On Friday, we got the December jobs number: +200,000. That’s good, but not good enough. I posted a graph from the Hamilton Project showing that, at that rate, the labor market wouldn’t recover till 2024.

But perhaps that’s too pessimistic. The Economic Policy Institute took a look at the same numbers and concluded that a growth rate of 200,000 jobs per month would lead to a full recovery in seven years or so. That’s nothing to celebrate, but it’s better than the Hamilton Project’s estimate of 12 years. It’s also a bit odd: Isn’t this a simple matter of taking job losses and dividing by monthly job gains? Well, no. The date of our eventual recovery depends on some crucial unknowables about the future of the American labor force.

I asked Heidi Shierholz, a labor economist at EPI, to explain how she got to seven years. “My method is really straightforward,” she e-mailed. “Assume an annual job growth rate of 1.8% (which is the annualized growth rate of the 200,000 we got last month) and given that growth rate, see how long it takes to get back to the December 2007 ratio of payroll to working-age population, (assuming the working-age population continues to grow as it has been growing recently).”

Note the “assuming the working-age population continues to grow as it has been growing recently” in Shierholz’s explanation. That’s about to become important.

Next, I asked Michael Greenstone, director of the Hamilton Project, how that equation could shave five years off the expected time it will take the labor market to recover. “Their calculation assumes a lower growth rate in the labor force and hence fewer jobs needed to get back to normal.”

Here’s the issue: Before the recession, the economy needed to produce 120,000 jobs a month just to keep up with new entrants into the labor market. Lately, that number has been closer to 90,000. Part of this is that immigration has fallen and many immigrants are leaving. Part of it is that some workers are leaving the labor force — either they can’t find a job and have given up, or they have decided to stay home with the kids or focus on other pursuits rather than take the sort of jobs they can get right now.

The question is whether the growth of the labor force bounces back or holds steady. And it turns out, this question matters a lot. This table that Greenstone sent along shows how long it will take to reverse the jobs losses at different rates of payroll and employment growth. As you can see, if we’re adding 200,000 jobs a month, it will take almost twice as long to recover if we’re adding 125,000 workers a month rather than 90,000:

(Michael Greenstone, Hamilton Project)

That said, this is a case where a longer recovery could mean a better recovery. If we continue growing at 90,000 jobs a month, it likely means that many of the long-term unemployed never made it back into the labor force, and that the economy is producing less than it otherwise could. It means, in other words, that the recovery is proving unable to reverse some of the deepest wounds inflicted by the recession. If we get catch-up growth in the labor market, that may push back the return of full employment, and it may even temporarily increase the unemployment rate, but it will mean we’re seeing a fuller recovery.