The unemployment rate is one of the most closely watched, yet least understood pieces of economic data. On the surface it appears simple: in December the Colorado Springs unemployment rate was 9.4 percent, while Pueblo’s was an even higher 10.3 percent. Doesn’t that mean that around 10 percent of adults in southern Colorado who wanted jobs were unable to find work?
Well, no it doesn’t. It simply tells us that 10 percent of southern Coloradans who didn’t have jobs in December, but had looked for one in the past four weeks, were unemployed. If you were out of work but didn’t actively look for a job, you may think you are unemployed, your spouse certainly thinks you are unemployed, your neighbors think you are unemployed, but the government says you aren’t. You are no longer in the labor force and thus don’t count in the unemployment rate.
You also aren’t considered unemployed if you went from a $150,000 a year job as a systems engineer to a 10-hour-a-week, minimum-wage job at a local fast-food restaurant. In both cases you count as one employed individual. In fact, if you now hold two 10-hour, minimum-wage jobs, you’ve gone from counting as one employed person in El Paso County to two. Wonderful, employment is up!
Employment data come from two separate surveys conducted by the Bureau of Labor Statistics. The unemployment rate comes from a monthly survey of about 60,000 households conducted by 2,200 highly trained labor department personnel. A household remains in the survey for four months. Sample data aren’t perfect, but for the nation, Colorado and metro Denver, there is a 90 percent chance that the error is extremely small and the data provide a useful picture of the job situation – if they are read with care.
In Colorado, there are 1,362 households in the survey sample, of which about 1,000 respond to the interviews each month. Since El Paso County contains about 12 percent of the state’s population and Pueblo County only 3.1 percent, it seems reasonable to assume that there are only about 163 and 42 households respectively in the local sample, and even fewer who respond and provide the unemployment data each month. That isn’t a large enough sample to be statistically significant, to use a technical term. In other words, the unemployment rate is about as likely to be wrong as it is to be right.
There are alternative measures of unemployment for the nation and the state (unfortunately not available at the county level) that tell us much more about the labor market situation than the simple unemployment rate statistic that garners the headlines. The broadest measure has the uninspired name of the U6 rate and can be found on the BLS web site, http://www.bls.gov/lau/stalt10q4.htm. It includes not only those unemployed who have looked for jobs in the past four weeks, but also those who want work and have looked in the past year and those who wanted a full-time job but could only find part-time work. Still excluded are those individuals who need a job but are so discouraged that they haven’t looked for work in the past 12 months. In 2010, the U6 rate for the U.S. was 16.7 percent rather than the reported 9.6 percent and the Colorado rate was 15.4 percent rather than 8.7 percent. In California, Michigan and Nevada almost a quarter of the workforce was unemployed by the broadest measure.
The fact that the unemployment rate must be taken with a large grain of salt doesn’t mean it should be ignored. Pay more attention to the national and state rates than you do to the local rate. Watch how the local rate moves over time, but ignore small changes. Someone once said that the reason economists use decimal points is to show that we have a sense of humor. A far more reliable and important figure is the number of people with jobs, although even that doesn’t distinguish between part-time and full-time work.
Adams, a Colorado Springs resident and longtime Colorado economist, is a senior partner at Summit Economics. She can be reached at firstname.lastname@example.org.