Policy Hub: Macroblog provides concise commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues for a broad audience.
Comments are moderated and will not appear until the moderator has approved them.
Please submit appropriate comments. Inappropriate comments include content that is abusive, harassing, or threatening; obscene, vulgar, or profane; an attack of a personal nature; or overtly political.
In addition, no off-topic remarks or spam is permitted.
December 1, 2022
Labor Supply, Wages, and Inequality Conference: Day 2 Overview
The second day of the Atlanta Fed Center for Human Capital Studies's recent conference on labor supply, wages, and inequality switched the focus from labor supply to wage setting. The day was kicked off by Christina Patterson, who presented her paper "National Wage Setting ," coauthored by Jonathon Hazell and Heather Sarsons. This research explores how large, multi-establishment firms, which are increasingly dominating local labor markets, set wages across space. Benchmark models suggest that firms would vary wages across space because of local differences in productivity, cost of living, and competition, resulting in variation across regions.
The authors use data from the job market analytics firm Burning Glass Technologies about posted job-level wages for online vacancies between 2010 and 2019, along with a survey of human resource managers and executives, self-reported wages from payscale.com (a compensation data site), and firm employment visa application data. Their findings suggest that a large minority of firms set wages nationally and adopt pay structures that do not differ geographically. The two most important reasons given by firms is management simplicity and the importance of nominal comparisons to workers.
The national wage setting is associated with 3 to 5 percent lower profits for firms, but evidence suggests that national wage setting reduces earnings inequality without negatively affecting employment. However, this reduced inequality holds primarily for low-wage regions. National wage setting is also associated with increased regional wage rigidity.
The second paper of the day, "Industries, Mega Firms, and Increasing Inequality," presented by John Haltiwanger and coauthored by Henry R. Hyatt and James R. Spletzer, provided a broader lens through which we can view earnings inequality, which has drastically increased over the past decades. The existing empirical studies have shown that most of this inequality increase came from the rising differences in earnings between firms. Using comprehensive matched employer-employee data from the Longitudinal Employer-Household Dynamics database at the US Census Bureau, the authors show that the rising between-firms earnings dispersion is almost entirely accounted for by the increasing earnings dispersion between industries.
Increasing dispersion among industries operates at the two tails of the income distribution and is almost entirely accounted for by just 30 four-digit NAICS industries (as defined by the Census Bureau's classification system) The employment share of low-paying industries—such as restaurants and other eating places as well as general merchandise and grocery stores—has increased substantially, while real, inflation-adjusted wages in those industries fell. As a result, the left tail of the income distribution has fallen farther behind. On the other hand, the employment share of high-pay industries—such as software publishers, computer system design, information services, and management of companies—increased and was accompanied by large growth in those industries' average pay, leading to even higher relative income of the right tail of the income distribution.
Underlying these changes are worker-industry sorting and segregation patterns. Over time, workers with less education are more likely to end up working in low-paying industries, while more educated workers are more likely to cluster in the high-paying industries. These results suggest important changes have occurred in how lowest- and highest-paying firms restructure and organize themselves. These trends are also likely to be a by-product of recent technological innovations, led largely by firms and workers in industries with high pay. Though these innovations led to hefty rewards for high-skill workers, they also facilitated the scalability and expansion of mega-firms at the bottom of low-pay service industries. During the pandemic, workers in these low-pay industries have seen significant wage gains, but it remains to be seen if these recent changes will affect future inequality.
The day's third paper, "The Distributional Impact of the Minimum Wage in the Short and Long Run ," was presented by Elena Pastorina and coauthored by Erik Hurst, Patrick Kehoe, and Thomas Winberry. Their research continues the focus on wages by developing a framework to explore the impact of a $15 minimum wage, which would be a substantial increase in the current minimum wage and would be binding for 40 percent of workers without a college degree. The framework incorporates a large degree of worker heterogeneity within education groups, monopsony power (or considerable employer hegemony) in the labor markets, and putty-clay frictions (that allow for differing short- and long-run impacts of changes in the minimum wage).
Their results suggest that increases in the minimum wage are beneficial in the short run as they increase the welfare of the target group—low-income, noncollege workers making close to the initial minimum wage—with no large employment effects. However, the authors find that in the long run, firms will reoptimize their capital investment to better fit the changed relative prices of capital and labor. Thus, this group's employment, income, and welfare will eventually decline.
The authors go on to show that the Earned Income Tax Credit (EITC), which is based on income and number of children, is more effective in improving the welfare of low-wage workers than merely increasing the minimum wage. However, they find that combining a modest increase in the minimum wage with the EITC improves welfare more than either program does alone.
The fourth and final paper of the second day of the conference was "Labor Market Fluidity and Human Capital Accumulation," by Niklas Engbom. Using panel data for 23 countries, Engbom finds a large degree of heterogeneity in labor market fluidity—specifically, job-to-job mobility across countries. He finds that mobility in highly fluid markets is about 2.5 times higher than in countries with low fluidity, and that higher fluidity is associated with higher real wage growth over a person's lifetime.
Engbom also documents that on-the-job training is more prevalent in countries that exhibit high fluidity and proposes a mechanism to explain the positive correlation among fluidity, wages, and training in which workers in highly fluid markets are able to accumulate more on-the-job skills and have higher productivity, resulting in higher wages.
The amount of labor market fluidity can also change over time, and Engbom notes that fluidity in the United States—while higher than many other countries—has declined significantly during the last 40 years. Engbom connects this secular decline to the flattening of worker lifetime wage profiles and estimates that reduced fluidity accounts for about half of this flattening.
One implication of this line of research is that there are potentially significant benefits to reducing barriers to job creation and allowing greater worker reallocation across jobs. Lower labor market fluidity reduces wage growth and human capital accumulation because it becomes harder for people to find jobs that fully utilize their skills, and it also discourages human capital accumulation.
That paper concluded the Atlanta Fed Center for Human Capital Studies's conference on labor supply, wages, and inequality. Next year's conference is already in the planning stages, so stay tuned for details.
November 30, 2022
Labor Supply, Wages, and Inequality Conference: Day 1 Overview
The Atlanta Fed's Center for Human Capital Studies held its annual employment conference in person this year. The conference, held October 13–14, was organized by Melinda Pitts, the center director, and two center advisers, Richard Rogerson of Princeton University and Robert Shimer of the University of Chicago. The conference's title was "Labor Supply, Wages, and Inequality," and the agenda and links to the eight papers presented can be found here. This Policy Hub: Macroblog post summarizes the four papers presented on day one of the conference. The next post will look at the four papers presented on the second day.
Raphael Bostic, president and CEO of the Atlanta Fed, opened the conference. His welcoming remarks addressed policy makers' desire to understand the changing labor market, mentioning the work done by researchers at the Atlanta Fed and encouraging the economists in the room to continue doing policy-relevant research to better inform decision makers. His welcome was followed by the first session, which featured two papers related to how the COVID-19 pandemic altered individuals' labor-supply decisions.
The first paper presented was "Has the Willingness to Work Fallen during the Covid Pandemic? ," by R. Jason Faberman, Andreas I. Mueller, and Ayșegül Șahin, and presented by Faberman. The answer to the question their title poses is "yes": desired hours fell dramatically during the pandemic and have not recovered to prepandemic levels. Using data from the US Census Bureau's Current Population Survey and the New York Fed's Survey of Consumer Expectations, the authors find that the decline was most pronounced among those with less than a college education, those whose current or most recent jobs posed more significant COVID exposure risk, and those not working or working only part-time.
An important implication of the results reported in this paper is that while the unemployment rate is again near historic lows, the labor market might be even tighter than the unemployment rate is making it appear. In other words, by adding together the desired hours of those working and not working, the potential labor supply has fallen farther than either the unemployment rate or the labor force participation rate, compared to prepandemic levels. As a result, the difficulty employers are having finding workers, or getting workers to work more hours, might not ease any time soon.
Another broader consideration is whether this decline in desired hours is a temporary blip or a fundamental shift in preferences. The latter would hold implications on several fronts: for potential growth in an economy fueled by labor; for the way policymakers might define full employment, when employment of those "wanting" work leaves a significant amount of labor resources on the sideline; and for discussion of what incentives might be brought to bear on reversing the shift in preferences. This paper joins a growing body of literature showing that the impact of this pandemic on individual behavior has been dramatic and unprecedented. Additionally, the decline in desired hours of work could prove to have lasting and profound implications for future economic growth.
Adam Blandin followed with the presentation of his paper, "Work from Home Before and After the COVID-19 Outbreak ," coauthored with Alexander Bick and Karel Mertens. The authors designed the Real-Time Population Survey, a national labor market survey of adults aged 18–64 that ran from April 2020 through June 2021. The authors find that the share of the US population working from home (WFH) increased from 14 percent just before the pandemic to 40 percent early in the pandemic and still represented 25 percent of all employment as of June 2021. Working with custom survey questions and a structural model, the authors attempt to determine how much of the shift to WFH was a short-term substitution to an inferior form of production driven by the exigencies of the pandemic, as opposed to firms making a one-time investment to learn how to produce with remote workers. Specific survey questions found that more than 60 percent of workers who transitioned to WFH believed they could have always done their job remotely but were required to come in by their employer. Employing a structural model with endogenous wages (that is, wages based on a number of discrete factors) based, in part, on WFH status; a COVID-period in-person production penalty; and a one-time switching cost to remote work, the authors attribute much of the shift in work location to firms adopting remote work production. Combined with survey responses, the model suggests that remote work will persist long after COVID has waned.
The second session of the first day continued the theme of labor supply but shifted away from pandemic-specific research. Eric French presented "Labor Supply and the Pension-Contribution Link ," coauthored with Attila S. Lindner, Cormac O'Dea, and Tom A. Zawisza. Public pensions in the United States and many other are unfunded, pay-as-you go systems with benefits determined by a formula based on earnings history. Many governments have considered proposals to reform this formula, but a key concern is whether workers would respond to changes in their future pension benefits by adjusting their labor supply. To answer this question, the authors examined a change in the Polish pension system that altered the benefit for workers younger than 50 on January 1, 1999, with neither changes in benefits for older workers nor changes in the other plan characteristics. The original formula based benefits on the highest 10 years of salary growth, and the new system took into account every year of earnings.
Using a regression discontinuity design (RDD) and all tax returns linked to the Polish population registry, the authors estimate labor supply responses occurring between 2000 and 2002. This empirical design identifies the effects of the policy change by comparing individuals who were born only a few days apart and who face a very similar labor market and economic environment but are assigned to different pension plans. They found that the net return to work fell by an additional 5.2 percent in high-growth regions relative to low-growth regions. At the same time, the RDD allowed them to estimate that employment declines between regions differed by 2.29 percent. Taken together, these figures imply that the employment elasticity with respect to work incentives is 0.44.
This elasticity is in the range of estimates we typically see in the literature. However, one novel aspect of this paper lies in the fact that the research observes labor supply changes in response to changes in benefits to be received many years in the future, whereas most of the literature estimates the labor-supply response to the contemporaneous return to work. These results provide constructive evidence that individuals' labor supply responds in a forward-looking way to incentives in the pension formula, suggesting that tightening the link between contributions and benefits has the potential to alleviate labor supply distortions caused by payroll taxes.
Rather than focusing on how workers respond to external policy changes, the final paper of the day explored how an individual's risk preference and (over)confidence alter their job search behavior and labor market outcomes. Laura Pilossoph presented the last paper of the day, "Gender Differences in Job Search and the Earnings Gap: Evidence from the Field and Lab ," coauthored with Patricia Corté, Jessica Pan, Ernesto Reuben, and Basit Zafar.
The authors collected data on the employment search behavior of recent (2012–19) bachelor's graduates from the Questrom School of Business at Boston University. They collected data on the standard demographics involved in job search outcomes, including timing of acceptance and both accepted and rejected offers, job search expectations, and measures of risk. They found that, on average, women accepted jobs earlier in the search process than men did, the initial accepted salary was higher for men than for women, and the willingness to accept risk is higher for males. The authors then developed a job search model that incorporated gender differences in the levels of risk aversion and overoptimism about prospective job offers. The model predicts that if women are more risk-averse than men, then they will have lower reservation wages (the lowest wage at which someone would accept a given job) and search earlier. Likewise, if men are overconfident, then they will have a higher reservation wage. In other words, the decline in the reservation wage and increased job finding are derived from female risk aversion and male learning (that is, updating expectations about job offers) or having less optimism. Controlling for the measures of risk and overconfidence reduced the gender gap in wages by 37 percent.
The findings from the field were replicated in a specially designed laboratory experiment that featured sequential job search. The lab experiment yielded very similar results, with the gender gap in wages reduced by 30 percent when accounting for risk preferences and overconfidence. The results from both analyses suggest that risk preferences place a significant role in the gender differential.
In tomorrow's post, we'll summarize the papers presented on day two of the conference.
May 31, 2018
Learning about an ML-Driven Economy
Developments in artificial intelligence (AI) and machine learning (ML) have drawn considerable attention from both the real and financial sides of the economy. The Atlanta Fed's recent Financial Markets Conference, Machines Learning Finance: Will They Change the Game?, explored the implications of AI/ML for the financial system and public policy. The conference also included two macroeconomics-related sessions. A presentation of an academic paper, and the subsequent discussion, looked at why AI/ML has not (yet) shown up in the productivity statistics. Also, a policy panel on the implications of AI/ML developments for monetary policy was part of the conference. This post summarizes the policy panel discussion.
Vincent Reinhart, chief economist at Standish Mellon Asset Management, opened the panel discussion with the observation that developments in AI/ML could affect the performance of the overall economy in a variety of ways. For example, advancing technology could better match workers with jobs and, as a result, boost employment. On the other hand, it could also complicate job matching by forcing jobs and workers to become more specialized.
A combination of three factors is driving the recent growth in AI/ML, explained Carolyn Evans, head economist and senior data scientist at Intel Corporation: increased data availability, faster computers, and improved algorithms for analyzing the data. Like Reinhart, she noted that AI/ML could have various effects on the economy. For example, AI/ML is helping to reduce cost and boost supply. On the demand side, AI/ML is increasing the efficiency of product searches by buyers. However, as some online sellers become better than others at using AI/ML to help customers find the products they want, customer relationships may become stickier. In addition, firms may come to value interactions with customers more highly because these interactions could provide them with valuable data to use with AI/ML to better serve current and future customers. Evans raised the question of whether these developments could change the nature of pricing.
Dallas Fed president Rob Kaplan said he believes AI/ML is causing a structural change. It is not the first new technology to affect the economy, but the economic effects of this technology are more pervasive. For instance, business pricing power is already more constrained than it used to be, but even businesses that seemingly have some power currently worry that they make themselves more vulnerable to AI/ML-enabled disruption if they raise prices. Kaplan also emphasized the importance of skills training and building human capital to alleviate what he views as the inevitable loss of jobs to AI/ML.
The issue of how monetary policymakers should think about AI/ML was the focus of a presentation by Chicago Fed president Charles Evans. He observed that the "sign, magnitude, and timing" of any resulting structural change are all uncertain. This uncertainty, he said, argues against the use of fixed policy rules such as the Taylor Rule. He suggested that the Federal Reserve should instead follow an "outcome-based policy," adjusting policy based on the evolution of expected inflation and unemployment relative to the policy objectives of stable prices and full employment.
You can download the available presentations from the 2018 Financial Markets Conference web pages. The videos will be posted as they become available. Read Notes from the Vault for a summary of sessions on the strengths and weaknesses of ML, some financial regulatory and broader ethical issues, and the use of ML by investors.
February 13, 2018
GDPNow's Forecast: Why Did It Spike Recently?
If you felt whipsawed by GDPNow recently, it's understandable. On February 1, the Atlanta Fed's GDPNow model estimate of first-quarter real gross domestic product (GDP) growth surged from 4.2 percent to 5.4 percent (annualized rates) after a manufacturing report from the Institute for Supply Management. GDPNow's estimate then fell to 4.0 percent on February 2 after the employment report from the U.S. Bureau of Labor Statistics. GDPNow displayed a similar undulating pattern early in the forecast cycle for fourth-quarter GDP growth.
What accounted for these sawtooth patterns? The answer lies in the treatment of the ISM manufacturing release. To forecast the yet-to-be released monthly GDP source data apart from inventories, GDPNow uses an indicator of growth in economic activity from a statistical model called a dynamic factor model. The factor is estimated from 127 monthly macroeconomic indicators, many of which are used to estimate the Chicago Fed National Activity Index (CFNAI). Indices like these can be helpful for forecasting macroeconomic data, as demonstrated here and here.
Perhaps not surprisingly, the CFNAI and the GDPNow factor are highly correlated, as the red and blue lines in the chart below indicate. Both indices, which are normalized to have an average of 0 and a standard deviation of 1, are usually lower in recessions than expansions.
A major difference in the indices is how yet-to-be-released values are handled for months in the recent past that have reported values for some, but not all, of the source data. For example, on February 2, January 2018 values had been released for data from the ISM manufacturing and employment reports but not from the industrial production or retail sales reports. The CFNAI is released around the end of each month when about two-thirds of the 85 indicators used to construct it have reported values for the previous month. For the remaining indicators, the Chicago Fed fills in statistical model forecasts for unreported values. In contrast, the GDPNow factor is updated continuously and extended a month after each ISM manufacturing release. On the dates of the ISM releases, around 17 of the 127 indicators GDPNow uses have reported values for the previous month, with six coming from the ISM manufacturing report.
[ Enlarge ]
For months with partially missing data, GDPNow updates its factor with an approach similar to the one used in a 2008 paper by economists Domenico Giannone, Lucrezia Reichlin and David Small. That paper describes a dynamic factor model used to nowcast GDP growth similar to the one that generates the New York Fed's staff nowcast of GDP growth. In the Atlanta Fed's GDPNow factor model, the last month of ISM manufacturing data have large weights when calculating the terminal factor value right after the ISM report. These ISM weights decrease significantly after the employment report, when about 50 of the indicators have reported values for the last month of data.
In the above figure, we see that the January 2018 GDPNow factor reading was 1.37 after the February 1 ISM release, the strongest reading since 1994 and well above either its forecasted value of 0.42 prior to the ISM release or its estimated value of 0.43 after the February 2 employment release. The aforementioned rise and decline in the GDPNow forecast of first-quarter growth is largely a function of the rise and decline in the January 2018 estimates of the dynamic factor.
Although the January 2018 reading of 59.2 for the composite ISM purchasing managers index (PMI) was higher than any reading from 2005 to 2016, it was little different than either a consensus forecast from professional economists (58.8) or the forecast from a simple model (58.9) that uses the strong reading in December 2017 (59.3). Moreover, it was well above the reading the GDPNow dynamic factor model was expecting (54.5).
A possible shortcoming of the GDPNow factor model is that it does not account for the previous month's forecast errors when forecasting the 127 indicators. For example, the predicted composite ISM PMI reading of 54.4 in December 2017 was nearly 5 points lower than the actual value. For this discussion, let's adjust GDPNow's factor model to account for these forecast errors and consider a forecast evaluation period with revised current vintage data after 1999. Then, the average absolute error of the 85–90 day-ahead adjusted model forecasts of GDP growth after ISM manufacturing releases (1.40 percentage points) is lower than the average absolute forecast error on those same dates for the standard version of GDPNow (1.49 percentage points). Moreover, the forecasts using the adjusted factor model are significantly more accurate than the GDPNow forecasts, according to a standard statistical test . If we decide to incorporate adjustments to GDPNow's factor model, we will do so at an initial forecast of quarterly GDP growth and note the change here .
Would the adjustment have made a big difference in the initial first-quarter GDP forecast? The February 1 GDP growth forecast of GDPNow with the adjusted factor model was "only" 4.7 percent. Its current (February 9) forecast of first-quarter GDP growth was the same as the standard version of GDPNow: 4.0 percent. These estimates are still much higher than both the recent trend in GDP growth and the median forecast of 3.0 percent from the Philadelphia Fed's Survey of Professional Forecasters (SPF).
Most of the difference between the GDPNow and SPF forecasts of GDP growth is the result of inventories. GDPNow anticipates inventories will contribute 1.2 percentage points to first-quarter growth, and the median SPF projection implies an inventory contribution of only 0.4 percentage points. It's not unusual to see some disagreement between these inventory forecasts and it wouldn't be surprising if one—or both—of them turn out to be off the mark.
- Business Cycles
- Business Inflation Expectations
- Capital and Investment
- Capital Markets
- Data Releases
- Economic conditions
- Economic Growth and Development
- Exchange Rates and the Dollar
- Fed Funds Futures
- Federal Debt and Deficits
- Federal Reserve and Monetary Policy
- Financial System
- Fiscal Policy
- Health Care
- Inflation Expectations
- Interest Rates
- Labor Markets
- Latin AmericaSouth America
- Monetary Policy
- Money Markets
- Real Estate
- Saving Capital and Investment
- Small Business
- Social Security
- This That and the Other
- Trade Deficit
- Wage Growth