Tuesday, December 17, 2013

The Employment Survey Offers a Clearer View

Mark Knold, Supervising Economist

I recently posted an article profiling the unemployment rate and some of the myths surrounding its perception. The unemployment rate has its shortcomings, and my conclusion is the job growth survey should be looked toward as the primary economic barometer. It offers a more direct view into current economic activity. Let’s look at why.

The unemployment rate comes through a monthly survey of households. The report on Utah’s job growth comes from a monthly survey of businesses. Economic commerce and responses to it are more sensitive in the business community than in households. For example, Google can make a product, thrive with it, hire workers, and advance the economy. It cannot do this while sitting idle. If Google went idle it would soon cease to exist. However, while Google is stimulating the economy, numerous households can sit idly by and do little within the economy, oblivious to the actions and impacts of Google. A household can be economically idle and still easily exist. By way of illustration, a retired household has income, yet may not be doing much to impact the economy. Therefore, I value the business survey as a more dynamic gauge of economic health than the household survey.


Surveys are merely a sampling of a greater population. If the entire population were sampled, then you have a census. But a census is costly and regularly unfeasible, so a survey stands in as a substitute. A strong sample is one that has its foundation upon a census. No census is ever performed at any time to strengthen the unemployment survey. Its results are a survey and always a survey. On the other hand, the employer survey is regularly checked against an employer census.

All states have an unemployment insurance system. Through it, laid-off workers may possibly receive unemployment benefits. Past earnings are often the criteria determining eligibility and weekly benefit amount. But how does each state know about any workers’ past earnings?

Four times a year, most businesses that hire workers in each state must report their payroll information into that state’s respective unemployment insurance program (there are some exclusions, like non-profits). This suddenly makes each state’s unemployment insurance system the closest thing this country sees to an employment census—and it is accumulated four times a year.

One drawback is the time it takes to collect and aggregate all this census employment. Employment counts for January, February, and March typically will not be summarized until at least July—four months after the fact. Economy watchers, however, want to know in January, February, and March what is happening in those months. To meet this need, something quicker and more nimble is required—like a survey.

That is the employer survey. To strengthen its accuracy, once a year the employer survey undergoes maintenance (called benchmarking) and is repopulated with the actual census employment counts. Then the survey is re-commenced upon this new foundation. In this manner the employment survey is regularly calibrated to match actual employment. No such procedure exists within the unemployment survey.

By way of further illustration, a worker cannot “hide” from the census employment count. If a job is cut, it will show up in the census employment count as one fewer position. But if the person who lost that job does not go looking for another job, that job loss will not show up in the unemployment rate. The lost job is counted, but the “unemployment” associated with it is not.

It is for this and myriad other reasons that the monthly employment estimate of Utah’s job performance is a stronger barometer of Utah’s current economic performance than is the seemingly more cited unemployment rate.

No comments:

Post a Comment

We welcome your comments! However, we do not post comments with "links" or those comments directing to your website. These will not be published. Thank you