Mike Licht/Flickr

Mike Licht/Flickr

Long-term unemployment is a terrifying trap that, even in the best of times, is difficult to escape. And it's a trap that you can get stuck in for no reason other than bad luck.

Today, there are still almost 3.5 million people who have been out of work for six months or longer and are looking for work. There isn't a more urgent crisis, and there are three things you should keep in mind about it.

1. As former CEA Chair Alan Krueger found, the long-term unemployed aren't much different from the short-term unemployed. They're a little older and more of them are African-Americans, but they're just about as educated and work in the same industries as everyone else who's trying to find a job.

2. The long-term unemployed have a hard time getting companies to even look at their job applications, let alone hire them. Rand Ghayad, a labor economist at Northeastern University, has tested this: he sent out thousands of fictitious resumés that were basically identical except for how long they said they'd been unemployed and what field they'd been in before. The results? Employers preferred people without any relevant experience but who'd been unemployed less than six months to people with experience who'd been unemployed longer than that. In other words, how long you'd been out of work trumped all else.

3.There's never been this much long-term unemployment before, at least not since they started keeping records in 1948. Right now, 35 percent of all unemployed people have been out of work for at least six months. That's actually down from the all-time high of 48 percent in 2010, but it's still well above the pre-Great Recession one of 28 percent in 1983.

Long-term unemployment is about macroeconomic bad luck

Long-term unemployment isn't a story about lazy people choosing to live on the dole instead of getting a job. It's a story about people who want a job not being able to find one, because there aren't enough of them—and then falling to the back of the jobs line. That is, it's a story about macroeconomic bad luck. Think about it this way. We know that companies discriminate against the long-term unemployed. That's why their ranks have been so slow to come down. But we also know that the long-term unemployed are like the short-term unemployed in every way except for how long they've been out of work. So we don't know why so many more short-term unemployed people are becoming long-term unemployed today than in the past—unless it's the economy, stupid.

And it is the economy. Ben Casselman of 538 has already shown that the best predictor for how long you were unemployed was the unemployment rate when you lost your job. So if you got laid off in 2009, when the economy was bad, there was a much, much higher chance that you'd end up long-term unemployed than if you'd gotten laid off in 2007, when the economy was still booming.

The chances you'll end up in the ranks of the long-term unemployed

To find out what our long-term unemployment problem is all about, I've extended Casselman's analysis using monthly rather than yearly microdata from the Current Population Survey to let us zoom in a little more. Specifically, we can see, going back to 1998, how many people became unemployed every month, and how many of them were still unemployed six months later.

Now, there are two quick notes. First, the monthly data is very volatile, in part because it's not seasonally adjusted, so I averaged it over three months to smooth things out. And second, these are survey results, which means the responses tend to cluster around six months or a year for people who've been out of work a long time. In other words, people can't always remember exactly how many weeks they've been unemployed, so they round to the nearest big number. This should upwardly bias our findings a little, but only a little.

Now let's get to those results. As you can see below, the odds of becoming long-term unemployed have closely tracked the business cycle. In the late 1990s, when, believe it or not, unemployment dipped below 4 percent, there was a less than a 5 percent chance that an unemployed person would still be six months later. Those odds shot up to 15 percent after the tech bubble burst, and more or less stayed there during the jobless recovery into 2003. The housing bubble finally pushed unemployment, and with it the probability of becoming long-term unemployed, down—but neither got as low as they had been when we were partying like it was 1999.

Then the Great Recession happened. Unemployment went up slowly, then all at once after Lehmangeddon. But it wasn't just the depth of the crisis that hurt so much. It was the length. And the lackluster recovery, too. Think about someone who lost their job in September 2008. The next six months were, to say the least, a tough time to find work—and if they hadn't by then, their chances would have fallen even further as companies started to discriminate against them. That's why people who lost their jobs in 2009, when unemployment peaked at 10 percent, had a 30 percent chance of ending up long-term unemployed. And those bleak odds have only slowly receded as unemployment has. Indeed, if you become unemployed today, you still have a higher chance of becoming long-term unemployed than you would have at the worst point of the tech bust.

LTE Odds.jpg

The last question is just how closely long-term unemployment follows the business cycle. It sure looks like it does, but what do the numbers say? Well, I ran regressions comparing the odds of becoming long-term unemployed with the unemployment rate and the broader unemployment rate, which includes discouraged people and part-time workers who can't find full-time jobs. Both were highly statistically significant, but the broad unemployment rate explained slightly more of the data—an R-squared of 88 percent. In other words, long-term joblessness is highly correlated with the business cycle, and there's no reason to think that the long-term unemployed are necessarily lazy or lacking skills. They're mostly caught in the tide of (bad) economic history. You can see this simple relationship below: more unemployment means more long-term unemployment.

U6 vs LTE Odds3.jpg

The lesson here is that we need stimulus. Sure, the economy will eventually recover on its own in the long run. But in the long run, our careers, if not ourselves, will be dead. High unemployment begets high long-term unemployment, which forces people into premature retirement. Spending more and printing more money can at least mitigate these vicissitudes of the business cycle—if, as the IMF points out, we're willing to be radical enough.

Nobody should lose their career because they lost their job at a bad time.