New EPI Economic Indicator: Monthly Updates of the Number of “Missing Workers” and What the Unemployment Rate Would Be If They Were Looking for Work

More than four years since the Great Recession officially ended in June 2009, the unemployment rate stands at 7.3 percent. This is still a percentage point above the highest unemployment rate of the early 2000s downturn, 6.3 percent. However, 7.3 percent is a big improvement from the high of 10.0 percent in the fall of 2009. Unfortunately, most of that improvement was for all the wrong reasons.

In today’s labor market, the unemployment rate drastically understates the weakness of job opportunities. This is because in the weak labor market of the aftermath of the Great Recession, there are a huge number of “missing workers”—potential workers who are neither employed nor actively seeking work simply because job opportunities remain so scarce. Because jobless workers are only counted as unemployed if they are actively seeking work, these missing workers are not reflected in the unemployment rate.

As part of its ongoing effort to create the metrics needed to assess how well the economy is working for America’s broad middle class, EPI is introducing its “missing workers” estimate. Our estimate shows there are currently nearly 5 million missing workers. These are workers who would be in the labor force if job opportunities were significantly expanded but, given the state of the labor market, are sidelined.

Exactly how many missing workers macroeconomic policymakers believe there are has enormous implications for their assessment of the strength of the job market, and therefore for their policy decisions. For example, if they underestimate the number of missing workers, they will overstate the strength of the labor market, and be less likely to provide the economy with the support it needs. As shown in the figure below, if the nearly 5 million missing workers were looking for work and thus counted as unemployed, the unemployment rate in August would have been 10.1 percent instead of 7.3 percent.

Missing Workers

The unemployment rate is vastly understating weakness in today’s labor market: Unemployment rate, actual and if missing workers* were looking for work, January 2006–July 2014

Date Actual If missing workers were looking for work
2006-01-01 4.7% 5.0%
2006-02-01 4.8% 4.8%
2006-03-01 4.7% 4.8%
2006-04-01 4.7% 4.9%
2006-05-01 4.6% 4.8%
2006-06-01 4.6% 4.7%
2006-07-01 4.7% 4.8%
2006-08-01 4.7% 4.6%
2006-09-01 4.5% 4.6%
2006-10-01 4.4% 4.4%
2006-11-01 4.5% 4.4%
2006-12-01 4.4% 4.1%
2007-01-01 4.6% 4.4%
2007-02-01 4.5% 4.4%
2007-03-01 4.4% 4.3%
2007-04-01 4.5% 4.9%
2007-05-01 4.4% 4.8%
2007-06-01 4.6% 4.8%
2007-07-01 4.7% 4.9%
2007-08-01 4.6% 5.1%
2007-09-01 4.7% 4.9%
2007-10-01 4.7% 5.2%
2007-11-01 4.7% 4.9%
2007-12-01 5.0% 5.1%
2008-01-01 5.0% 4.8%
2008-02-01 4.9% 5.0%
2008-03-01 5.1% 5.1%
2008-04-01 5.0% 5.2%
2008-05-01 5.4% 5.4%
2008-06-01 5.6% 5.6%
2008-07-01 5.8% 5.7%
2008-08-01 6.1% 6.0%
2008-09-01 6.1% 6.3%
2008-10-01 6.5% 6.5%
2008-11-01 6.8% 7.1%
2008-12-01 7.3% 7.5%
2009-01-01 7.8% 8.2%
2009-02-01 8.3% 8.7%
2009-03-01 8.7% 9.3%
2009-04-01 9.0% 9.4%
2009-05-01 9.4% 9.7%
2009-06-01 9.5% 9.9%
2009-07-01 9.5% 10.1%
2009-08-01 9.6% 10.4%
2009-09-01 9.8% 10.9%
2009-10-01 10.0% 11.3%
2009-11-01 9.9% 11.2%
2009-12-01 9.9% 11.7%
2010-01-01 9.7% 11.3%
2010-02-01 9.8% 11.4%
2010-03-01 9.9% 11.3%
2010-04-01 9.9% 11.0%
2010-05-01 9.6% 11.1%
2010-06-01 9.4% 11.1%
2010-07-01 9.5% 11.3%
2010-08-01 9.5% 11.1%
2010-09-01 9.5% 11.3%
2010-10-01 9.5% 11.5%
2010-11-01 9.8% 11.7%
2010-12-01 9.4% 11.6%
2011-01-01 9.1% 11.4%
2011-02-01 9.0% 11.4%
2011-03-01 9.0% 11.3%
2011-04-01 9.1% 11.4%
2011-05-01 9.0% 11.4%
2011-06-01 9.1% 11.5%
2011-07-11 9.0% 11.7%
2011-08-20 9.0% 11.4%
2011-09-01 9.0% 11.3%
2011-10-11 8.8% 11.2%
2011-11-20 8.6% 11.0%
2011-12-30 8.5% 11.0%
2012-01-12 8.2% 10.8%
2012-02-12 8.3% 10.7%
2012-03-12 8.2% 10.7%
2012-04-12 8.2% 10.9%
2012-05-12 8.2% 10.6%
2012-06-12 8.2% 10.5%
2012-07-12 8.2% 10.8%
2012-08-12 8.1% 10.8%
2012-09-12 7.8% 10.4%
2012-10-12 7.8% 10.0%
2012-11-12 7.8% 10.3%
2012-12-12 7.9% 10.3%
2013-01-12 7.9% 10.4%
2013-02-12 7.7% 10.5%
2013-03-12 7.5% 10.6%
2013-04-12 7.5% 10.5%
2013-05-12 7.5% 10.3%
2013-06-12 7.5% 10.3%
2013-07-12 7.3% 10.2%
2013-08-12 7.2% 10.3%
2013-09-12 7.2% 10.3%
2013-10-12 7.2% 10.7%
2013-11-12 7.0% 10.3%
2013-12-12 6.7% 10.2%
2014-01-12 6.6% 10.0%
2014-02-12 6.7% 10.0
2014-03-12 6.7% 9.8%
2014-04-12 6.3% 9.9%
2014-05-12 6.3% 9.7%
2014-06-12 6.1% 9.6%
2014-07-12 6.2% 9.6%

* Potential workers who, due to weak job opportunities, are neither employed nor actively seeking work

Source: EPI analysis of Mitra Toossi, “Labor Force Projections to 2016: More Workers in Their Golden Years,” Bureau of Labor Statistics Monthly Labor Review, November 2007; and Current Population Survey public data series

Show table Copy data to Excel Embed this chart

Estimating the number of missing workers is not straightforward because some changes in labor force participation over the last five years have nothing at all to do with the weak labor market (for example, baby boomers beginning to reach retirement age). Our estimate of the number of missing workers isolates the cyclical component of the decline in the labor force participation rate since the start of the Great Recession. In other words, it counts just those missing workers who would be in the labor force if job opportunities were strong. It doesn’t count, for example, those retiring baby boomers who would have left the labor force whether or not the Great Recession happened.1

We will update these estimates on the first Friday of every month immediately after the Bureau of Labor Statistics releases the monthly jobs numbers. In particular, we will update the following three figures each month at EPI’s Missing Workers page:

1. The trend in the total number of missing workers, currently nearly 5 million:

Missing Workers

Millions of potential workers sidelined: Missing workers,* January 2006–July 2014

Date Missing workers
Jan-2006 530,000
Feb-2006 110,000
Mar-2006 110,000
Apr-2006 250,000
May-2006 210,000
Jun-2006 110,000
Jul-2006 60,000
Aug-2006 -120,000
Sep-2006 120,000
Oct-2006 -50,000
Nov-2006 -220,000
Dec-2006 -500,000
Jan-2007 -460,000
Feb-2007 -210,000
Mar-2007 -150,000
Apr-2007 650,000
May-2007 560,000
Jun-2007 360,000
Jul-2007 370,000
Aug-2007 840,000
Sep-2007 410,000
Oct-2007 800,000
Nov-2007 280,000
Dec-2007 250,000
Jan-2008 -320,000
Feb-2008 220,000
Mar-2008 50,000
Apr-2008 340,000
May-2008 -60,000
Jun-2008 20,000
Jul-2008 -70,000
Aug-2008 -90,000
Sep-2008 180,000
Oct-2008 60,000
Nov-2008 420,000
Dec-2008 420,000
Jan-2009 710,000
Feb-2009 620,000
Mar-2009 1,050,000
Apr-2009 750,000
May-2009 650,000
Jun-2009 650,000
Jul-2009 1,040,000
Aug-2009 1,320,000
Sep-2009 2,050,000
Oct-2009 2,270,000
Nov-2009 2,300,000
Dec-2009 3,120,000
Jan-2010 2,770,000
Feb-2010 2,680,000
Mar-2010 2,460,000
Apr-2010 1,940,000
May-2010 2,510,000
Jun-2010 2,960,000
Jul-2010 3,210,000
Aug-2010 2,830,000
Sep-2010 3,200,000
Oct-2010 3,570,000
Nov-2010 3,340,000
Dec-2010 3,830,000
Jan-2011 3,950,000
Feb-2011 4,080,000
Mar-2011 3,960,000
Apr-2011 4,020,000
May-2011 4,070,000
Jun-2011 4,220,000
Jul-2011 4,650,000
Aug-2011 4,130,000
Sep-2011 3,970,000
Oct-2011 4,010,000
Nov-2011 4,150,000
Dec-2011 4,230,000
Jan-2012 4,490,000
Feb-2012 4,120,000
Mar-2012 4,220,000
Apr-2012 4,690,000
May-2012 4,190,000
Jun-2012 4,070,000
Jul-2012 4,540,000
Aug-2012 4,690,000
Sep-2012 4,480,000
Oct-2012 3,840,000
Nov-2012 4,400,000
Dec-2012 4,180,000
Jan-2013 4,370,000
Feb-2013 4,700,000
Mar-2013 5,240,000
Apr-2013 5,130,000
May-2013 4,780,000
Jun-2013 4,710,000
Jul-2013 5,050,000
Aug-2013 5,230,000
Sep-2013 5,380,000
Oct-2013 6,060,000
Nov-2013 5,710,000
Dec-2013 6,100,000
Jan-2014 5,850,000
Feb-2014 5,660,000
Mar-2014 5,290,000
Apr-2014 6,220,000
May-2014 5,950,000
Jun-2014 5,980,000
Jul-2014 5,860,000

* Potential workers who, due to weak job opportunities, are neither employed nor actively seeking work

Note: Volatility in the number of missing workers in 2006–2008, including cases of negative numbers of missing workers, is simply the result of month-to-month variability in the sample. The Great Recession–induced pool of missing workers began to form and grow starting in late 2008.

Source: EPI analysis of Mitra Toossi, “Labor Force Projections to 2016: More Workers in Their Golden Years,” Bureau of Labor Statistics Monthly Labor Review, November 2007; and Current Population Survey public data series

Show table Copy data to Excel Embed this chart

2. The breakdown of missing workers by gender and age, showing most missing workers are of prime working age:

Missing Workers

Roughly half of missing workers are of prime working age: Missing workers,* by age and gender, July 2014

 

Missing workers
Men under 25 630,000
Women under 25 420,000
Men 25–54 1,850,000
Women 25–54 1,380,000
Men 55+ 580,000
Women 55+ 1,000,000

* Potential workers who, due to weak job opportunities, are neither employed nor actively seeking work

Source: EPI analysis of Mitra Toossi, “Labor Force Projections to 2016: More Workers in Their Golden Years,” Bureau of Labor Statistics Monthly Labor Review, November 2007; and Current Population Survey public data series

Show table Copy data to Excel Embed this chart

3. The earlier figure depicting what the unemployment rate would be if the missing workers were looking for jobs.

Methodology

1. How do we estimate the number of missing workers? Labor force participation rate projections published by the Bureau of Labor Statistics in November 2007—before the start of the Great Recession—are available in Table 3 of Mitra Toossi, “Labor Force Projections to 2016: More Workers in Their Golden Years,” Bureau of Labor Statistics Monthly Labor Review, November 2007. The projections assumed a healthy labor market over the period in question, 2006–2016, so the participation rate changes it forecasts reflect purely non-cyclical factors (e.g., the impact of retiring baby boomers). The difference between these projections and the actual labor force participation rate is thus a good measure of the cyclical change in the labor force participation rate, i.e., the change that is a direct result of the weak labor market in the Great Recession and its aftermath. Based on this logic, missing workers are estimated in the following way: The labor force participation rate projections for 2016 by gender and age group (age groups 16–19, 20–24, 25–34, 35–44, 45–54, 55+) available in Table 3 of Toossi (2007) are assumed to be structural rates. The current month’s structural rates (by gender and age group) are calculated by linearly interpolating between 2006 and 2016. The size of the potential labor force is calculated by multiplying the current month’s structural rates by actual population numbers (available by gender and age group from the Current Population Survey public data series). The difference between the size of the potential labor force and size of the actual labor force (also available by gender and age group from the Current Population Survey public data series) is the number of missing workers.

 

Tagged

  • benleet

    The labor force participation rate for August 2013, 63.2%, was a low not seen since 1978 when it was a historical high, and the employment to population ratio, 58.6%, also is at 1977 level, an historical high at that time. Back to the 1970s. I think that the 67.1% participation ratio of the four years 1997, ’98. ’99. ’00, is possible again. In April 2000 the employment to population ratio, 64.7%, a historical peak, was 6.1% higher than today’s, and if we had that rate today then 14.6 million would be working, not just the 5 million of this article, and total working would be 158.8 million. Full employment used to be the political top priority. Robert Kuttner’s article The Task Rabbit Economy is the unfortunate future of our labor force, unless. That’s his recent article in the American Prospect. He provides a great solution section at the end.

  • benleet

    The labor force participation rate for August 2013, 63.2%, was a low not seen since 1978 when it was a historical high, and the employment to population ratio, 58.6%, also is at 1977 level, an historical high at that time. Back to the 1970s. I think that the 67.1% participation ratio of the four years 1997, ’98. ’99. ’00, is possible again. In April 2000 the employment to population ratio, 64.7%, a historical peak, was 6.1% higher than today’s, and if we had that rate today then 14.6 million would be working, not just the 5 million of this article, and total working would be 158.8 million. Full employment used to be the political top priority. Robert Kuttner’s article The Task Rabbit Economy is the unfortunate future of our labor force, unless. That’s his recent article in the American Prospect. He provides a great solution section at the end.

  • Nicholas Thompson

    This “missing workers” sounds a lot like the definitions of U4 and U5 (subtracting the normal unemployment rate). U4 is discouraged workers plus the normal unemployment rate, and U5 is U4 plus marginally attached workers.

    Contrary to this work, the number of discouraged and marginally attached workers have been dropping since after 2011, not increasing.

    The 2007 BLS projection has a very rosy outlook on future employment, the difference between U5-U3 and the “Missing workers” shown here is an illustration of just how overly optimistic the BLS report actually was.