Department of Labor Logo United States Department of Labor
Dot gov

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Topic Archives: Survey Respondents

How Much Does a Cup of Coffee Cost? It’s Complicated

We have a guest blogger for this edition of Commissioner’s Corner. Rob Cage is the Assistant Commissioner for Consumer Prices and Price Indexes at the U.S. Bureau of Labor Statistics.

The pandemic has changed my morning routine. Before the outbreak of the COVID-19 pandemic and full-time telework at BLS, two things motivated me each morning.

Person holding mobile phone and ordering coffee on an app.

First, I was always on a mission to minimize my commute to work. I would do things each night so I wouldn’t waste time in the morning. Things like shaving, setting out clothes, and preparing the next day’s lunch. I timed my alarm to go off to allow just enough time to shower, suit up, grab that sandwich, and catch my commuter train as it rolled into the station.

The second thing I needed to start each work day was a strong, fresh, hot cup of joe—actually more like two or three cups. Not one of those fancy drinks with mocha, caramel, steamed milk, or anything like that. Ordinary drip-brewed, filtered coffee. Medium to dark roast and like Betty MacDonald, my coffee had to be “…so strong it snarled as it lurched out of the pot.” Then I add some cream (and by cream, I mean half-and-half), but no sugar. But by far the most important element of the drink: temperature. I like coffee precisely at a certain temperature. If it’s too hot, you taste nothing but a scalded tongue. If it’s too cold, you’re met with an overwhelming sense of disappointment. In that ideal temperate zone, you are jolted alive with a satisfying sip of silky cocoa and nutty fragranced bliss.

Through trial and error, I eventually unearthed a way to satisfy both of these morning habits efficiently: getting my coffee along the commute. Brewing the coffee at home took too much time, and I’d drink most of it on the train, arriving at my desk empty handed. Getting my first cup after I arrived was also uneconomical since I’d have to backtrack to get it. The simple solution? Find a place to get the coffee along the way, and preferably as close to my office as possible. This way, the temperature of the drink was in that sweet spot as I turned on my computer.

With four different coffee shops located along my route in Washington’s Union Station, one would think I could easily achieve this. But no, I’m foiled by impatience. According to a 2014 Journal of Consumer Behavior study, the time before ordering has the greatest influence on how customers perceive waiting times and service quality. A customer who has to wait 10 minutes in line before ordering will feel more dissatisfied than a customer waiting 10 minutes after ordering, even if the total wait time is the same. I couldn’t agree more. Queues at Union Station during the morning rush were just too long and unpredictable to meet my needs. I didn’t have the patience to wait behind customers pondering through a long order recital: Quad Grande nonfat extra hot caramel macchiato upside down, please. I needed my expeditiously stated, two-word order quickly. Luckily, the employee cafeteria in my building—conveniently located just off the lobby—had self-serve coffee. No competing commuters. No preorder queue. No postorder queue. Only a payment queue. I had found my routine: a 55-minute total commute, landing at my desk with strong, hot coffee in hand.

Then one day, I bumped into a coworker on the train. As we walked through Union Station and approached the maze of coffee shops with the insufferable queues, she stopped in front of one; took two steps to the left, scanned the drinks on top of a cart, found one with her name on it, picked it up, and met me back in stride. Amazed, I asked her how she pulled off this sensational stunt. She had placed her order on the coffee shop’s mobile app, of course. That was her routine. Curious but unconvinced, I asked her if she was concerned the coffee would be too cold by the time she picked it up. Through trial and error, she had figured out that if she placed her order on the app as the train rolled out of the L’Enfant Plaza stop, her drink would typically be hot and ready as she passed the cart. Could this be coffee-ordering nirvana? Guaranteed no-wait service, with guaranteed handoff at perfect temperature? Surely this improvement in the quality of the purchasing experience would cost more, which was my next question. And the astonishing answer: the coffee was the same price! I immediately downloaded the app, copied her process, and shaved three minutes off my morning routine. An equilibrium commute down to 52 minutes, about a 5-percent improvement!

Which brings me to how this tortured story relates to the business of BLS and specifically the measurement of the cost of living and consumer inflation. If the cost of my preferred cup of coffee was identical ($2.45 before sales tax) whether I stood in line to get it or not, then surely I would be better off by ordering on the app. Doing that resulted in a 5-percent time savings on my commute—an attribute of purchasing coffee that was critically important to me. In other words, the app-ordered coffee represented a higher-quality product, even though the price was the same. Using the federal minimum wage rate of $7.25\hour (or 12 cents a minute), an estimate of the time savings is 3 minutes x $0.12 = $0.36. One could say $0.36 is a reasonable estimate of the difference in quality. So what is the correct measure of price change between these two choices?

ApproachWalk-up purchaseApp purchasePrice changeNote

Ignore purchase time

$2.45 $2.45 0%No change in price

Add purchase time

$2.81 $2.45 -13%Deflation

Assume purchase time is built into market price, and adjust prices to reflect zero purchase time

$2.09 $2.45 17%Inflation

This is the million dollar question in consumer price index measurement, and the answer depends on how a unique consumer good—in this case a prepared cup of coffee—is defined. In the price index literature, the buzzword is homogeneity. To measure inflation accurately, goods that are homogenous must be identified and grouped together for proper treatment. This is at the core of getting the CPI right. Homogenous is defined as “of the same kind, alike; consisting of parts all of the same kind.” In CPI jargon, the component “parts” of a unique item in the sample are called “attributes.” So what are the attributes that define a cup of coffee? We could consider a list of attributes that most baristas might say are important, like size, bean variety, country of origin, blend, roast, freshness, or caffeine content; and a couple you already know that are important to me: temperature and queue time.

How many of these attributes do we explicitly control for in the CPI as obvious, overt, and separate variables used in scientifically selecting a sample of coffee drinks from quick service establishments, for use in calculating the index each month? You might be surprised by the answer: none! How, then, do we capture constant-quality price change for prepared coffee drinks accurately in the CPI?

We implicitly account for all of these characteristics one way or another. The CPI uses the matched-model approach to index measurement. We select a sample of 100,000+ unique, well-specified, strictly homogenous goods and services for the sample. Then we compare the price of each unique sampled item to the price of the exact same item in subsequent months. The key, of course, is defining and selecting the unique items. Generally speaking, sample selection has two major components: selection of the establishments (for example, a coffee shop) and then selection of a unique item (for example, 16-ounce dark roast drip coffee) at the selected establishments. Limited budget requires BLS to take a sample rather than a census of all goods and services consumers purchase. Thus, we group unique products into broadly homogeneous categories so the selected products can accurately reflect price change for unsampled items in those categories. We bundle prepared coffee from quick service establishments into the elementary category “limited service meals and snacks.” Comparatively, this is one of the more broadly defined components in the CPI basket. With a variety of different food and beverage items eligible for the sample, there are simply too many attributes to consider as separate selection steps to create the sample of unique items. Instead, we base the selection largely on the descriptions of different items listed on the menu. This is how we would distinguish an ordinary brewed coffee drink from other coffee drinks, such as a latte and cappuccino.

Any attribute expressly identified in the description of the menu item becomes a characteristic defining the unique item. For example, “12-ounce Cup of Organic Single Origin Light Roast Coffee” and “12-ounce Cup of Organic Classic Blend Medium Roast Coffee” may be two different menu items at a coffee shop. By rule, they are treated as distinct, unique, separate products for CPI sample selection. Then each month, CPI data collectors meticulously capture the price of the exact same product. If any of the characteristics change, that would trigger a quality review. Suppose medium roast was no longer available. A decision would have to be made to substitute the most comparable item to the originally selected item. Then a commodity analyst in the national office would have to decide if the new item was comparable to the old item. For example, is there a difference in quality between the light roast and the medium roast? Obviously, consumer taste and preferences are idiosyncratic, and the difference in quality of light roast and medium roast is a function of individual preference. But to the average consumer, perhaps not. In fact, prices tend not to vary by roast type. So in this situation, the analyst might judge medium and light to be comparable, and the price of the light will be matched to the previous price of the medium and used in the index. However, if a single-origin coffee was selected, a different outcome might result, especially if the price of the single-origin coffee was considerably different from a previously selected blend coffee, with all other characteristics being the same. Then a decision would need to be made as to how much of the difference was a quality difference (single origin versus blend), and how much was pure price change.

But what about the other factors that are not expressly identified in the description of the menu item, like temperature, freshness, and queue time? These are ostensibly identified, and held constant month after month, by the selection of the establishment. The outlet itself is associated with many attributes of product quality which are not observed. Over time, customers come to expect a certain level of service or product quality within each specific store, or at specific locations of chain stores. So, by controlling for the outlet, we are effectively able to hold constant these unobservable attributes.

Now that I am teleworking, my morning habits are out of equilibrium. My commute time is drastically shorter, reduced to the time it takes me to walk from my bedroom to the guest room, which has been hastily converted into a home office. My problem is the coffee. I haven’t figured out the roast, or the precise coffee-to-water ratio for the perfect strength; I don’t like spending time grinding whole bean, so I substitute ground coffee instead. My barista tells me that’s a quality decrease.

I’d say I am better off commute-timing wise but worse off coffee wise. A push. All in all, I can’t wait to return to on-premises work, mostly for that reliable cup of java.

A Closer Look at Recent Employment Trends

BLS has closely tracked the upheaval in the U.S. job market in recent months, most notably through the monthly “payroll jobs” data. These data, from the Current Employment Statistics survey, provide detail on the change in employment in each industry. We count jobs by asking thousands of employers every month the number of employees on their payroll for the pay period that includes the 12th of the month. For August, we reported that employers added 1.4 million jobs. Today I want to scratch beneath that surface and examine recent employment trends in several industries.

But before I go on, let me take a moment to thank all those businesses that respond voluntarily to our request for information every month. With so much going on, responding to a BLS survey may not be your highest priority. Yet, you continue to come through every month, and for that we extend our sincere thanks.

Using February 2020 as our starting point, let’s look at the job losses that occurred through April. From the nearly 152 million jobs recorded in February, we lost just over 22 million by the end of April. That’s a drop of 14.5 percent in total nonfarm employment. But that decline varied across industries. The leisure and hospitality industry, including restaurants, hotels, and amusements, saw the largest percentage decline, down 49.3 percent from February. Other industries saw percentage declines similar to the overall total, such as retail trade (decline of 15.2 percent) and construction (decline of 14.2 percent). And some industries experienced small declines, such as financial activities (decline of 3.2 percent). These differences stem from many factors, including stay-at-home orders, the need for workers in essential industries, the ability for some work to be done remotely, and on and on.

Editor’s note: Data for this chart are available in the table below.

Following large losses through April, many industries gained jobs over the next four months. By August, about 10.6 million jobs were added to employer payrolls. One way to look at these figures is to consider what share of the March/April job loss was “recovered” by the May/June/July/August job gain. Overall, 47.9 percent of the decline was recovered. The retail trade industry restored the greatest percentage of job losses, 72.5 percent, followed by other services (including barbers and salons, 61.2 percent) and construction (60.8 percent). Education and health services recovered 47.6 percent of lost jobs, nearly equal to the overall percentage of jobs recovered, as did manufacturing (47.2 percent). Utilities, mining and logging, and the information industry had fewer jobs in August than in April.

Editor’s note: Data for this chart are available in the table below.

While the percentages let you compare industries, digging a little deeper uncovers other interesting stories. For example, three sectors, professional and business services; manufacturing; and transportation and warehousing, each lost between 10 and 11 percent of jobs from February to April 2020. But those losses amounted to vastly different numbers of jobs: 2.3 million in professional and business services; 1.4 million in manufacturing; and 570,000 in transportation and warehousing.

Some detailed industries provide interesting contrasts. Within health care from February to April, hospital employment showed a slight decline while offices of physicians lost about 11 percent of jobs. In contrast, offices of dentists declined by 56 percent, losing more than half a million jobs. As of August, employment had rebounded in most health care industries, with the notable exception of nursing and residential care facilities, which has declined each month since February.

Editor’s note: Data for this chart are available in the table below.

Americans were encouraged to stay at home and only venture out for essential items, which is reflected in employment in various retail industries. For example, food and beverage stores showed little employment change from February to August. In contrast, clothing store employment declined by 62 percent through April, and only half of that loss had been recovered by August. Jobs in electronics and appliance stores declined through May and in August stood at about 90 percent of their February total.

Editor’s note: Data for this chart are available in the table below.

A reminder that Current Employment Statistics data are updated as new information becomes available. Thus, the July and August data shown here are preliminary and will be revised. Employment data by industry are also available for states and localities.

When looking for trends or comparing industries of different sizes, the comparisons shown here can be helpful. The detailed data are available for you to compare other industries, too. Get the data through the BLS data query system.

Percent decline in payroll employment from February through April 2020, by major industry
IndustryPercent decline

Leisure and hospitality

-49.3

Other services

-23.1

Retail trade

-15.2

Total nonfarm

-14.5

Construction

-14.2

Education and health services

-11.3

Professional and business services

-10.7

Manufacturing

-10.6

Transportation and warehousing

-10.0

Information

-9.8

Mining and logging

-8.5

Wholesale trade

-6.7

Government

-4.3

Financial activities

-3.2

Utilities

-0.7
Percent of payroll employment decline from February to April 2020 that was recovered by August 2020, by major industry
IndustryPercent recovered

Retail trade

72.5

Other services

61.2

Construction

60.8

Leisure and hospitality

50.2

Total nonfarm

47.9

Education and health services

47.6

Manufacturing

47.2

Professional and business services

35.8

Transportation and warehousing

33.2

Financial activities

31.5

Wholesale trade

17.4

Government

14.2

Information

-9.5

Mining and logging

-59.0

Utilities

-86.8
Percent of February 2020 employment level in months after February, selected health care industries
IndustryAprilMayJuneJulyAugust

Offices of physicians

89.291.594.195.296.2

Offices of dentists

43.869.289.093.996.1

Hospitals

97.797.097.197.697.8

Nursing and residential care facilities

96.494.994.393.793.2
Percent of February 2020 employment level in months after February, selected retail industries
IndustryAprilMayJuneJulyAugust

Electronics and appliance stores

89.874.780.286.290.5

Building material and garden supply stores

97.3101.8104.3105.1106.1

Food and beverage stores

98.6100.4101.7101.0101.2

Clothing and clothing accessories stores

38.244.562.470.371.1

Department stores

75.279.490.094.597.5

General merchandise stores, including warehouse stores

104.6106.2109.0105.8110.1

Labor Day 2020 Fast Facts

I have been Commissioner of Labor Statistics for about a year and a half now, and what a time it has been! BLS has faced many challenges throughout its history, but none quite like those from the COVID-19 pandemic. All of our staff moved to full-time telework March 16, and I am so proud of how well they have worked under trying circumstances. In a very short time—days, not weeks—we had to change our data collection processes to eliminate in-person collection and move to a combination of telephone, internet, and video. We recognize how challenging it is for our survey respondents to provide data during the pandemic, and I am very grateful for their cooperation. Response rates have dipped a bit in some programs, but the quality of our samples remains strong across the board. Despite all of the challenges, BLS has been able to produce all of our economic reports without interruption.

The pandemic has taught us there’s an unlimited appetite for data. The U.S. statistical system is working to satisfy that appetite. At BLS, we strive for more and better data to understand the hardships caused by the pandemic. Starting in May we added new questions to our monthly survey of households. The questions ask whether people teleworked or worked from home because of the pandemic; whether people were unable to work because their employers closed or lost business; whether they were paid for that missed work; and whether the pandemic prevented job-seeking activities. We continue to gather new data from those questions.

We collaborated with our partners at other U.S. statistical agencies to find out how many people received payments from the Coronavirus Aid, Relief, and Economic Security (CARES) Act, signed into law on March 27, 2020. For those who received payments, we asked how they used them.

Soon we will have new data about how businesses have responded to the pandemic. These data are from a brand new survey that seeks to identify changes to business operations, employment, workforce flexibilities, and benefits as a result of the pandemic.

These are just a few examples of how our data collection has responded to the pandemic. Good data are essential for identifying problems, guiding policymakers, and gauging whether and how fast conditions improve for workers, jobseekers, families, and businesses.

Labor Day is a good time to reflect on where we are. Despite these difficult times, I hope you are able to enjoy the long holiday weekend. Take a moment to look at some fast facts we’ve compiled on the current picture of our labor market.

Working

Our monthly payroll survey shows that employment had been increasing through February 2020. With March came the pandemic and the job losses related to it. We lost more than 22 million jobs in March and April and then regained about 48 percent of them in May, June, July, and August.

The employment–population ratio was 56.5 percent in August. This ratio is the number of people employed as a percent of the population age 16 and older. The ratio was 61.1 percent in February.

There were 7.6 million people working part time for economic reasons in August 2020. These are people who would have preferred full-time employment but were working part time because their hours had been reduced or they were unable to find full-time jobs. This number was down from 10.9 million in April. The number was 4.3 million in February.

Not Working

The unemployment rate reached 14.7 percent in April 2020. That was the highest rate, and the largest over-the-month increase, in the history of the data back to January 1948. The rate has fallen since then, reaching 8.4 percent in August. The rate was 3.5 percent back in February, the lowest since 1969.

We have noted the challenges of measuring unemployment during this pandemic. The rates we have seen since March likely understate unemployment, but the trend is clear. The rate rose sharply in March and even more sharply in April and has trended down since April.

Among the major worker groups in August 2020, the unemployment rate was 8.4 percent for adult women and 8.0 percent for adult men. The rate for teenagers was 16.1 percent. The unemployment rate was 13.0 percent for Blacks or African Americans, 10.7 percent for Asians, 10.5 percent for Hispanics or Latinos, and 7.3 percent for Whites.

Job Openings

On the last business day of June 2020, the number of nonfarm job openings was 5.9 million. That was a decline of 18 percent from June 2019.

The ratio of unemployed people per job opening was 3.0 in June 2020. Since the most recent peak of 4.6 in April 2020, the ratio of unemployed people per job opening declined in May and June. In February 2020, there was 0.8 unemployment person per job opening.

Pay and Benefits

Civilian compensation (wage and benefit) costs increased 2.7 percent in June 2020 from a year earlier. After adjusting for inflation, real compensation costs rose 2.1 percent over the year.

Paid leave benefits are available to most private industry workers. The access rates in March 2019 were 73 percent for sick leave, 79 percent for vacation, and 79 percent for holidays.

In March 2019, civilian workers with employer-provided medical plans paid 20 percent of the cost of medical care premiums for single coverage and 33 percent for family coverage.

Productivity

Labor productivity—output per hour worked—in the U.S. nonfarm business sector grew 2.8 percent from the second quarter of 2019 to the second quarter of 2020. That increase reflects large pandemic-related declines in output (−11.2 percent) and hours worked (−13.6 percent).

Safety and Health

In 2018, there were 5,250 fatal workplace injuries. That was a 2-percent increase from 2017 and was the highest number of fatal work injuries in a decade. It was, however, below the numbers of workplace deaths in the 1990s, when over 6,000 fatalities occurred per year.

There were about 2.8 million nonfatal workplace injuries and illnesses reported in 2018 by private industry employers. This resulted in an incidence rate of 2.8 cases per 100 full-time workers in 2018. The rate is down from 9.2 cases per 100 full-time workers in 1976.

Unionization

The union membership rate—the percent of wage and salary workers who were members of unions—was 10.3 percent in 2019, down by 0.2 percentage point from 2018. In 1983, the first year for which comparable union data are available, the union membership rate was 20.1 percent.

Total employer compensation costs for private-industry union workers were $48.57 and for nonunion workers $34.16 per employee hour worked in March 2020. The cost of benefits accounted for 40.5 percent of total compensation (or $19.65) for union workers and 28.4 percent (or $9.71) for nonunion workers.

Looking to the Future

We released our latest set of long-term employment projections September 1. We project employment to grow by 6.0 million jobs from 2019 to 2029. That is an annual growth rate of 0.4 percent, slower than the 2009–19 annual growth rate of 1.3 percent. The healthcare and social assistance sector is projected to add the most new jobs, and 6 of the 10 fastest growing occupations are related to healthcare. These projections do not include impacts of the COVID-19 pandemic and response efforts. We develop the projections using models based on historical data. The historical data for this set of projections cover the period through 2019, so all input data precede the pandemic. We will continue to examine the effects of the pandemic as we update our projections next year and the years that follow.

From an American worker’s first job to retirement and everything in between, BLS has a stat for that! Want to learn more? Follow us on Twitter @BLS_gov.

New Recommendations on Improving Data on Contingent and Alternative Work Arrangements

The workplace is changing. We have seen more evidence of that in recent months as workplaces have adapted to the COVID-19 pandemic. Even before the pandemic, many of us wanted to learn more about telework, flexible work hours, and independent contracting. We also wanted to know more about intermittent or short-term work found through mobile devices, unpredictable work schedules, and other employment relationships we might not think of as traditional. It’s our job at BLS to keep up with these new work relationships and figure out how to measure them.

In 2018, we released data collected in 2017 about people in contingent and alternative work arrangements. Contingent workers are people who do not expect their jobs to last or who report their jobs are temporary. Alternative work arrangements include independent contractors, on-call workers, temporary help agency workers, and workers provided by contract firms. We also published data in 2018 about electronically mediated work. All of these data reflect the rapidly changing workplace.

Those reports received a lot of attention, but policymakers, employers, researchers, and others told us they want to know more about these nontraditional workers. We need to understand people in jobs that often involve doing short-term tasks, such as ridesharing or data-entry services. Our 2017 survey included a few questions about these arrangements, but this work can be complex and varied. That makes it hard to measure nontraditional work arrangements with just a few questions.

To effectively analyze these hard-to-measure work arrangement, BLS sought out experts on nontraditional work. In 2019, we contracted with the Committee on National Statistics to explore what we should measure if we had the funding to collect and publish more data about these workers. We asked the committee not to recommend changes to the main Current Population Survey, the large monthly survey of U.S. households from which we measure the unemployment rate and other important labor market measures. The committee had free rein, however, to recommend topics we should examine in any future edition of the Contingent Worker Supplement to the Current Population Survey. We also asked the committee to recommend changes to the survey design and methods of data collection if we were to conduct the supplement again.

The Committee on National Statistics is a federally supported independent organization whose mission is to improve the statistical methods and information that guide public policies. The committee moved quickly to form a group of experts on the relevant topics. I asked these experts to review the Contingent Worker Supplement and consider other sources of information on nontraditional work arrangements. The group was impressive and included a former BLS Commissioner, a former Administrator of the U.S. Department of Labor Wage and Hour Division, and several experts in economics and survey methods. They all volunteered their time to help us improve the Contingent Worker Supplement.

The group held public meetings and a workshop, hearing from experts, data users, and policymakers to understand what data would be the most valuable. At the end of their year-long review, they produced a report with specific recommendations in July of 2020 about measurement objectives and data collection.

BLS thanks the Committee on National Statistics and the expert panel for the time and effort they put into the report. Their recommendations thoughtfully balanced the desire to measure everything about this important topic with the limited time and information survey respondents can give us. In the coming months, we will study the report. It will guide us as we consider how to update the Contingent Worker Supplement to reflect the variety of work arrangements in the U.S. labor market.

Update on the Misclassification that Affected the Unemployment Rate

How hard can it be to figure out whether a person is employed or unemployed? Turns out, it can be hard. When BLS put out the employment and unemployment numbers for March, April, and May 2020, we also provided information about misclassification of some people. I want to spend some time to explain this issue, how it affected the data, and how we are addressing it.

In the monthly Current Population Survey of U.S. households, people age 16 and older are placed into one of three categories:

  • Employed — they worked at least one hour “for pay or profit” during the past week.
  • Unemployed — they did not work but actively looked for work during the past 4 weeks OR they were on temporary layoff and expect to return to work.
  • Not in the labor force — everyone else (including students, retirees, those who have given up their job search, and others).

Again, how hard can this be? It starts to get tricky when we talk to people who say they have a steady job but did not work any hours during the past week. In normal times, this might include people on vacation, home sick, or on jury duty. And we would continue to count them as employed. But during the COVID-19 pandemic, the collapse of labor markets created challenges the likes of which BLS has never encountered. People who reported zero hours of work offered such explanations as “I work at a sports arena and everything is postponed” or “the restaurant I work at is closed.” These people should be counted as unemployed on temporary layoff. As it turns out, a large number of people—we estimate about 4.9 million in May—were misclassified.

With the onset of the COVID-19 pandemic, the unemployment rate—at a 50-year low of 3.5 percent in February—rose sharply to 4.4 percent in March and to 14.7 percent in April, before easing to 13.3 percent in May. Despite the stark difference from February, we believe the unemployment rate likely was higher than reported in March, April, and May. As stated in our Employment Situation news releases for each of those months, some people in the Current Population Survey (also known as the CPS or household survey) were classified as employed but probably should have been classified as unemployed.

How did the misclassification happen?

We uncovered the misclassification because we saw a sharp rise in the number of people who were employed but were absent from their jobs for the entire reference week for “other reasons.” The misclassification hinges on how survey interviewers record answers to a question on why people who had a job were absent from work the previous week.

According to special pandemic-related interviewer instructions for this question, answers from people who said they were absent because of pandemic-related business closures should have been recorded as “on layoff (temporary or indefinite).” Instead, many of these answers were recorded as “other reasons.” Recording these answers as “on layoff (temporary or indefinite)” ensures that people are asked the follow-up questions needed to classify them as unemployed. It does not necessarily mean they would be classified as unemployed on temporary layoff, but I’ll get into that in a moment.

When interviewers record a response of “other reasons” to this question, they also add a few words describing that other reason. BLS reviewed these descriptions to better understand the large increase in the number of people absent from work for “other reasons.” Our analysis suggests this group of people included many who were on layoff because of the pandemic. They would have been classified as unemployed on temporary layoff had their answers been recorded correctly.

What are BLS and the Census Bureau doing to address the misclassification?

BLS and our partners at the U.S. Census Bureau take misclassification very seriously. We’re taking more steps to fix this problem. (The Census Bureau is responsible for collecting the household survey data, and BLS is responsible for analyzing and publishing the labor market data from the survey.) Both agencies are continuing to investigate why the misclassification occurred.

Before the March data collection, we anticipated some issues with certain questions in the survey because of the unprecedented nature of this national crisis. As a result, interviewers received special instructions on how to answer the temporarily absent question if a person said they had a job but did not work because of the pandemic. Nevertheless, we determined that not all of the responses to this question in March were coded according to the special instructions. Therefore, before the April data collection, all interviewers received an email that included instructions with more detailed examples, along with a reference table to help them code responses to this question. However, the misclassification was still evident in the April data. Before the May data collection, every field supervisor had a conference call with the interviewers they manage. In these conference calls, the supervisors reviewed the detailed instructions, provided examples to clarify the instructions, and answered interviewers’ questions.

Although we noticed some improvement for May, the misclassification persisted. Therefore, we have taken more steps to correct the problem. Before the June collection, the Census Bureau provided more training to review the guidance to the interviewers. The interviewers also received extra training aids. The electronic survey questionnaire also now has new special instructions that will be more accessible during survey interviews.

Why doesn’t BLS adjust the unemployment rate to account for the misclassification?

As I explained above, we know some workers classified as absent from work for “other reasons” are misclassified. People have asked why we just don’t reclassify these people from employed to unemployed. The answer is there is no easy correction we could have made. Changing a person’s labor force classification would involve more than changing the response to the question about why people were absent from their jobs.

Although we believe many responses to the question on why people were absent from their jobs appear to have been incorrectly recorded, we do not have enough information to reclassify each person’s labor force status. To begin with, we don’t know the exact information provided by the person responding to the survey. We know the brief descriptions included in the “other reasons” category often appear to go against the guidance provided to the survey interviewers. But we don’t have all of the information the respondent might have provided during the interview.

Also, we don’t know the answers to the questions respondents would have been asked if their answers to the question on the reason not at work had been coded differently. This is because people whose answers were recorded as absent from work for “other reasons” were not asked the follow-up questions needed to determine whether they should be classified as unemployed. Specifically, we don’t know whether they expected to be recalled to work and whether they could return to work if recalled. Therefore, shifting people’s answers from “other reasons” to “on layoff (temporary or indefinite)” would not have been enough to change their classification from employed to unemployed. We would have had to assume how they would have responded to the follow-up questions. Had we changed answers based on wrong assumptions, we would have introduced more error.

In addition, our usual practice is to accept data from the household survey as recorded. In the 80-year history of the household survey, we do not know of any actions taken on an ad hoc basis to change respondents’ answers to the labor force questions. Any ad hoc adjustment we could have made would have relied on assumptions instead of data. If BLS were to make ad hoc changes, it could also appear we were manipulating the data. That’s something we’ll never do.

How much did the misclassification affect the unemployment rate?

We don’t know the exact extent of this misclassification. To figure out what the unemployment rate might have been if there were no misclassification, we have to make some assumptions. These assumptions involve deciding (1) how many people in the “other reasons” category actually were misclassified, (2) how many people who were misclassified expected to be recalled, and (3) how many people who were misclassified were available to return to work.

In the material that accompanied our Employment Situation news releases for March, April, and May, we provided an estimate of the potential size of the misclassification and its impact on the unemployment rate. Here we assumed all of the increase in the number of employed people who were not at work for “other reasons,” when compared with the average for recent years, was due solely to misclassification. We also assumed all of these people expected to be recalled and were available to return to work.

For example, there were 5.4 million workers with a job but not at work who were included in the “other reasons” category in May 2020. That was about 4.9 million higher than the average for May 2016–19. If we assume this 4.9 million increase was entirely due to misclassification and all of these misclassified workers expected to be recalled and were available for work, the unemployment rate for May would have been 16.4 percent. (For more information about this, see items 12 and 13 in our note for May. We made similar calculations for March and April.)

These broad assumptions represent the upper bound of our estimate of misclassification. These assumptions result in the largest number of people being classified as unemployed and the largest increase in the unemployment rate. However, these assumptions probably overstate the size of the misclassification. It is unlikely that everyone who was misclassified expected to be recalled and was available to return to work. It is also unlikely that all of the increase in the number of employed people not at work for “other reasons” was due to misclassification. People may be correctly classified in the “other reasons” category. For example, someone who owns a business (and does not have another job) is classified as employed in the household survey. Business owners who are absent from work due to labor market downturns (or in this case, pandemic-related business closures) should be classified as employed but absent from work for “other reasons.”

Regardless of the assumptions we might make about misclassification, the trend in the unemployment rate over the period in question is the same; the rate increased in March and April and eased in May. BLS will continue to investigate the issue, attempting both to ensure that data are correctly recorded in future months and to provide more information about the effect of misclassification on the unemployment rate.