How Apps, Entrepreneurship and a Steady Boom Have Brought Unemployment Claims to an All-Time Low

The Facts

The 4-week moving average of initial unemployment claims is at an all-time low.

In raw numbers, it was actually lower in 1969 but at that point the labor force was exactly half (well, okay, 50.7%) what it is now. So as a percentage, it has never been lower in recorded history.

This statistic is a measure of how many people walk into an unemployment office to say, "I've lost my job and don't have another one to go to."

In April of 2009, 658,000 people filed for unemployment each week. That was the worst of the Great Recession. October of 1982 was even worse, with 671,750 walking into unemployment offices around the country during a single week.

But the American economy is always shedding and creating jobs, at a rate of about 2 million per month. It's remarkable that such a small number of folks laid off or quitting one job don't end up in an unemployment office before they get their next job.

The week of 24 February, only 220,500 people filed initial unemployment claims. The last time it was lower was 27 December 1969 - nearly 50 years ago - when it hit 219,750. As a percentage of the labor force, though, that 1969 number equated to about 1.1% of the labor force showing up in an unemployment office during the month whereas this latest number suggests only 0.5% of the labor force filed for unemployment in the month. It's a stunning number.

The Theories

The most obvious explanation is an uninterrupted rise in jobs created that has now gone on for 88 months (and counting). Every month that results in more jobs created than destroyed means that many fewer people unemployed or unable to find a job. I think there is more to even than that, though.

Simply put, the economy has never been more efficient at creating new jobs and then matching unemployed people to those new jobs. I don't know why but I have a theory that it is because of apps and entrepreneurship.

Uber, Lyft, Mechanical Turk, PostMates, and other apps quickly match people to tasks they can do. Even the websites like Monster, and Indeed have accelerated the time it takes for employers to find qualified employees to hire. Some apps quickly find someone to perform a task; other sites accelerate the time it takes to find a new employee.

Once upon a time, in small communities, you knew that Todd could help carry heavy loads and Melissa could repair fragile things. You could easily find help and they could find work. As the world got bigger and more advanced, it became harder to know who could program in Java vs. C, or who could design period-furniture and who could repair modern furniture. It took a long time for the unemployed to find jobs and for employers to find help. It would take months to find the right person for a job and for tasks that might take only minutes or days, you might never find a match. There was a lot of friction in job markets even a decade ago.

Now software lets a person who wants to drive you to the airport find the person who wants a ride to the airport within about 15 minutes. It's easier than ever to find a match and this makes for nearly friction-less labor markets. This means that more people are downloading an app to make money (I know, to qualify as an Uber driver is not as easy as downloading an app) rather than waiting for a person to hire them. One result is a lower number of folks who file for unemployment.

Another element is increased levels of entrepreneurship. It is easier than ever to start or expand a business. Once upon a time you had to get loans to buy a store or build a factory to start a business; now you can fund a software startup with six laptops. More than half of American employees now work at least part of the week at home. This suggests that the overhead for office space per person is dropping, one less barrier to starting or expanding a business. (I know. It's not THAT cheap or simple. Still, even renting a cubicle is cheaper than setting up a factory.) A great number of employees are hired as contractors; some because that is now how corporations are engaging employees and some because that is an increasingly common way to put someone through the equivalent of a probationary period. There is less commitment and expense in "hiring" an employee and thus less hesitancy to do so. Employees still face a great deal of uncertainty about particular contracts or income levels but less likely to go long stretches without some kind of income.

In this way it's a bit like the move from a bank account that offers 3% annual rate vs. a stock that could rise or fall 30% in a year. Income will fluctuate more but employment will not. People are less likely to turn to unemployment insurance than to another job or task that could mean a temporary rise or fall in pay. If not already, I suspect that fewer people will look back at the last ten years of employment as a steady rise of 4% in annual pay and will instead see rise and falls more akin to the performance of a 401(k).

What It Means

One thing that no one would suggest, though, is that the Uber drivers are fine with just their cars and Uber app. Among other things, they need roads to drive on. Why mention this obvious thing? As the economy and information systems become more adept at matching supply and demand, it's important to support the infrastructure that makes it work. In this case, it's not just roads. People who are more likely to be getting their income from contract jobs and apps need things like universal healthcare to replace the standard benefits once provided by corporate employers and job training programs to make them steadily more productive. The good news is that these more efficient markets will mean less reliance on government unemployment insurance; the bad news is that more tenuous income streams suggests a greater need for things like government health insurance and education.

Another implication of more efficient labor markets is the very real possibility that the natural rate of unemployment has dropped. The ideal rate of unemployment would not be zero for the simple fact that finding a great fit between employee and job is not an instantaneous process. Given it takes a little while to find a great match, it makes sense that somewhere between 2 to 5% of the labor force would be unemployed at any given time. If it is true that it's easier for people to find work, it makes sense that this rate has gone down. What that means is that if string of uninterrupted job creation continues another 6 to 24 months, the unemployment rate could approach 3%. [I've already forecast about a 33% chance of a recession but if that doesn't hit this year, unemployment could steadily trend downwards.]

In all, more efficient labor markets is yet another great sign of progress. It doesn't mean that business cycles are over but it does mean that in any given month fewer people face the prospect of unemployment. That's pretty cool.