1. Home
  2. Meetings of the Commission
  3. Meeting of April 5, 2017 - EEOC to Examine the State of the Workforce and the Future of Work
  4. Written Testimony of Dr. Aparna Mathur Resident Scholar in Economic Policy Studies at the American Enterprise Institute

Written Testimony of Dr. Aparna Mathur Resident Scholar in Economic Policy Studies at the American Enterprise Institute

Meeting of 4-5-17 EEOC to Examine the State of the Workforce and the Future of Work

I. Introduction

The U.S. labor market has largely recovered from the downturn during the Great Recession. Unemployment has returned to long-term historical rates near 5% from a peak of 10% in October 2009. Average wages have started to show signs of improvement, growing at over 2.5 percent per annum over the previous year. Other indicators reported by the Bureau of Labor Statistics still show signs of slack in the labor market but are trending in the right direction. The U-6 rate which captures not just the unemployed but also involuntary part-time workers is at 9.4 percent today, marginally higher than pre-recession levels. After steadily declining since 2008, the labor force participation rate has levelled out at 63 percent today. According to the latest jobs report from the BLS, there are 7.5 million unemployed persons, of which 1.8 million are long-term unemployed i.e. they have been jobless for 27 weeks or longer. An additional 5.7 million workers would prefer full-time work but are working part-time because their hours have been cut or they are unable to find full-time jobs.

While the jobs report provides a snapshot of how workers are faring in the economy, the BLS also reports information on available job openings and firms' ability to hire workers in the labor market, relative to the number of vacancies they create. According to the latest data produced as part of the Job Openings and Labor Turnover Survey (JOLTS), there were 5.6 million job openings at the end of January in 2017. Job openings data show the number of unfilled vacancies at firms at the end of each month. Positive trends in job openings can signal either that companies are optimistic about the economy and are therefore creating jobs and attempting to recruit workers to fill these vacancies, or they can signal an inability of firms to hire talented employees, a phenomenon referred to as the "skills-gap". In this testimony, I explore trends in job openings across industries and nationwide, relate them to the number of unemployed workers and explore potential reasons why the matching between job openings and unemployed workers might not be taking place. I also explore what these trends imply for the future of work as well as opportunities for American workers.

II. What do We Know About the Unemployed?

While the official unemployment rate has come down for all workers to 4.7 percent, this masks significant differences across demographic groups.i In the latest jobs report, the unemployment rates for men and women were at 4.3 percent. However, the rates remain elevated for Blacks, Hispanics and American Indians, and are almost double the rates for Whites. Figure 1, the unemployment rate for Blacks is at 8.1 percent, and for Hispanics, at 5.6 percent. Further, there are significant disparities across age groups. Younger workers are still at a disadvantage in the market, with teenagers facing unemployment rates of 15 percent and those between the ages of 16 and 24 facing rates higher than 10 percent.




Some of this can be explained by educational attainment. The unemployment rate for workers with a Bachelor's degree or higher is approximately half of the official unemployment rate, at 2.5 percent.ii For workers with less than a high school diploma, the unemployment rate is 7.5 percent. These trends have persisted over time, with more educated workers typically experiencing much lower rates of unemployment.


A 2014 report from the Pew Research Center shows that these differences in education levels not only translate into differences in unemployment rates, but also in earnings and job satisfaction.iii This report, which focused on workers between 25 and 32 years of age, showed that college- educated workers were more likely to have full-time jobs relative to those with only a high- school diploma. Their earnings were significantly higher, and they were much more likely to be in jobs that offered a career track. Moreover, they reported higher levels of satisfaction with their job and felt that they had enough education and training to move up in their careers. Interestingly, when these college graduates were asked what would have helped prepare them for their current job, a majority answered more on-the-job training or more work experience.

Skill differences also play a significant role in explaining wage differentials in the U.S. and unemployment outcomes, as shown in a recent paper.iv The authors show that the gender wage gap in the U.S. is partly explained by the lower skill levels of women, and skill gaps are a primary reason for the racial wage gap.

Moreover, Black unemployment rates have been consistently above those of Whites for the last several decades.v One potential explanation is that Blacks live in areas where there is a scarcity of jobs. This is called spatial mismatch, and studies find that this is an important issue for low- skilled Black males. However, a recent paper finds that even if Blacks live in areas which are dense in jobs at their skill level, their jobs are more likely to be held by Whites, a phenomenon known as racial

It is also instructive to look at employment demographics, shown in Table 1 below.

              Table 1: Percent of Total Employed by Industry, 2016

Industry Women Black Asian Hispani
Total, 16+ 46.8 11.9 6.1 16.7
Agriculture, Forestry, Fishing/Hunting 25.3 2.7 1.1 24.5
Mining, quarrying, oil & gas 13.4 6.4 2.3 17.8
Construction 9.1 5.8 1.9 28.9
Manufacturing 28.9 10 6.8 16.1
Trade 44.5 11.3 5.9 17.1
Transportation & Utilities 23.5 18 5.3 17.3
Information 40.6 11.1 7.6 11.3
Finance 52.3 10.4 7 12.2
Professional & Business Services 41.4 9.7 8.3 16.6
Education Services 68.4 11.1 4.6 11.2
Health Care & Social Assistance 78.7 17.4 6.9 12.7
Leisure & Hospitality 50.6 12.6 6.5 22.9
Government 45.1 16.7 4.5 12.2

Women make up less than half the labor force in 2016. They are primarily employed in social sectors like healthcare and education. By race, Blacks, Asians and Hispanics made up a small proportion of the total labor force at 11.9 percent for Blacks, and 16.7 percent for Hispanics. Blacks are a larger fraction of all employed in healthcare while Hispanics are much more likely to work in construction.

Disability also plays a role in labor force participation. The BLS reports that participation rates are only 20.4 percent for people with a disability, and 68.4 percent for those without a disability.vii As I have written earlier, the Social Security Disability Insurance system's benefit structure disincentivizes work and results in low rates of exit and high rates of entry into the program.viii Reforming the disability system so that workers who are able to continue to participate in the workforce are encouraged to do so is critical.

Family structure is another important factor. In 2015, families maintained by women with no spouse present-single mother families-remained less likely to have an employed member (75 percent) than families maintained by men with no spouse present (82.9 percent) or married- couple families (81.4 percent). Among married-couple families, both the husband and wife were employed in 48.0 percent of families; in 19.8 percent of married-couple families only the husband was employed, and in 7.1 percent only the wife was employed.ix

III. Job Openings

Next we look at the number of job openings or vacancies across the country to see how they relate to the number of unemployed workers. Job openings information is collected in the JOLTS data for the last business day of the reference month. It includes openings for full-time, part-time, permanent, short-term and seasonal openings. As shown in Figure 2 below, there were 5.6 million non-farm job openings as of January 2017. Job openings peaked prior to the recession, then experienced a steep decline over the course of the Great Recession, and are now significantly above pre-recession levels.




To see how job openings relate to the number of unemployed persons, the BLS also produces a chart showing the ratio of unemployed persons per job opening. This is reproduced in Figure 3 below. At the start of the recession in December 2007, there were 1.9 unemployed people per job opening. This reached a peak in June 2009, at a ratio of 6.6 persons per job opening and since then has trended downward. In January 2017, the ratio was at 1.4.





The Beveridge curve captures this relationship by plotting the unemployment rate against the job openings rate. As Figure 4 shows, at the peak of the recession, the Beveridge Curve was far to the right and low, showing high unemployment rates with low job opening rates. Through January 2017, the curve has shifted left and up, showing the decline in the unemployment rate and the increase in the job openings rate. The Beveridge Curve can shift because of factors like skills mismatches, increased labor force participation and economic and policy uncertainty, which would cause the curve to shift outward.





IV. Employment Trends by Industry


To observe how these job vacancies are distributed across industries, we calculate the job openings rate for each industry, shown in Figure 5. The job openings rate is defined as the total number of job openings divided by the sum of job openings and employment in the industry. Over the last fifteen years, the job openings rate has increased in nearly all industries, with the largest increases in professional and business services, construction, manufacturing, and trade, transportation and utilities. This suggests that employers have become more willing to hire. With a similar increase in separations and quits, employees have become more confident about leaving their jobs and looking for new positions.




The regional distribution of the job openings rate is shown in Figure 6. The chart suggests generally higher job openings rates in the South, West and Midwest than the Northeast.





As job openings have increased, employment has naturally followed a similar trajectory. Employment has increased after dropping during the Great Recession and is now above pre- recession levels. The chart below shows the distribution of workers by industry. In 2016, the largest sources of employment in the private sector were in trade, transportation and utility industries, followed by professional and business services and health care and social assistance.





As previously noted, the industries with the highest rates of job openings are in healthcare, professional and business services, and leisure and hospitality.


The BLS projects that the fastest growing occupations between 2014 and 2024 are in health care.x The aging of the population will contribute to additional demand for healthcare workers, and healthcare employment will grow by 26 percent between now and 2024, an increase of over 4 million jobs. This growth includes 296,000 physicians and surgeons and over 1 million registered nurses. Moreover, because healthcare jobs require personal interaction with patients, this industry is not susceptible to automation or offshoring. For this industry, completing at minimum a bachelor's degree program is key. On the job training is also essential.xi

The construction industry is also projected to expand substantially. According to the BLS, the most rapid growth occupations in construction are mechanical insulation workers, helpers of brick masons, block masons, stonemasons, and tile and marble setters. While many occupations in manufacturing are expected to see declining employment, manufacturing linked to construction is expected to grow. Such work will be much more high-skilled since it requires working with high-tech machines.

In the next section of my testimony, I will look at the reasons for the current gap between job seekers and vacancies. Given that across the country, there are 5.6 million job openings and about 7.5 million unemployed people, why is the matching between workers and jobs not taking place? Further, given what we know about the demand for future workers, what policies could help place workers in these jobs?

V. Skills Gap

Over the course of the last several decades, there has been a shift away from middle-skill jobs to a polarization between high- and low-skill jobs. The decline in middle skill jobs is largely attributed to an increase in the automation of tasks, scarcity of workers with appropriate skills, and to some extent relocation of jobs overseas.xii While middle-skill jobs are described as routine in nature and often involve manual work, high-skill jobs are non-routine and cognitive, requiring problem-solving skills. One study finds that between 1981 and 2011, the share of middle-skill employment has declined from 58 percent to 44 percent, while the shares of high-skill and low- skill jobs have increased over that period. This is also the finding in Jaimovich and Siu (2012). The employment share of non?routine cognitive occupations increased from 29% to 39%, the share of non?routine manual occupations increased from 13% to 19%, and the shares in routine cognitive skill occupations decreased from 58% to 44%.xiii The loss of these middle-skill jobs requires that workers either upgrade their skills so that they can apply for high-skill jobs, or they are left to work at low-skill jobs.

At the same time, since jobs are becoming more skill-oriented, the demand for workers with these skills is rising. The skills gap is simply the idea that job openings exist for workers with certain skills, but employers are unable to find or match with these workers.

The notion of a skills gap in the workplace is now several decades old. In a survey conducted in the late 1990s, Lynch and Black (1998) reported that 53% of non-manufacturing workers and 46% of manufacturing employers said that skills associated with their work had increased over the period 1993-1996.xiv This appeared related to the diffusion of computerization and the reorganization of work to increase employee involvement in problem solving and decision making. More than a third of U.S. employers reported that 25% or more of their workers were not fully proficient in their current job.

Workers themselves appear aware of this problem. Leuven and Oosterbeek (1999) found that one in four workers reported they were under-trained.xv More recently, a McKinsey study found that 30% of the graduates surveyed felt that they were underprepared by their college education for the "world of work"; and many employers see these new hires as poorly prepared for life on the job.xvi New research has also shown that the proportion of the labor force employed in occupations that make intensive use of non-routine skills has increased significantly while those employed in routine intensive occupations has declined. Similar results from the 2013 OECD Survey of Adult Skills has shown that the employment share of occupations with minimal skills has declined over the last couple of decades while the demand for relatively more complex cognitive skills has increased.xvii

With changes in technology, there has to be a simultaneous transformation in worker skills to implement and operate the new technology and business models. I will begin by taking a look at manufacturing. According to a study on the Skills Gap by the Manufacturing Institute and Deloitte, over the next decade nearly three and a half million manufacturing jobs will need to be filled, and 2 million of these will remain unfilled due to the skills gap.xviii In surveys, manufacturing executives reported that six out of ten manufacturing positions remain unfilled due to the talent shortage. Eighty percent of executives reported that this was despite their willingness to pay more than the market rates in areas with larger gaps. Additionally, executives reported that it takes an average of 94 days to recruit people for the engineer/scientist field and an average of 70 days to recruit skilled production workers. The types of skills that employers report lacking in the workforce are technology, computer and technical training skills. This is followed by a lack of basic problem solving and math skills. The issue, according to a recent study in the Journal of Economic Perspectives, is that manufacturing in general has moved away from low-skill to high skill jobs.xix

But the skills gap is not just about workers not having the right skills. An interesting problem with manufacturing is what I have termed the image gap. Many younger workers are unwilling to take up jobs in manufacturing because their perception of these jobs is tainted by their image of what factory jobs used to be several decades ago. As I have written earlier, many workers are no longer interested in manufacturing jobs and there appears to be a stigma attached to manufacturing work.xx A survey on the Public Perception of Manufacturing shows that while most Americans perceive manufacturing as the backbone of a strong domestic economy, few parents want their children to work in this industry, and manufacturing is the last career choice for people between the ages of 19 and 33.

To overcome this challenge, companies are increasingly using apprenticeship programs to recruit workers at younger ages from community colleges and high schools. There is also an effort to target workers who would traditionally face discrimination in the labor market, including the long-term unemployed, single mothers or ex-convicts. Companies are using innovative means to attract millennials and convince them that manufacturing jobs are high-tech, well-paying and challenging, not "dirty, grimy, repetitive or dangerous".

All of this suggests that the problem in manufacturing is not the lack of available jobs, but something bigger that requires a two-pronged approach. We must encourage workers to upgrade their skills with training in math, science and computing. For younger workers, paid apprenticeships with companies could produce big results. But aside from the skills gap, we also need to tackle the "image-gap"-the unwillingness of some workers to take up these jobs because of their inherent bias against working in jobs that they perceive as similar to the factory jobs of the past.

This is generally true for many skilled trades. As Mike Rowe has highlighted through his work and writings, many younger workers view jobs in the skilled trades as dirty jobs when many of them pay well and offer a good life.xxi Manufacturing employs about 60% of all skilled trade workers, another 17% are in construction.xxii Traditionally, baby boomers have occupied the majority of these skilled trade positions, but the majority of younger workers are in non- vocational careers.xxiii Hence the demand for these jobs among younger workers is declining.

VI. Workforce of the Future: Policy Response

It is evident that the nature of work is changing. The development of automation, enabled by technologies including robotics and artificial intelligence, has allowed for higher productivity and possibly earnings for certain workers. However, it has also posed a larger challenge about the broader impact on jobs, skills, wages, and the nature of work itself. Some recent studies suggest that about 45 to 47 percent of jobs in the U.S. are susceptible to automation.xxiv The impact of this on jobs and wages is likely negative. In a new paper, authors suggest that industrial robots that replace manual labor like welding, painting, assembling or packaging have increased four-fold between 1993 and 2007 and are likely to continue to increase significantly over the next decade. The effect on jobs would be to reduce about 3-6 workers for every new robot and wages would go down by about 0.5 percent.xxv We must encourage individuals to upgrade their skills so that they can complements the jobs being done by robots more easily, rather than have their jobs be automated away.

For younger workers, paid apprenticeship programs are key. Experience in states like South Carolina suggests that these types of programs are beneficial by allowing both employers and workers to invest in the type of training that is needed. The advantage of apprenticeship programs is that they ease the transition from school to work and follow an earn-while-you-learn model. While not used extensively in the U.S., where apprentices make up only 0.3% of the total workforce, cross- country evidence from many European countries suggests that such programs boost worker earnings, improve productivity, allow apprentices to achieve better job matches and typically allow shorter periods of unemployment before the first job. Moreover, the skills are transferable across firms. For the U.S., a survey by the Department of Labor of the Registered Apprenticeship program found that participants in the program had substantially higher earnings than similar non-participants. Further, economists David Neumark and Donna Rothstein find that internship and apprenticeship programs could help improve labor market outcomes particularly for disadvantaged youth and those in the "forgotten half" less likely to go to college.xxvi

Equally important are investments in basic math and reading. As reported in Hulten and Ramey, the 2013 National Assessment of Educational Progress (NAEP), a "report card" on the math and reading skills of some 92,000 American 12th graders, found that only 26% scored in the proficient or advanced range in math, and only 37% in reading. Further, 35% of 12th graders scored in the lowest proficiency category, "below basic", in the math assessment and 25% in reading. This is troubling when thinking about the preparation necessary for success in the emerging knowledge-based economy.

Soft skills, such as the ability to interact patiently with people and work collaboratively, self- discipline, perseverance, and employability skills such as showing up to work on time, are an equally important part of the process of skill upgradation.xxvii Again, these cannot be automated away and will become more important as the availability of certain types of jobs, particularly in healthcare, grows over time.

Technology has also been and will continue to be a boon in some aspects of the labor market. Job- matching sites such as LinkedIn and Monster are changing and expanding the ways individuals look for work and companies identify and recruit talent. Independent workers are increasingly choosing to offer their services on digital platforms such as Upwork, Uber, and Etsy. Many workers are in contingent work or digital platforms. A recent survey by McKinsey suggests that these platform jobs are helping workers supplement incomes, gain more flexibility and feel like their own boss. It seems likely that this type of contractual or alternative work arrangement will grow over time.

Other policies likely to help workers are the adoption of a paid family leave policy that allows workers, especially mothers, to take time off at the birth of a child or for own illness or caregiving. Research suggests that the lack of these types of family-friendly policies explains one-third of the gap between the US and the OECD in women's labor force participation. Under the current system, only 13 percent of workers have access to paid leave through their employers and the most disadvantaged workers are the least likely to get access to paid leave. I have written extensively about this and am also directing the AEI-Brookings Working Group on Paid Family Leave.xxviii Such a policy will help working families balance work and family responsibilities.

Finally, we need to engage in skills training and re-entry programs for the 650,000 men and women who leave prison and return home every year but have little by way of resources to help them cope with life outside prison.xxix Their history of incarceration precludes them from applying for jobs in many cases or actually obtaining one. They are excluded from Pell Grants which prevents them from getting federal funds for returning to school. Finally, many types of assistance programs are also unavailable to these individuals which impedes their assimilation back into society. As a result, by some estimates, 50 percent to 70 percent of these individuals are back in prison five years after their release.

VII. Conclusion

To conclude, there is a significant amount of work that policymakers need to do to enable the current labor force to access the workforce opportunities of the future. Closing the skills gap is key to this challenge. While we often hear talk about how America is losing jobs to workers from overseas, it is important to remember that there are still a lot of jobs being created in America. The difference is that the nature of those jobs has changed. They are more skill-intensive and often require workers to operate high-tech machinery and have computing and problem solving skills. The skills gap requires upgrading the skills of the existing workforce so that they can more easily match into these jobs.

At the same time, there are certain workers who are being left out of the opportunities available even today, perhaps because of their history of incarceration, or because their disability status prevents them from working, or because they are unable to access benefits such as paid leave that would allow them to take much needed time off and then return to work to continue on their career path. Addressing these current challenges through policy is critical as well.


i Bureau of Labor Statistics, "The Employment Situation," news release, March 10, 2017.

ii Bureau of Labor Statistics, "Unemployment rates for persons 25 years and older by educational attainment," March 10, 2017,

iii Pew Research Center, "The Rising Cost of Not Going to College," February 11, 2014.

iv Stijn Broecke, Glenda Quintini and Marieke Vandeweyer, "Wage Inequality and Cognitive Skills: Re-Opening the Debate," NBER Working Paper No. 21965, February 2016.

v Drew Desilver, "Black unemployment rate is consistently twice that of whites," Pew Research Center Fact Tank, August 31, 2013.

vi Judith Hellerstein, David Neumark and Melissa McInerney, "Spatial Mismatch or Racial Mismatch?" NBER Working Paper No. 13161, June 2007.

vii Bureau of Labor Statistics, "Table A-6. Employment status of the civilian population by sex, age, and disability status, not seasonally adjusted," March 10, 2017,

viii Aparna Mathur, "Poor Work Incentives in Disability: What To Do," Forbes, September 22, 2016.

ix Bureau of Labor Statistics, "Employment Characteristics of Families - 2015," news release, April 22, 2016.

x Bureau of Labor Statistics, "Table 1.3 Fastest growing occupations, 2014 and projected 2024," April 18, 2016,

xi Charlotte Oslund, "Which industries need workers? Exploring differences in labor market activity," Bureau of Labor Statistics Monthly Labor Review, January 2016.

xii Demetrio Scopelliti, "Middle-skill jobs decline as U.S. labor market becomes more polarized," Bureau of Labor Statistics Monthly Labor Review, October 2014.

xiii Nir Jaimovich and Henry E. Siu, "The Trend is the Cycle: Job Polarization and Jobless Recoveries," NBER Working Paper No. 18334, August 2012.

xiv Lisa M. Lynch, "Development Intermediaries and the Training of Low-Wage Workers," in Emerging Labor Market Institutions for the Twenty-First Century, ed. Rchad B. Freeman, Joni Hersch and Lawrence Mishel (University of Chicago Press, 2005).

xv Edwin Leuven and Hessel Oosterbeek, "The Demand and Supply of Work-Related Training: Evidence from Four Countries," Research in Labor Economics 18 (1999): 303-330.

xvi Charles Hulten and Valerie Ramey, "Skills, Education, and U.S. Economic Growth: Are U.S. Workers Being Adequately Prepared for the 21st Century World of Work?" in Education, Skills, and Technical Change: Implications for Future U.S. GDP Growth, ed. Charles Hulten and Valerie Ramey (National Bureau of Economic Research Book Series, forthcoming).

xvii Ibid.

xviii Craig Giffi et al., "The skills gap in U.S. manufacturing: 2015 and beyond," Manufacturing Institute, 2015.

xix Marcel P. Timmer et al., "Slicing Up Global Value Chains," Journal of Economic Perspectives 28, no. 2 (Spring 2014): 99-118.

xx Aparna Mathur, "Making Manufacturing Great Again Will Require A Two-Pronged Approach," Forbes, December 21, 2016.

xxi Mike Rowe, "5.6 Million Reasons to Stop Ignoring the Skills Gap," blog, February 15, 2016,

xxii Joshua Wright, "America's Skilled Trades Dilemma: Shortages Loom As Most-In-Demand Group Of Workers Ages," Forbes, March 7, 2013.

xxiii Adecco, "Vocational skills - skilled trades are in demand as boomers retire," October 4, 2016,

xxiv Daron Acemoglu and Pascual Restrepo, "Robots and Jobs: Evidence from US Labor Markets," NBER Working Paper No. 23285, March 2017.

xxv Ibid.

xxvi Aparna Mathur, "Apprenticeships as the pathway to entrepreneurship for Millennials," Forbes, March 11, 2015.

xxvii Hulten and Ramey, "Skills, Education, and U.S. Economic Growth: Are U.S. Workers Being Adequately Prepared for the 21st Century World of Work?"