What Is the Unemployment Rate Really?

Kent Sellers

August: 4.6. May: 5.1. January: 6.0.

What do all of these unemployment rate data points mean?

Labor market information (LMI) can help workforce professionals, educators and job seekers distinguish where there are strengths and/or weaknesses in their economies. All of the attention to economic data may allow for more informed decision-making at all levels: informed job seekers, business owners, educators and policymakers.

The unemployment numbers (and rates) are some of the most monitored economic data across the United States. The monthly unemployment rate attracts lots of media coverage. Consequently, it is important to have a good understanding of the unemployment and labor force estimates, how they are calculated, and what these numbers can (and don’t) mean. This article looks briefly at the sources and covers the annual and monthly revision process. To conclude, this article offers some “best practices” so that the reader will be more fully prepared to understand and interpret the monthly unemployment rate and labor force estimates.

Where Does the Unemployment Rate Come From?

Workforce agencies, such as the Indiana Department of Workforce Development (DWD), exist as part of state government in all 50 states and have cooperative data agreements with the Bureau of Labor Statistics (BLS) of the U.S. Department of Labor. Each state has analysts who work cooperatively and uniformly in each of the employment and wage statistical program areas. The Local Area Unemployment Statistics (LAUS) is just one of several programs, and this is the program from which the monthly unemployment and labor force estimates are produced.graphic

Each month, BLS announces the total number of employed and unemployed persons in the United States for the most recently completed month, along with many characteristics of such individuals. This LAUS program is a federal-state cooperative program and estimates several components of the labor force each month: civilian labor force, employment, unemployment and unemployment rate. The data can be drilled down to the county level, as well as cities with populations exceeding 25,000. Some states also publish data for smaller cities and towns, but this data can be highly volatile due to the small survey sample size. Indiana publishes data for some of these small areas throughout the state.

One of the biggest misconceptions is the notion that the unemployment numbers are derived solely from those who file claims for unemployment insurance (UI). However, there are people without jobs who delay or do not file for UI benefits; there may also be some that have exhausted their benefits and still do not have a job, and individuals that just aren’t eligible to receive UI benefits. As you can imagine, this would not be an accurate depiction of the total count of the unemployed population.

Instead, LAUS estimates are based on a model using three primary variables in addition to historical trends and current surveyed estimates:

  1. The Current Population Survey (CPS), a phone survey conducted by the U.S. Census Bureau. There are about 60,000 households in the U.S. and about 1,000 households in Indiana that make up the sample in this survey each month. The survey participants are asked detailed questions about their current work or work search activities.graphic
  2. The statewide and metropolitan statistical area (MSA) estimates of Current Employment Statistics (CES) for each month, plus CES-type estimates for non-MSA counties. CES is another federal-state survey-based cooperative program that estimates monthly employment by industry and area (statewide and MSAs) based on reports for around 5,000 establishments each month in Indiana. The CES produces the monthly “jobs” numbers we learn about each month and are typically released at the same time as the LAUS estimates by BLS and the states.
  3. The UI continued claims counts by area (these are used in part to allocate unemployment estimates to MSAs, counties and smaller cities).

The LAUS model is heavily driven by the CPS sample of about 1,000 households in Indiana, but CES estimates, claims counts and historical trends serve to mediate the volatility of the sample and provide more real-time local data.

As more data becomes available, we revise.

How Is the Unemployment Rate Revised?

The LAUS preliminary estimates are revised each month to take advantage of the revised CES estimates, which are used as one major variable for the LAUS model, and of the additional unemployment claims data—another major variable. Sometimes firms surveyed report their data late, and sometimes there are other delays in modeling or processing that contribute to these minor monthly revisions.

In addition, at the beginning of each year, the previous year’s estimates are thoroughly revised again through a process known as benchmarking. The Quarterly Census of Employment & Wages (QCEW) is known as the “universe” (counts) of employment and wage data. These data come from Unemployment Insurance (UI) covered employers, sometimes referred to as payroll employers, who provide quarterly employment levels and payroll to the states. The QCEW is a much more complete set of data that provides employment levels of all payroll establishments. The CES (monthly jobs numbers) draws its sample of establishments from QCEW. The QCEW program publishes a quarterly count of employment and wages reported by employers, covering approximately 95 percent of wage and salary workers (not including self-employment). Unfortunately, the QCEW is lagged by about six months (e.g., January–March data isn’t available until early September). Once this complete data set becomes available, it is used to benchmark the employment levels used for the monthly estimates prior to beginning another year of estimating.

How Much Can the Preliminary Unemployment Rate Differ from the Benchmarked Unemployment Rate?

Let’s illustrate the extent of the differences between the monthly estimates and the benchmarked (final revision) estimates. Table 1 shows those counties with some of the largest revisions during the most recent benchmark process, along with Indiana as a comparison point. Whereas Indiana had a relatively minor final revision of half a percentage point, the counties highlighted saw revisions of more than 1 percentage point, either plus or minus. That upward revision for Orange County means that the rate for March 2014 was actually significantly higher than originally estimated—8.7 percent instead of 6.9 percent. Other counties had their unemployment rates revised downward; for example, Tipton’s rate was revised to 5.4 percent from the original 6.6 percent.

Therefore, it would be wise to keep in mind that unemployment rates are best used when analyzing trends over a period of time rather than analyzing the most recent unemployment rate only.

Table 1: Monthly Unemployment Rates and Revisions in 2014

Geography Month Preliminary Unemployment Rate Benchmarked Unemployment Rate Percentage Point Difference
Indiana August 5.5% 6.0% +0.5%
Orange County March 6.9% 8.7% +1.8%
Greene County March 7.1% 8.7% +1.6%
Blackford County March 6.8% 8.3% +1.5%
Floyd County April 6.0% 4.8% -1.2%
Tipton County June 6.6% 5.4% -1.2%
Harrison County April 6.5% 5.4% -1.1%

Source: Local Area Unemployment Statistics, Indiana Department of Workforce Development

What Are Some Best Practices When Looking at the Unemployment Rate?

Now that a better understanding of the unemployment data and caveats have been established, here are a few tips to keep in mind to best utilize unemployment data:

  1. Understand that there is a statistically acceptable margin of error for these estimates and all economic models. For example, we hear this all the time when political polls are released with a “plus or minus” 3 percent or 5 percent.
  2. Look for the trends. Instead of looking at each month compared to the last, identify trends over the course of many months. Year-to-year trends are best when looking at non-seasonally adjusted rates, as it is less subjective to the volatility of monthly preliminary estimates.
  3. Recognize that most of the monthly economic indicators from the U.S. Department of Labor published in the media are the preliminary estimates. Further exploration will have to be done in order to learn about the revised or benchmarked unemployment rates.

What Are Some Alternative Measures to the Unemployment Rate?

During the current economic recovery, increased attention has been given to the BLS’ Alternative Measures of Labor Underutilization.1 These measures offer an alternative to the more commonly reviewed total unemployment rate as an estimate of all labor underutilization, including people that are considered underemployed. BLS has published these for the U.S. on a monthly basis for many years. They can be found at the end of the monthly employment situation press release in Table A-12.2 Whereas the total unemployment rate only considers people unemployed if they are actively seeking work, these measures offer broader definitions of “unemployed,” as well as some narrowly defined targeted measures of the long-term unemployed. The U-1 measure is the narrowest definition and estimates the percentage of the labor force unemployed for 15 weeks or longer.

One of the broadest alternative measures of labor underutilization, the U-6, is expanded to include unemployed or underemployed people that may be working part-time for economic reasons, people that are considered discouraged and are no longer looking for work, and all others that are only “marginally” attached to the labor force. The U-3 is the commonly known unemployment rate. Indiana's value has now fallen below the national average. The U-6 measure has followed the same trend, indicating a lower percent of total unemployed, and a lower percent of discouraged and underemployed workers than the national average. The most recent release for Indiana and all states is a four-quarter rolling average ending in the first quarter of 2015. For this time frame, Indiana’s “regular” unemployment rate was 5.5 percent compared to 5.7 percent nationally. The U-6 measure was still nearly twice as high, at 10.3 percent for Indiana and 11.3 percent nationally.

Economic developers can hope that this indicates there is still room for discouraged and part-time workers to re-enter the labor force as the recovery continues.

Learn More

  • To keep up-to-date with unemployment numbers and other labor market information, check out the Hoosiers by the Numbers website: www.hoosierdata.in.gov.

  • For more information and FAQs about the BLS Local Area Unemployment Statistics program, visit www.bls.gov/lau/.

  • To learn more about Alternative Measures of Labor Underutilization/Unemployment (also known as underemployment), go to www.bls.gov/lau/stalt.htm.

Notes

  1. Alternative Measures of Labor Underutilization for States: www.bls.gov/lau/stalt.htm.
  2. October Employment Situation, Table A-12: www.bls.gov/news.release/empsit.t12.htm.