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
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 25, 2019
Tariff Worries and U.S. Business Investment, Take Two
Last summer, we reported that one fifth of firms in the July Survey of Business Uncertainty (SBU) were reassessing capital expenditure plans in light of then-recent tariff hikes and retaliation concerns. Roughly 6 percent had already cut or deferred capital spending as a result of tariff worries.
Since then, tariff hikes and trade policy tensions have continued to mount, as recounted in the Peterson Institute's Trade War Timeline. U.S. stock market volatility also rose sharply in the last four months of 2018, partly in reaction to trade policy concerns. These developments led us to pose another round of questions about trade policy and investment in the January 2019 SBU.
We first asked each firm if tariff hikes and trade policy tensions caused it to alter its capital expenditures in 2018 and, if so, in which direction and by how much. We use the responses to estimate the net impact of tariff hikes and trade policy tensions on U.S. business investment in 2018.
We estimate that tariff hikes and trade policy tensions lowered gross investment in 2018 by 1.2 percent in the U.S. private sector and by 4.2 percent in the manufacturing sector. The larger response for manufacturing makes sense, given its relatively high exposure to international trade. In constructing these estimates, we consider firms that raised and lowered investment due to trade policy, and we weight each firm by its size.
To estimate the dollar impact of trade policy developments, we multiply the percentage amounts by aggregate investment values. The resulting amounts for U.S. business investment in 2018—minus $32.5 billion for the private sector and minus $22 billion for manufacturing—are modest in magnitude, in line with our forward-looking assessment last summer.
In January, we also asked forward-looking questions about the potential impact of trade policy worries on business investment. As reported in Exhibit 2 below, 20 percent of firms said they are reassessing their capital expenditure plans in 2019 because of tariff hikes and trade policy tensions, a share very similar to what we obtained in our forward-looking question last July. As before, manufacturing firms were more likely to reassess their capital spending plans due to trade policy concerns.
Exhibit 3 below speaks to the question of how firms have reassessed their capital expenditure plans. Here, too, results are similar to what we reported last summer, with one important exception. Among firms reassessing, more than half have either postponed or dropped some portion of their capital spending for 2019, compared to just 31 percent in July 2018. Thus, it appears that firms anticipate somewhat larger negative effects of trade policy developments on capital expenditures in 2019 than they did in 2018.
All told, our results continue to suggest that tariff hikes and trade policy tensions have had a rather modest impact on U.S. business investment. Of course, tariffs and other trade barriers affect U.S. and foreign economies through multiple channels. Even if the near-term business investment effects of trade policy developments are modest in magnitude, trade barriers can disrupt supply chains, raise input prices, and lead to higher prices for consumer goods. That's important to keep in mind as the trade policy outlook remains murky.
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 16, 2019
I didn't coin the title of this blog post. It was the label on a chart of the Federal Reserve's balance sheet that appeared in an issue of The Wall Street Journal last week. I've led with this phrase because it does seem to capture some of the sentiment around what has become the elephant in the monetary policy room: Is the rundown in the size of the Fed's balance sheet causing an unanticipated, and unwarranted, tightening of monetary policy conditions? I think the answer to that question is "no." Let me explain why.
In June 2017, the Federal Open Market Committee (FOMC) determined that it was appropriate to begin the process of reducing the size of the Fed's balance sheet, which had more than quadrupled as a result of efforts to combat the financial crisis and support the subsequent recovery from a very deep recession.
As I noted in a speech last November, I see the Committee's strategy for shrinking the balance sheet as having two essential elements.
- First, the normalization process is designed to be gradual. It was phased in over the course of about a year and a half and is now subject to monthly caps so the run-down is not too rapid.
- Second, the normalization process is designed to be as predictable as possible. The schedule of security retirements was announced in advance so that uncertainty about the pace of normalization can be minimized. (In other words, "quantitative tightening" is decidedly not on the QT.) As a result, the normalization process also reduces complexity. Balance-sheet reduction has moved into the background so that ongoing policy adjustments can focus solely on the traditional interest-rate channel.
In his recent remarks at the annual meeting of the American Economic Association, Chairman Jerome Powell was very clear that the fairly mechanical balance-sheet strategy adopted by the FOMC thus far should not be interpreted as inflexibility in the conduct of monetary policy or an unwillingness to recognize that balance-sheet reduction is in fact monetary policy tightening.
I will speak for myself. Balance-sheet policy is an element of the monetary policy mix. The decision to adopt a relatively deterministic approach to balance-sheet reduction is not a decision to ignore the possibility that it has led or might lead to a somewhat more restrictive stance of monetary policy. It is a decision to make whatever adjustments are necessary through the Fed's primary interest-rate tools to the greatest extent possible.
I maintain that there is still wisdom in this approach. The effects of our interest-rate tools are much more familiar to both policymakers and markets than balance-sheet tools are. That, to my mind, makes them the superior instrument for reaching and maintaining our dual goals of stable inflation and maximum employment. It is my belief that reducing the number of moving pieces makes monetary policy more transparent and predictable, which enhances the Committee's capacity for a smooth transition toward those goals.
It should now be clear that nothing is written in stone. Whether the FOMC uses active interest-rate policy with passive balance-sheet policy or uses both instruments actively, policy decisions will ultimately be driven by the facts on the ground as best Committee members can judge, and by assessments of risks that surround those judgments.
In my own judgment, it is far from clear that the ongoing reduction in the balance sheet is having an outsized impact on the stance of monetary policy. I think it is widely accepted that one of the ways balance-sheet policies work is by affecting the term premia associated with holding longer-term securities. (There are many good discussions about balance-sheet mechanisms, including this one by Edison Yu, which can be found in the first quarter 2016 edition of the Philadelphia Fed's Economic Insights, or this article by Jane Ihrig, Elizabeth Klee, Canlin Li, Min Wei, and Joe Kachovec in the March 2018 issue of the International Journal of Central Banking.)
Lots of things can push term premia up and down. But one of the factors is the presumed willingness of the central bank to purchase long-term securities in scale—or not. The idea that running down the balance sheet tightens monetary policy is that, in so doing, the FOMC is removing a crucial measure of support to the bond market. This makes longer-term securities riskier by transferring more duration risk back to the market, which raises term premia and, all else equal, pushes rates higher.
Although estimating term premia is as much art as science, I don't think the evidence supports the argument that these premia have been materially rising as a result of our normalization process. The New York Fed publishes one well-known real-time estimate of the term premia associated with 10-year Treasury securities. But isolating and quantifying the effect of balance-sheet changes on term premia is challenging. It is possible that a number of factors, such as the continued high demand for U.S. Treasuries by financial institutions and a low inflation risk premium, might have dampened the independent effect of balance sheet run-off. But if the term premia channel is a critical piece of what makes balance-sheet policy work, I'm hard pressed to see much evidence of financial tightening via rising term premia in the data so far.
Lest anyone think I am overly influenced by one particular theory, I will emphasize that I am not taking anything for granted. In addition to my monitoring of developments on Main Street, I will be watching financial conditions and term premia as I assess the outlook for the economy. My view is that a patient approach to monetary policy adjustments in the coming year is fully warranted in light of the uncertainties about the state of the economy, about what level of policy rates is consistent with a neutral stance, and about the overall impact of balance-sheet normalization. This patience is one of the characteristics of what I mean by data dependence.
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 29, 2018
Cryptocurrency and Central Bank E-Money
The Atlanta Fed recently hosted a workshop, "Financial Stability Implications of New Technology," which was cosponsored by the Center for the Economic Analysis of Risk at Georgia State University. This macroblog post discusses the workshop's panel on cryptocurrency and central bank e-money. A companion Notes from the Vault post provides some highlights from the rest of the workshop.
The panel began with Douglas Elliot, a partner at Oliver Wyman, discussing some of the public policy issues associated with cryptoassets. Drawing on a recent paper he cowrote, Elliot observed that there are "at least four substantial market segments" that provide long-term support for cryptoassets:
- libertarians and techno-anarchists who, for ideological reasons, want a currency without a government;
- people who deeply distrust their government's economic management;
- seekers of anonymity, who don't want their names associated with transactions and investments; and
- technical users who find cryptoassets useful for some blockchain applications.
Besides these groups are the speculators and investors who hope to benefit from price appreciation of these assets.
Given the strong interest of these four groups, Elliot argues that cryptoassets are here to stay, but he also asserts that these assets raise public policy issues that regulation should address. Some issues, such as anti–money laundering, are being addressed, but all would benefit from a coordinated global approach. However, he observes that of the four long-term support groups, only the technical users are likely to favor such regulations.
Another paper, by University of Chicago professor Gina C. Pieters, analyzed the extent to which the cryptocurrency market is global using purchases of cryptocurrency by state-issued currencies. She finds that more than 90 percent of all cryptocurrency transactions occur using one of three currencies: the U.S. dollar, the South Korean won, and the Japanese yen. She further finds that the dominance of these three currencies cannot be explained by economic size, financial openness, or internet access. Pieters also observed that transactions involving bitcoin, the largest cryptocurrency by market value, do not necessarily represent a country's cryptomarket share.
Warren Weber, former Minneapolis Fed economist and a visiting scholar at the Atlanta Fed, discussed so-called "stable coins," one type of cryptocurrency. The value of many cryptocurrencies has fluctuated widely in recent years, with the price of one bitcoin soaring from under $6,000 to more than $19,000 and then plunging to just over $6,000—all within the period from October 2017 to October 2018. This extreme price volatility creates a significant impediment to Elliot's technical users who would like some method of buying blockchain services with a currency controlled by a blockchain. In an attempt to meet this demand, a number of "stable coins" have been issued or are under development.
Drawing on a preliminary paper, Weber discussed three types of stable coins. One type backs all of the currency it issues with holdings of a state-issued currency, such as the U.S. dollar. A potential weakness of these coins is that they incur operational costs that require payment. Weber observed that interest earnings might cover part of these expenses if the stable coin issuer holds the dollars in an interest-bearing asset. Additionally, charging redemption fees might offset some or all of the expense.
The other two alternatives involve the creation of cryptofinancial entities or crypto "central banks." Both of these approaches seek to adjust the quantity of the cryptocurrency outstanding to stabilize its price in another currency. However, Weber observed that both of these approaches are subject to the problem that the cryptocurrency could take on many values depending upon people's expectations. If people come to expect that a coin will lose its value, neither of these approaches can prevent the coin from becoming worthless.
The question of whether existing central banks should issue e-money was the topic of a presentation by Francisco Rivadeneyra of the Bank of Canada. Summarizing the results of his paper, Rivadeneyra observed that central banks could provide e-money that looks like a token or a more traditional account. The potential for central banks to offer widely available account-based services has long existed. However, after considering the tradeoffs, central banks have elected not to provide these accounts, and recent technological developments have not changed this calculus. However, new technologies may have changed the tradeoff for token-based systems. Many issues will need to be addressed first, though.
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.
November 14, 2018
Polarization through the Prism of the Wage Growth Tracker
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?
An earlier macroblog post explored this question. Unfortunately, answering it is not as easy as it might appear. 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.
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 to 2017 wage growth distribution of all workers in the Wage Growth Tracker data. In the chart, the first quartile (labeled <$13.8) 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. For example, the median wage growth from 2016 to 2017 for the lowest quartile is 3.9 percent, 1.6 percent for the second quartile, 1.9 percent for the third quartile, and 3.2 percent for the top quartile.
The pattern of higher median wage growth in the lower and upper quartiles, compared with the middle part of the wage distribution, is reasonably uniform over time. However, there is a cyclical difference between the median wage growth of high- and low-wage earners. 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, first quartile) currently exceeds that of high-wage workers (the blue line, fourth quartile), but it was below the median for high-wage 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. In contrast, median wage growth for workers in the middle of the wage distribution (the orange and purple lines) remains lower than for either high- or low-wage workers. Overall, these findings reinforce the idea of polarization, where the demand for workers has generally grown more in the tails of the skill/wage distribution than in the middle.
- 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
- Tariff Worries and U.S. Business Investment, Take Two
- Trends in Hispanic Labor Force Participation
- Quantitative Frightening?
- Are Employers Focusing More on Staff Retention?
- June 2019
- May 2019
- March 2019
- February 2019
- January 2019
- December 2018
- November 2018
- October 2018
- August 2018
- July 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