by Nicholas Dobbins, Sanjukta Chaudhuri, Oriane Casale and Jim Hegman
June 2020
Minnesota's labor market hummed along in 2019, with stable unemployment rates over the period but slowing job growth. Then, in mid-March 2020, a severe shock hit as measures to contain COVID-19 began to be implemented. This summary touches on the state of Minnesota's labor markets prior to March 2020 but focuses on the most recent period for which data are available, March, April and May 2020. This summary relies on data from Current Employment Statistics (CES) conducted by the U.S. Bureau of Labor Statistics (BLS), the Local Area Unemployment Statistics, conducted by the BLS and Unemployment Insurance (UI) data from the Minnesota Unemployment Insurance Program.
Minnesota nonfarm employment growth continued to slow over-the-year in 2019 before dipping well into the negative in early 2020 as the first effects of fallout from the COVID-19 crisis hit the labor market. Preliminary April estimates from the CES program put the statewide over-the-year (OTY) job loss at 13.1%, a decline of 387,894 jobs from April 2019. February employment, which is the last month before we began to see the effects of the pandemic, was up 0.1% on the year, while March employment declined 0.7%, showing the leading edge of pandemic effects. In Minnesota, temporary business closures and other pandemic control measures were instituted during the second half of March. Because the reference period for the nonfarm employment survey is based on the 12thof the month, most March losses did not appear in the CES estimates until April.
The April decline marked the single-worst OTY job loss in the CES series' history, which dates back to 1950. The second-worst OTY job loss came in September 2009 when employment was down by 5%. While unprecedented, the OTY job declines in Minnesota were in line with nationwide estimates, as the country lost 13.2% of jobs OTY in April 2020, also the largest decline on record. Because the Minnesota series only goes back to 1950, and the U.S. to 1939, both miss the effects of the Great Depression, which is likely the best historical analogue for current job losses.
These sharp declines overwhelmed any previous trends in the labor market. Prior to the effects of the pandemic, statewide job growth had been lagging nationwide growth, and had posted gains of less than 1% fairly consistently through 2019 and most of 2018, with a 2019 average growth of 0.5%. Nationwide, employment grew by 1.6% OTY on average in 2019.
The first quarter of 2020 capped what had been over two years of Minnesota consistently underperforming national job growth, due in part to Minnesota's extremely tight labor market constraining job growth. However, it is likely that any overall trends in total nonfarm employment present before April 2020 have been permanently subsumed, and that the month will mark the beginning of a mostly new set of trends in our labor markets.
Large employment losses occurred in most industry groups in April, with OTY declines in every supersector save mining and logging. Service providing industries bore the brunt of the job losses, off by 13.9%OTY, but goods producers were not immune to the effects as they shed 8.5% of jobs.
Leisure and hospitality was the hardest hit supersector, with more than half of all jobs in the industry group lost in April (down 55.5% OTY, and 53.3% over the month), at a time when employment is usually ramping up for the summer. While extreme, the losses were not entirely unexpected, as the supersector contains many of the industries most effected by declines in nonessential activities such as dining at restaurants, traveling and attending sporting events and concerts. Leisure and hospitality had been one of the stronger-performing supersectors prior to the pandemic, with annual employment growth of 1.3% in February.
Other services, which was arguably the strongest-performing supersector in the state in the months leading up to the pandemic, was the second hardest hit in April. OTY growth plummeted from 3.4% in February (the best in the state) to -28.3% in April. Personal and laundry services employment was down 63.2% on the year, with much of that loss likely owing to the temporary closure of hairdressers and other personal care service businesses across the state. Another other services component, religious, grant making, civic, professional, and similar organizations, also saw a steeper-than-normal decline, off by 19.7% on the year.
No geographic region in the state was immune to the job losses. Employment was down in every published metropolitan area in the state, with declines of more than 10% in all MSAs.
While the job losses in Minnesota and across the nation are striking, it is important to note that there are signs that some of the losses may be short-lived. At the time of this writing, preliminary estimates showing Minnesota added 27,500 private sector payroll jobs on a seasonally adjusted basis in May, up 1.3%, during the month. U.S. estimates show that a surprising 2.5 million jobs were added on a seasonally adjusted basis nationwide in May. As we progress through this unprecedented upheaval in the market, myriad factors are likely to affect employment in different industries as individual parts of our economy work to reopen, or are hit by lagging negative consequences of the situation. CES estimates are designed to give us information as quickly as possible, which means that even under ideal circumstances, the information is often subject to change as the data submitted by employers becomes more complete over time. While early estimates provide a helpful snapshot of our labor market, the picture will become much clearer in the coming months, as we continue to gather data and assess the new employment situation in Minnesota.
The year 2019 and the first two months of 2020 were not out of the ordinary for Minnesota's labor market. In 2019, there were an estimated 3.01 million employed people in Minnesota's labor force and 100,492 unemployed people for a total labor force of 3.11 million and an unemployment rate of 3.2% compared to 3.7% nationally. With a civilian non-institutional population of 4.42 million (age 16 and over), the annual average labor force participation rate was 70.3% compared to 63.1% nationally and the employment to population ratio was 68.0% compared to 60.8% nationally. These key labor market statistics were very stable throughout the year on a seasonally adjusted basis.
2020 started on a similar note. The number of employed ticked up to 3.02 million and the number unemployed ticked down to 97,444 for an unemployment rate of 3.2% in January and 3.1% in February on a seasonally adjusted basis. The labor force participation rate ticked down to 70.2% in February while the employment-to-population ratio remained stable at 68.1%.
By mid-March, however, pandemic containment measures disrupted the stable labor market at a speed and scope that was impossible to capture using traditional labor market measures, namely those of the Current Population Survey (CPS). Estimation of these impacts has presented a number of challenges for labor market statistics, mainly driven by the swiftness of the downturn which caused a departure from the empirically stable relationships between labor market variables (see side bar Labor Force Statistics During the Covid-19 Pandemic for a brief overview).
With this backdrop, estimates for April 2020 provide an early glimpse of the labor market situation during the ongoing pandemic. Seasonally adjusted unemployment reached 8.7% (revised) in April. Only one period in the history of the series dating back to the mid-1970s saw higher unemployment rates: During the 1982-83 recession the unemployment rate reached 8.9% in Minnesota and stayed there for 3 months, only gradually dropping under 7.0% by January 1984. The unemployment rate never exceeded 8.0% during the Great Recession in 2008 and 2009. Estimates for May 2020 show Minnesota unemployment at 9.9%, the highest ever recorded in the state.
Increased unemployment was not the only effect of the pandemic on Minnesota's labor market; a growing number of people are leaving the labor force. According to CPS statistics, from February to May, the number employed dropped by 259,946, the number unemployed rose by 206,616 and 53,330 people dropped out of the labor force altogether. Dropping out of the labor force means they were neither working nor were they seeking work or on temporary layoff from a job to which they expected to return. As a result, the labor force participation rate dropped from 70.2% in February to 68.9% in May, while the employment to population ratio plummeted from 67.6% in March to 62.1% in May.
As mentioned in the sidebar, there are specific problems related to labor force estimation due to the pandemic. These include miscoding of layoffs as still employed, an increase in labor underutilization, and a surge in discouraged workers leaving the labor market who are therefore not included in the official unemployment rate. In the context of the unique and unprecedented job losses during this period, it has become even more important than before to follow the alternative measures of unemployment. The main objective of these alternative measures is to capture a broader measure of unemployment including underemployment, unavailability for work and worker discouragement that are not included in the headline unemployment rate.
For example, the CPS reveals that in April, the number of persons who usually work full time and were working full time in Minnesota declined by 166,579 over the month, and the number of workers who usually work part time and were working part time declined by 187,416. The number of persons at work part time for economic reasons more than doubled over the month from 72,951 in March to 191,291 in April. These individuals, who would have preferred full-time employment, were working part time because their hours had been reduced or they were unable to find full-time jobs. This group includes persons who usually work full time and persons who usually work part time but would prefer to work full time.
The number of persons not in the labor force who currently want a job was 119,700 in April, more than double March's estimates of 55,400. These individuals were not counted as unemployed because they were not actively looking for work during the last four weeks or were unavailable to take a job. Persons marginally attached to the labor force, a subset of persons not working or looking for work but who currently want a job, numbered 21,700 in April, up by 10,700 over the month. These individuals wanted and were available for work and had looked for a job sometime in the prior 12 months but had not looked for work in the four weeks preceding the survey.
While the available CPS data does not yet allow us to assess the impact of COVID-19 by various subgroups in the labor market, it is noteworthy that for much of 2019 and early 2020, the various subgroups more or less mirrored the low unemployment rate across Minnesota's labor market. Prior to COVID-19 impacts, for women, the unemployment rate ranged from 2.1% to 2.6%; for men it was 3.6% to 4.0%. By race, the unemployment range was between 2.7% and 3% for Whites, 4.5% and 6.8% for Blacks, and 3.0% and 5.4% for Hispanics. In the coming months, more data will aid in developing a deeper understanding of the labor market impact of COVID-19 on these groups. (An article on employment disparities will be posted soon.)
With the state of Minnesota's labor market changing so rapidly in March 2020, we turned to a timelier source of information to better gauge these changes on a daily and weekly basis: data from our state UI program. There are multiple types of UI data measuring different trends. This article looks at two types of data: applicants and continued claims.
Applicants are a count of people applying for UI. Using UI applications results in an over count of layoffs because these applications have not yet been screened for eligibility. Many will not be determined eligible for benefits, some are not fully completed applications, and some will never ask for payment. In addition, some people apply as a result of a reduction in hours but are still working.
The second UI measure used in this article comes from the weekly benefit request data, also known as continued claims. To receive a UI benefit payment, applicants must not only file an initial application, but must also request benefits weekly. The weekly request process is used to gather earnings and hours worked information along with information about other forms of income that reduce UI payments, such as pension payments. The process is also used to pick up work search activities, job refusals, quits, discharges and layoffs. Weekly benefit requests extend information about the experience of workers from initial job loss through the weeks that follow. Although they are designed to allow for the administration of program eligibility requirements of UI rather than as a survey tool, these data nonetheless provide insight into the experiences of workers until stronger data becomes available.
The first measure, applicants, provides a gauge of how many workers overall self-report that they have lost employment or had their hours reduced. The loss of employment may have been due to a layoff, discharge, quit, leave of absence or suspension imposed by their employer. Beginning the week of March 16, the UI program added an additional category that it called COVID-19 separation. This allows applicants to self-identify their unemployment as having been caused by the pandemic.
Typically, Minnesota workers report their reason for separation accurately. In part this is because the UI program validates that information with their employer(s). Due to the unprecedented surge in new claims, this employer validation has not been practical. It may be supposed that most of those that reported a COVID-19 separation did so accurately. It is probably the case that some applicants also misreported their reason for separation and may either have still been working or were unemployed for reasons unrelated to the pandemic.
Since the week of March 16, when the first COVID-19-related order limiting business operations went into effect, through May 30, we've seen a cumulative 728,369 UI applications. That represents almost one in every four persons in the labor force, however as noted above this does not mean that many people are unemployed.
Use with Caution: Notes on UI Data as a Labor Market GaugeUI data are transactional and administrative in nature: They are defined, collected and compiled to run a complex program, not to create a gauge of economic or labor market activity. Moreover, they are continually in flux – especially during the early periods of a downturn. Finally, most of the data are not cleaned as an economic data series, as are our other sources of data on the labor market. For these reasons, they should be used with caution for the purpose of understanding the extent of employment loss in Minnesota. UI data cannot be used alone to calculate an unemployment rate because not every UI applicant is unemployed – some are working reduced hours but still working, and others have applied erroneously – and not every unemployed person applies for UI. Some people who are unemployed have never had a paid job or have not worked in many months and thus could not qualify for UI. Others quit, or were fired rather than laid off, or for some other reason may or may not qualify for benefits after their separation is adjudicated. Because UI is an eligibility-based program, there is an inviolate right to apply. Any individual can apply for UI whether or not they have ever worked and regardless of the reason for their unemployment. Unemployed workers are discouraged from not applying or "self-selecting out" of UI by not applying. That said, some unemployed workers who might otherwise be eligible simply choose not to apply for a variety of reasons, including personal choice. Currently, people may be eligible for UI if they meet any of the regular eligibility requirements, or if they are on COVID-19 leave either because of their own illness or because they need to care for a family member. In addition, the temporary extra $600 per week granted under the CARES Act fundamentally changes the incentive structure of the program, incenting people to apply for UI who might not qualify or otherwise bother applying, as well as incenting people who could go back to work to stay on UI without giving employers an incentive to get people off UI and back to work. It is also important to note that, with the unprecedented high volume of applications and changes in UI rules during this period, the rigorous application screening that occurs normally was not possible or required. An additional, confounding element of UI applications during this period is the very large presence of individuals not ordinarily found in UI data: self-employed individuals and workers whose employer is not covered by the UI program. During this period, self-employed individuals who have not elected to have their work covered by the UI program and workers who work for employers who have the option of not being covered by UI also applied as part of the process to request benefits from a temporary, emergency UI-like federal program for those not eligible for UI, called "Pandemic Unemployment Assistance" (PUA). For this reason, the number of UI applications is inflated by PUA applicants by as much 100,000. These applicants assert that they are unemployed, but would not ordinarily be part of the UI population and tend to skew current UI data trends when compared to historical UI trends. In summary, during the period examined here, UI applications undoubtedly overestimate the number of layoffs in the economy in part because of the nature of UI and in part because of the changes in law governing UI and a changed incentive structure for potential applicants during the current crisis. |
| Industry Sector | Cumulative UI Applications | Share of Workforce that has Applied for UI Benefits |
|---|---|---|
| Other Services, Ex. Public Admin | 43,592 | 48% |
| Information | 23,135 | 46% |
| Accommodation and Food Services | 107,514 | 46% |
| Arts, Entertainment, and Recreation | 24,220 | 45% |
| Mining | 1,923 | 33% |
| Construction | 38,528 | 29% |
| Retail Trade | 83,945 | 29% |
| Health Care and Social Assistance | 132,570 | 26% |
| Administrative and Waste Services | 35,753 | 26% |
| Transportation and Warehousing | 27,316 | 25% |
| Real Estate and Rental and Leasing | 8,372 | 24% |
| Manufacturing | 73,617 | 23% |
| Management of Companies and Enterprises | 17,942 | 20% |
| Wholesale Trade | 22,258 | 17% |
| Professional and Technical Services | 26,563 | 17% |
| Educational Services | 27,259 | 12% |
| Public Administration | 11,604 | 9% |
| Agriculture, Forestry, Fishing and Hunting | 1,780 | 8% |
| Utilities | 726 | 5% |
| Finance and Insurance | 7,072 | 5% |
The number of UI applications has been highest in health care and social assistance, accommodation and food services and retail trade, three very large sectors. Applications peaked in these industries at different times: In health care and social assistance, applications peaked close to the middle of the period where as for accommodation and food services applications peaked during the first week and in retail trade they peaked during the first month1.
In terms of share of jobs in the industry, applications have been most concentrated in other services (including personal services like hairdressers and barbers), information, accommodation and food services, and arts, entertainment and recreation, sectors heavily impacted by the various executive orders limiting business operations and enforcing social distancing measures, with the exception of information. Nearly 50% workers in these sectors have applied for UI during this period.
By occupation, applications follow the same pattern with 60% of arts and entertainment workers, almost half of personal care and services workers, 40% of food prep and serving workers and 25% of sales workers applying for UI during this period.
There are stark differences in rates of UI application among different demographic groups in Minnesota. Much of this is based on the demographic makeup of industries and occupations in Minnesota. Younger people, who are concentrated in three of the hardest hit sectors: accommodation and food service, arts, entertainment and recreation, and retail trade, have been harder hit by the pandemic measures than older people in Minnesota. Just over 26% of workers under age 25 have applied for UI benefits during this period, with the share at 31.8% for those between 20 and 24. This is the highest of any age group, but all age groups have been impacted, with 24% of workers over age 65 having applied for UI benefits.
By educational attainment, a much lower share of those with a 4-year college degree have applied for UI compared to those with less education. Of the workforce with a high school diploma or less, 42.6% have applied for UI. This group is also concentrated in the industries hardest hit by pandemic containment measures.
One note here is that while significantly more women than men applied for UI during the early weeks of the pandemic containment measures, the number of women applying for UI benefits has slowed more quickly than the number of men applying since then. There are several possible reasons. First, applications in female dominated sectors like health care and social assistance, accommodation and food service and other services peaked early and have now slowed.
Second, UI applications from male-dominated occupations like Construction and Manufacturing are highly seasonal and peak during winter months. Many of these workers applied during the winter so that if they did later have a COVID-19-related reason to initiate a claim, it would not be counted as an application but instead as a continuing claim. These industries may also be experiencing secondary impacts of the pandemic recession due to decreased demand.
| - | Cumulative | Share of Total Labor Force | Share of Total Applicants |
|---|---|---|---|
| Demographic Group | |||
| Under 25 | 116,701 | 26.3% | 16.0% |
| 25-54 | 453,080 | 23.7% | 62.2% |
| 55+ | 158,588 | 21.8% | 21.8% |
| High school Diploma or less | 267,118 | 42.6% | 36.7% |
| Some college or 2-year degree | 286,736 | 33.4% | 39.4% |
| 4-year degree or more | 174,515 | 16.1% | 24.0% |
| Female | 389,813 | 27.9% | 53.5% |
| Male | 338,522 | 22.1% | 46.5% |
| People of Color or Indigenous | 174,599 | 36.8% | 24.0% |
| White | 524,351 | 19.8% | 72.0% |
| All | 728,369 | 23.7% | 100.0% |
| Detailed Race/Ethnicity | |||
| American Indian Alaska Native | 8,534 | 34.7% | 1.2% |
| Asian | 39,908 | 28.4% | 5.5% |
| Black | 70,634 | 40.3% | 9.7% |
| Hispanic | 36,746 | 23.3% | 5.0% |
| Native Hawaiian or Pacific Islander | 1,246 | NA | 0.2% |
| Non-Hispanic White | 524,351 | 19.8% | 72.0% |
| More than one race | 17,531 | 29.5% | 2.4% |
| Choose not to answer | 29,419 | NA | 4.0% |
| All | 728,369 | 23.7% | 100.0% |
| Note: NA = population not available | |||
Cumulatively, 72% of UI applicants from mid-March through the end of May in Minnesota were non-Hispanic white. However, 36.8% of people of color in Minnesota's labor force have applied for UI compared to 19.8% of non-Hispanic whites. UI applicants as a share of the labor force varies widely by race and ethnicity: Blacks (40.3%) and American Indians (34.7%) have the highest UI applications as a share of total labor force size. (An article on employment disparities will be posted soon.)
Again, this is because of the industries in which different demographic groups have concentrations of employment. A high share of American Indian workers are employed in casinos, which remain closed through this period, and hard hit accommodation and food service. In Minnesota, Black workers are concentrated in the hard hit health care sector as well as accommodations and food service, and administrative support. Temp help falls within the administrative support industry grouping and temp help has seen dramatic declines in employment over the period.
The above applicant numbers include both applicants for the regular UI program and applicants for the PUA program. Of the 728,369 UI applicants over this period, 99,585 were self-employed PUA applicants.
| Week Ending | Applications | Continued Claims |
|---|---|---|
| 21-Mar | 115,773 | 76,527 |
| 28-Mar | 110,171 | 186,127 |
| 4-Apr | 112,436 | 288,054 |
| 11-Apr | 92,860 | 397,062 |
| 18-Apr | 77,020 | 479,481 |
| 25-Apr | 51,867 | 502,984 |
| 2-May | 46,761 | 524,389 |
| 9-May | 39,465 | 548,846 |
| 16-May | 32,298 | 559,023 |
| 23-May | 26,106 | 541,593 |
| 30-May | 20,462 | 519,844 |
Continued claims, the second UI data measure noted above, provides a gauge of how many people continue to be affected by the pandemic induced recession. Each week UI applicants who remain unemployed must request their UI benefit payment. In Minnesota, continued claims reached a high of 559,023 during the week ending May 16 and dropped moderately during each of the next two weeks. This is a sign that people are beginning to return to work as restrictions were lifted and companies began to use Paycheck Protection Program (PPP) funds to hire back employees. (An article on the employment impacts of the Paycheck Protection Program will be posted soon).
As more information comes in on the U.S. economy, evidence builds that the labor market will bounce back more quickly than was forecast earlier in the COVID-19 period. The U.S. unemployment rate, much to the astonishment of many economists and labor market analysts, dropped in May and the number of payroll jobs grew. This may mean that the worst is behind us and that as states begin to open back up and as the PPP begins to draw employees back onto payroll, the economy will bounce back quickly. But at this point in time, that bright outlook is still far from certain, given unknown factors such as a resurgence of COVID-19 infections and hospitalizations.
Because the COVID-19 containment measures had a non-uniform impact on different industries and occupations, they also impacted different groups of workers more severely than others. Therefore, it is likely that there will be a long period of time when many Minnesotans, in particular young Minnesotans and Minnesotans of color who make up a large share of the most deeply impacted industries and occupations, will remain unemployed. At this point in time, there is no comprehensive data available on the number of businesses that have shut down permanently, and no one knows with certainly when sectors like restaurants or arts and entertainment will be operating at full capacity again. It will be important, for both equity and sustained economic growth reasons, that Minnesota policymakers do not lose sight of those workers who are unable to get back to work quickly through no fault of their own.
1Note that assigning applications to industry is inexact due to multiple job holding and high turnover.
