Opinion: Carolina Beat

No. 739: Will the Real Job Count Please Stand Up?

More than two million jobs have been lost nationwide in the last three years. No, almost 800,000 net new jobs have been created. Huh? Are both statements correct? Or, is someone cooking the books or manipulating statistics?

Actually, both numbers are right. They come from different government surveys of the job market. They differ because they measure the job market in different ways. However, knowing what each survey does and doesn’t measure is vitally important in knowing what’s happened in the labor market.

The survey showing a job loss of more than two million positions in the previous three years is based on going to existing businesses and counting jobs. This survey is called the establishment, or payroll, job survey.

However, the government does another count of jobs by going to people’s residences and asking whether they’re working. People who say they are working are counted as having a job. This measure of jobs is called the household survey.

So why can the two surveys give different employment totals? Easy. The establishment survey counts jobs at existing businesses, but the list of current businesses is always somewhat out of date. This means the establishment survey misses jobs at newly created companies. It also misses workers who are self-employed.

The establishment survey of jobs will eventually catch up with the new companies, but since new businesses are always being added, the survey will always under-count total jobs.

But the under-counting is especially a problem in the years following a recession. It’s then that some people who lost jobs during the recession will start their own businesses. And, the increase in consumer and business confidence following a recession will motivate new business starts.

The household survey is not without its critics. The number of people contacted at home is far less than the number of companies visited for the establishment survey. Also, some economists have questions about whether self-employment is overstated in the household survey.

The same two job surveys are also done in North Carolina, but interestingly, they are giving much closer results than at the national level. Over the past three years, both surveys show job losses in our state. The establishment survey shows net job cuts of 135,000, whereas the household survey shows jobs down by 94,000. But notice, the job losses are lower in the household survey than in the establishment survey.

What conclusions should we draw about the debate over counting jobs? I think there are three bottom lines. First, in recent years, the number of jobs in existing companies has dropped. Second, during the same time period, jobs have been added at newly started firms and through self-employment. Third, whether these new jobs have replaced all the jobs cut at longstanding companies is a matter of contention between economists and statisticians. My own sense is the answer is “yes” in the nation, but “no” in North Carolina.

Most people judge the economy by the job market. The government’s two job surveys have been giving conflicting signals about the nationwide job market, although the signals have been consistent for North Carolina’s employment situation. When you hear politicians quote job statistics this year, make sure you know what count they’re using.