The U.S. Labor Department continues to release wonderful news for U.S. workers.

Just this week, on Sept. 20, the agency said that the number of Americans filing for unemployment benefits reached the lowest level in almost 49 years. Unemployment benefits track changes in the number of workers who are laid off by companies. When few companies lay off workers, this signals a tight labor market and often means unemployment rates are likely to drop further.

As for the official U.S. unemployment rate, it continues to trickle downward and was last at just 3.9 percent during August, which was the lowest level since the late 1960s.

With the economy still strong, just how low could it go? And could the unemployment rate ever get to zero?

Economics often gets a bad rap as the “dismal science,” I and many of my colleagues are actually quite optimistic people who often search for how good things can get – not how bad.

The definition of unemployed

Many people may think that they know what the U.S. unemployment rate means and what it measures. After all, how complicated could it be, people either have a job or they don’t?

The official definition, however, is far from simple. To be considered unemployed, a person must pass three tests.

First, the person has to be immediately available to work. If a company asks you to start today, you must be able to say “yes” to be considered unemployed. Searching for a job that starts weeks or months from now does not count.

Second, to be unemployed someone cannot be working, even a tiny bit. Driving for a car company like Uber or Lyft while looking for a job means a person is not classified as unemployed but instead is counted as working.

Lastly, a person has to be actively searching for a job. Active searching means doing something that could result in an employer contacting a job seeker. Spending hours cruising internet job boards doesn’t count, unless a person sends in at least one resume or contacts a company directly.

Types of unemployment

Economists divide the reasons people are unemployed into five reasons: cyclical, structural, seasonal, frictional and institutional. For the unemployment rate to become zero, all five would have to disappear.

Cyclical unemployment happens because the economy goes through periodic cycles of booms and busts. During the Great Recession from 2007 to 2009, businesses needed far fewer workers because they were selling less. With the economy expanding today, businesses now need more workers, so currently almost no one is unemployed for this reason.

Structural unemployment occurs when a worker’s skills no longer match any business need. The economy is constantly evolving and new types of jobs are being created, while old jobs are destroyed.

Decades ago when I was in high school, some of my friends took stenography classes to get jobs as secretaries. Stenography was a type of shorthand used so bosses could dictate letters, which were typed up later.

Today voice recognition and word processing software has eliminated the need for stenographers. If any of my friends are still looking for a stenography position, they would be considered structurally unemployed.

The only way to eliminate structural unemployment is to prevent new ideas and inventions. This is not something most people want so the economy always has some structural unemployment.

Seasonal unemployment is when there is no work because of weather or time conditions. For example, lifeguards who protect swimmers are usually hired in summertime and then laid off in the fall. Little can be done about seasonal unemployment since control of the weather and seasons are beyond human control. This means we always will have some seasonal unemployment.

Frictional unemployment arises because searching for a job does not always provide instantaneous results. It takes time for businesses looking for workers and people wanting jobs to find each other.

The U.S. government tracks the number of open jobs each month via the JOLTS survey. During this past summer, there were almost 7 million open jobs at the same time that 6.3 million people were considered unemployed.

Frictional unemployment explains why millions could be without work even if there are more vacancies. Faster communication helps reduce frictional unemployment, but as long as people and businesses take time to interview and make up their minds, this type of unemployment too will exist.

Finally, institutional unemployment arises when wages are too high and cannot fall. This is one reason why some people argue against raising the minimum wage rate to US$15 per hour.

Critics claim that if businesses are forced to pay higher wages, they will cut back their hiring of the low-skilled and boost unemployment. Others argue changing the minimum wage has little impact on employment.

In my mind, institutional unemployment could theoretically be zero. The only question is at what minimum wage rate this happens.

More room to fall

Even though some types of unemployment could zero out, others will always remain – meaning the overall rate will never reach zero percent. But then what’s the minimum rate of unemployment even the healthiest economy should expect?

The Congressional Budget Office takes a rather pessimistic view of the matter and concludes that the bare minimum of unemployment is over 4 percent – perhaps viewing the recent figures as anomalies.

Still, past experience suggests the jobless rate could continue to fall, despite the dour Congressional Budget Office perspective.

Using data that stretch back to 1948 shows that the unemployment rate has, in fact, been quite a bit lower than the current level. For a period of 13 months in 1952 and 1953, the rate was consistently below 3 percent and fell to just 2.5 percent in May of 1953, the lowest recorded value as the economy expanded to support the Korean War.

And in the late ‘60s, the last time the rate was so low, it was under 3.5 percent from September 1968 to May 1969. In total, the unemployment rate has been below the current level for 88 months since 1948.

Just how low the unemployment rate will go today is still an open question. But, if the economy keeps growing at the current pace, I believe there is a small chance of reaching the old record of 2.5 percent.

Jay L. Zagorsky does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.