Maine Data Glimpse: Underestimates of the Labor Force

Analysis from Jessica Carson, Ph.D. and Sarah Boege at Carsey School of Public Policy at the University of New Hampshire and commissioned by the John T. Gorman Foundation.

This Data Glimpse focuses on an important measurement issue: how the Bureau of Labor Statistics collects official labor force statistics and how these measurements are complicated by the pandemic labor market context. Inspired by the unique uptick in Maine’s July unemployment rate (see Figure 1), we trace how underestimates of the labor force may be artificially suppressing unemployment rates during the pandemic.

Unemployment Rate

Since the COVID-19 pandemic did not really hit the United States until mid-March 2020, unemployment rates in January and February 2020 represent the pre-pandemic economic situation (as shown in Figure 1). For many states, including Maine, even the March 2020 unemployment rate was consistent with pre-pandemic levels. However, April brought historic spikes in unemployment both in Maine and across the nation. April has represented the peak of unemployment nationwide, and in each month since, unemployment rates have been trending downward (although still high compared with pre-pandemic). In contrast to this broad trend, Maine saw unemployment rise between June and July, up more than three percentage points in the month (6.7 to 9.9 percent).


Labor force data suggest the driver of Maine’s uptick was twofold. First, On-the-ground changes in the labor market were partly responsible for this increase. Chiefly, a shipbuilders’ strike in Bath contributed to the loss of 3,500 jobs in the Manufacturing sector from June to July.1 These job losses represent more than one-in-seven of the newly unemployed Mainers in July. Second, Maine’s civilian labor force increased from 669,900 in June to 698,000 in July, which altered
counts of both total workers and unemployed workers used in the official unemployment calculations.2 This phenomenon is explored in more detail below.

Definitions and Measurement

The U.S. Bureau of Labor Statistics (BLS) collects labor force statistics through the Current Population Survey (CPS). Survey respondents are asked basic demographic questions as well as information about their employment situation over the “reference week,” which is usually the previous week. Unemployment may be colloquially defined as “those who do not have a job,” but the technical definition is more nuanced. To be considered “unemployed” a person must not have a job and (1) be available for work and (2) have engaged in at least one active job search activity in the last four weeks. To be considered ‘available for work,” a person must be able to start a job if they were offered one. (For example, a parent without child care might be unavailable for work). Active job search activities include interviewing with employers, contacting an employment agency or employment center, submitting resumes, filling out job applications, or otherwise actively searching for a job.3 An exception to the active job search requirement is made for temporarily laid-off workers who expect to be recalled to work; these workers are considered unemployed, even if they are not searching for another job. If the respondent does not have a job but is either not available for work or has not been actively looking for a job, they are considered “not in the labor force,” and excluded from unemployment calculations (see below):

In usual times, labor force nonparticipation encompasses retirees, stay-at-home parents, people with disabilities, volunteers, interns, and people enrolled in school or training without accompanying employment (among others).4 In the present context, this also includes people who are ill, those who are precluded from work by health risks, those caring for sick family members, or those caring for children. The relationships between work activity, availability, and official labor force status are summarized in Figure 2, below.

Maine’s Labor Force

The size of Maine’s civilian labor force was consistent between January and March 2020 but decreased drastically in April (as shown in Figure 3). Maine’s labor force saw minor changes in May and June, before a big increase in July returned the labor force size to pre-pandemic levels. Given BLS definitions, the April labor force constriction was likely due to a combination of people who were no longer available to work, and those who were not working and available to work but had not yet begun actively looking for a job.

Although not reliably available for Maine specifically, national BLS data identifies people’s reasons for not being in the labor force. Survey respondents who want a job but are not working and have not looked for work in the past four weeks are asked to select their main reason for not looking for work. These options included items like “believes no work available in area of expertise,” “family responsibilities,” and “ill-health, physical disability.” Survey respondents in April through July 2020 indicate remarkably varied answers to this question, with no clearly dominant response, aside from “other,” selected by 45% of respondents. The pandemic was not included on the list of options, so it is possible that pandemic-induced factors such as concerns over personal health and safety might be captured in this “other” category. That some feature of the pandemic drove this response is supported by the fact that the number of “other” responses spiked in April and has decreased into July. To better understand the impact of the pandemic on the labor market specifically, the BLS added five questions to the Current Population Survey starting in May,5 asking people specifically whether they were not looking for work because of the pandemic. Unlike the questions discussed above, these items are asked of all people not working and not looking for work, regardless of whether they want a job. Nationwide, 9.5 percent of those not in the labor force did not look for work because of the pandemic; by July, that figure had fallen to 6.5 percent. This supports the idea that more people were leaving the labor force due to COVID19 and its impacts towards the beginning of the pandemic.


Although Maine’s July unemployment rate is higher than that of May or June, there is cause for optimism. For one, it is possible that the April, May, and June unemployment rates are artificially low because fewer Mainers were in the labor force to be counted as unemployed. In this case, the July unemployment rate may be much less of an unemployment spike and more of a reflection of underlying conditions already present since spring. Since the July labor force has reached pre-pandemic levels again, the July unemployment rate is likely more accurate. Additionally, the return of the labor force to pre-pandemic levels in itself can be interpreted as a positive sign. During earlier months of the pandemic, Mainers may have dropped out of the labor force due to a perception that the labor market was so bad it did not make sense to look for a job, a lack of child care, family responsibilities, poor health, concerns about safety at work, and lots of other reasons. The increase in the labor force may reflect that more Mainers feel it is worth actively looking for a job again and have the supports and ability to do so.


1 As discussed here The strike has since ended on August 23
with the approval of a three-year contract, as detailed here
2 As discussed here
3 Additional details available at
4 Additional details available at