position. As chart 3 shows, at the Great Recession's peak, there were 6.5 unemployed workers for each job posting and 5.7 unemployed not on temporary layoff for each job posting. I think unemployed workers not on temporary layoff is a more useful measure of unemployed job seekers because those on temporary layoff expect to be recalled by their employer and hence are not necessarily looking for another job. Contrast that with January 2021, when there were 1.5 unemployed workers for each opening and 1.1 unemployed workers not on temporary layoff for each job vacancy. In this sense, the labor demand and supply during the COVID-19 pandemic has been more in balance than during the Great Recession. Compared with the Great Recession, apart from the period during the initial lockdown, total vacancies by firms has scaled back relatively modestly during the pandemic while the number of workers looking for a job has increased by less.

Nonetheless, during both the Great Recession and the COVID-19 pandemic, many workers who remained employed have experienced an involuntary reduction in their work hours, which has dragged down workers' weekly paychecks even when their hourly rate of pay hasn't fallen. In February 2021, about 6.5 million workers were classified by the U.S. Bureau of Labor Statistics (BLS) as working part-time for economic reasons—almost 2 million more than in February 2020, just before the pandemic hit the U.S. economy. For this reason, I've constructed an alternate version of the Wage Growth Tracker, which shows the median growth of individual weekly earnings. This new measure uses the same data (from the Current Population Survey, jointly administered by the BLS and the U.S. Census Bureau) as the hourly earnings measure, and I show both series in chart 4 for comparison.

Generally, the two series move in tandem, with the weekly series slightly outpacing the hourly series during economic expansions as hours worked tend to rise. However, as we see here, during both the Great Recession and the COVID-19 pandemic, reduced hours worked each week lowered many workers' median growth in weekly earnings relative to hourly earnings.

As the economy recovers from the COVID-19 pandemic, watching both the hourly and weekly versions of the Wage Growth Tracker will be useful. As fewer worker face reduced hours, I expect to see median weekly wage growth recover and at least match the pace of hourly wage growth. A tighter labor market should result in higher wage growth on both an hourly and weekly basis. I'll write about the developments using new Wage Growth Tracker data we'll post soon, so check back.

Note: If you are interested in tracking the hourly and weekly versions of the Wage Growth Tracker you can do that here, or via the EconomyNow app, which also features several other Atlanta Fed data tools.

By John Robertson, a senior policy adviser in the Atlanta Fed's Research Department

• March 23, 2021 in
  • Employment
  • Labor Markets
  • Unemployment
  • Wage Growth
• Permalink
• Comments