Lies, damned lies, and unemployment statistics

Front and center in the upcoming campaign will undoubtedly be discussions of unemployment. Prominent links on Obama’s and Romney’s campaign pages are their respective plans to “create” jobs.

Talking about jobs can be somewhat irksome to economists partly because employment and unemployment statistics can be easily misinterpreted. The most recent (July) Bureau of Labor Statistics (BLS) report says

Total nonfarm payroll employment rose by 163,000 in July, and the unemployment rate was essentially unchanged at 8.3 percent.

The press release mentions the change in the level of employment but in the percent of unemployment. The flavor of the story might change if it was reversed.

163 thousand sounds like a big number (Tempe, Arizona’s population in July 2011 was 164 thousand), and even if employment increased by just one job, if you were the person who got that job you’d be happy. But what if the report stated “Total nonfarm payroll employment rose from 133.1 million to 133.2 million from June to July” (Table B-1)? Suddenly the picture seems more bland when you consider that it rose only 0.1%.

The bland story is supposed to be the “essentially unchanged” unemployment rate, but the rate actually increased from 8.2% to 8.3% because there were 45 thousand more people unemployed (Table A-1). If you were one of these 45 thousand people who newly lost your job you wouldn’t consider your own financial situation to be “essentially unchanged.” Add to this the fact that there were already 12.75 million people unemployed, more than the total populations of New York City and Los Angeles.

Even though it is cited so often, I doubt whether most people have a clear understanding of how the unemployment rate is actually calculated. The BLS surveys about 60,000 households every month on their employment situation (so there is already the possibility of user response error). To be considered employed doesn’t require very much: you have to work for pay at least one hour a week or can even be unpaid if you work for a family business. Thus, a manager who lost a good full-time job and now works ten hours a week part-time at a restaurant will not affect the employment number. To be considered unemployed means you don’t have a job but have looked for one in the past month. Thus, a spouse who loses a full-time job in April who decides to stay at home for the summer months is not unemployed.

The unemployment rate itself is calculated as the percentage of people in the labor force (employed + unemployed) who are unemployed. We usually presume that drops in the unemployment rate mean the economy is better and rises in the rate mean the economy is worse; but the opposite can happen. If some unemployed people stop looking for work and fall out of the labor force because their job prospects are dismally poor, then the drop in the numerator will be proportionally larger than the drop in the denominator and the unemployment rate will fall, usually signaling good times. Likewise, if things pick up and more people start to look for work again, the rise in the numerator will be proportionally larger than the rise in the denominator so the unemployment rate will rise, giving the impression that things are worse.

So, using the unemployment rate as your go-to single statistic on how the economy is doing is a bad idea. If you just want to focus on jobs, at least look also at the employment rate (= employment/labor force); the employment-population ratio, the number of discouraged workers, the number of long-term unemployed (27 weeks or more), and involuntary part-time workers. Better yet, look at productivity, changes in the inventory-to-sales ratio, or real GDP. Of course, all of these statistics have their problems too but tend to be less subject to misinterpretation than the unemployment rate.

Above all, remember that jobs are not the cause of economic growth, they are the effect. Jobs cannot be created without businesses being successful, and businesses in general are rarely successful in the long term because of things government can do for them (besides protect private property and enforce contracts). The assessment of political candidates should be based not on promises to create so many millions of jobs, but on a plan to let businesses get back to business without government meddling, cronyism, and bailouts.



  • Poor Rich Folks?

    Lack of evidence? You gotta be kidding. I am weeping right now for the money sorrows you think rich people have to endure. Where’s the evidence that rich folks would reinvest more of their already more than substantial, privileged and tax sheltered gains if additionally capital gains taxes were abolished? And luxury items are based more on high profit items for which skilled workers don’t get proportionately or fairly paid.

  • Common Cents

    Maybe unemployment rates are contingent upon the extent to which the super rich horde their profits in tax sheltered offshore bank accounts, heh? And how much they lavish themselves spending on luxuries instead of reinvesting in their businesses and in new business endeavors.

    • Tim Shaughnessy

      Despite your lack of evidence and emotional appeal to envy, I’ll play along with your game of if…
      IF the “super rich” (like Oprah? “horde their profits in tax-sheltered offshore accounts,” then doesn’t that speak to the onshore tax system that penalizes entrepreneurship? If we eliminated the double taxation of capital gains taxes, there would be a flood of money back home that would be reinvested.
      IF the “super rich” lavish themselves with luxuries, who do you think produces those luxuries? Do you think the “super rich” produce and sell yachts? More than likely, skilled tradesmen do, and more than likely, the makers of luxuries are probably not all rich.

  • Christopher

    Tim, you are spot on some of the analysis. However you must add what it truly means to be unemployed & looking for a job. If a teacher has a signed contract or an agreement to return in September, we counted them as employed. If you were unemployed & all you did was look at ads in the newspaper, job boards or online, but didn’t contact anyone, send a resume, fill out am application, talk to a employer or network with friends to find an opening, your not counted as unemployed even if you are receiving unemployment. Your counted as lazy, discouraged or taking care of home & family. Out of the workforce. That number/definition is the only reason, the UI index isn’t 1930’s highs. I should know, I’ll be doing the survey later next month.

    • Tim Shaughnessy

      I ask all CV readers to keep people in those situations in their prayers. My goal was to point out some of the ways that unemployment statistics are misleading, and you helped to add an additional layer to the discussion.



Receive our updates via email.