In a mysterious labor market development, the December expiration of extended unemployment benefits has yet to show up in the data for early 2014.

February's jobs report from the Bureau of Labor Statistics showed labor market resilience, as businesses added 175,000 jobs and the labor force participation rate rose despite snowstorms across the United States during the survey week.

On the other hand, the unemployment rate rose slightly, from 6.6 percent to 6.7 percent, as did the number of long-term unemployed workers, by 203,000 to 3.8 million.

Those results are the opposite of what would be expected in the months after two million Americans — as estimated by House Ways and Means Committee Democrats — lost extended unemployment benefits.

The extended unemployment insurance program, which was first authorized in 2008 and then renewed several times before expiring in 2013, provided benefits for workers who had been unemployed for longer than 26 weeks.

It’s often thought that by in effect paying workers to not have jobs, unemployment benefits incentivize workers to remain unemployed.

And there is empirical evidence to that effect. One recent examination of the data by Princeton's Henry Farber and the Federal Reserve Bank of San Francisco's Robert Valetta found reductions in workers exiting unemployment and increases in the average duration of unemployment in states that extended unemployment insurance. They concluded that extra benefits raised the overall unemployment rate by about 0.4 percentage points over the course of the recession.

Other analyses have suggested that eliminating unemployment insurance leads workers to quit the labor force entirely. And some studies, like one published by the Congressional Budget Office in late 2012, suggest that extended unemployment benefits boost economic growth and employment when unemployment is high, by allowing those out of work amid scarce openings to spend more than they would otherwise.

But the logic is clear that ending unemployment benefits should lower the unemployment rate, either by forcing the unemployed to take whatever jobs are available, or by the unemployed giving up the job search entirely and falling out of the official labor force.

North Carolina served as a test case for the impact of unemployment insurance by cutting off extended benefits at the state level in 2013. Over the rest of the year, unemployment fell sharply, from 8.8 to 6.9 percent.

It’s not yet clear how much of that decline represented unemployed workers finding and taking jobs versus ones quitting the job hunt. But either way, the effect on the unemployment rate seemed clear.

Yet, two months after the North Carolina experiment was replicated at the national level, the labor markets have run in the opposite direction.

Headline unemployment, at 6.7 percent, is where it was in December. The number of long-term unemployed increased, as did the average duration of unemployment. When extended unemployment benefits had previously been scaled back over the second half of 2012, reducing the maximum number of weeks of unemployment benefits from 99 to 73, the average unemployment spell also fell:

Clearly, there has been no rush of workers from the ranks of the long-term unemployed to take jobs.

But neither have those workers quit looking for work as a group. In fact, February was the first time in nearly four years that more unemployed workers found jobs (2.2 million) than left the labor force (2.1 million).

Those facts create a bit of a puzzle regarding what has happened in labor markets since December. One possible explanation is that the labor market is stronger than it appears from the headline numbers, and that as more data comes in and the winter weather subsides it will become clear that the jobs recovery has been strong enough to keep the long-term unemployed looking for work despite their loss of benefits.

Federal Reserve economists have previously suggested that a similar dynamic may have inflated the number of long-term unemployed. The Census Bureau counts as unemployed anyone who says that he is out of work and has looked for work in the past month. Normally, the vast majority of new people in that category report having been unemployed for fewer than five weeks, meaning that they have been laid off or otherwise lost work and began looking for new work.

But during the current recovery, a disproportionate share of those newly counted as unemployed say that they’ve been unemployed for 27 weeks or longer — meaning they’re immediately counted among the long-term unemployed.

In a late 2011 paper, economists in the Federal Reserve System offered an explanation: With the labor market slowly improving, workers who had used up their unemployment benefits during the depths of the recession and quit the job search, thereby falling out of the Census' “unemployed” category, were regaining hope about finding jobs and starting to look for openings again. The result was an odd situation in which they were technically newly among the ranks of the unemployed, but telling pollsters that they had been out of work for years.

But previously discouraged workers rejoining the official labor force as long-term unemployed were likely not part of the reason for February’s surge in the number of long-term unemployed. The number of workers going from “out of the labor force” to “unemployed” in February fell from 2.6 million to 2.5 million.

In other words, February’s jobs report, which otherwise was relatively strong, only raises new questions about the fate of the long-term jobless in the wake of the recession.