Academic journal article Review - Federal Reserve Bank of St. Louis

Industrial and Occupational Employment Changes during the Great Recession

Academic journal article Review - Federal Reserve Bank of St. Louis

Industrial and Occupational Employment Changes during the Great Recession

Article excerpt

(ProQuest: ... denotes formulae omitted.)


The Great Recession left a deep, persistent mark on the U.S. labor market. By October 2009, the unemployment rate shot up to 10 percent from its pre-recession level of 5 percent in as late as April 2008. The recovery was painfully slow-it was not until October 2015 that the unemployment rate returned to its pre-recession level. Other measures of the labor market-for example, total hours worked and payroll-tell the same story.

There has been no shortage of commentary and academic papers on the slow recovery of the labor market that followed the Great Recession. But most focus on aggregate, or economywide, statistics. Some do explore changes in employment across industrial sectors, but few attempt to dissect employment fluctuations across occupations.1 In this paper, we consider employment changes across industries and occupations jointly, aiming to provide further insights on the labor market dynamics during and after the Great Recession.2

We use the American Community Survey (ACS), which has a large enough sample size that enables a detailed industrial and occupational level analysis of the labor market. One shortcoming is that it is annual data and hence not suitable for higher-frequency (i.e., monthly and quarterly) business-cycle analysis. However, for our study of the Great Recession, which is a big, discontinuous event, annual data may still produce a conspicuous pattern. Indeed, our analysis finds a striking, hitherto undocumented pattern.

We find that the sharp contraction in the labor market during the Great Recession was largely driven by the construction sector, which shed employment across all occupations. This was only partially mitigated by rising employment in the food services, education, and health industries, which added workers and hours across all occupations.

In contrast, the recovery since 2010 is not led by particular industries but by low-skill occupations (between 2010 and 2012) and high-skill occupations (after 2012) across all industries. This pattern of recovery is a continuation of the underlying trend of polarization across occupations-employment moving away from middle-skill occupations to low- and high-skill ones-that started in the 1980s.3

This "contraction by sectors and expansion by occupations" asymmetry has never been documented and is a worthwhile observation in its own right. We conjecture, however, that the sharp contraction along the sector dimension and the ensuing slow rise along the occupation dimension will eventually offer an insight into the slow recovery from the Great Recession and, more broadly, the negative skewness of business cycle fluctuations (i.e., sudden recessions and slow recoveries).

The rest of the paper is organized as follows. We describe the data and define the employment shares of interest in Section 2. In Section 3, we show employment changes at the occupation level and decompose them into between- and within-sector components. In Section 4, we show employment changes across industrial sectors. We follow up in Section 5 with a detailed look at year-to-year changes in employment for selected sectors, breaking down a sector's employment change by occupations of different skill levels. Finally, we summarize our findings and discuss avenues for future research in Section 6.


Employment shares by occupation and industry are constructed from the 2000 Census and the 2001 to 2015 ACSs. We restrict our sample to 16 to 64 year olds who report themselves as employed at the time of each survey. We measure employment in each industryoccupation combination by summing up the hours worked (over a year) of all workers in a given industry-occupation cell.4 That is, by employment, we mean total hours worked-or, equivalently, hours per worker times the number of workers. Hence in our analysis, a fall in employment occurs along both the intensive (change in hours only) and extensive (change in the number of workers) margins. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.