For the past six months, the unemployment rate has toggled between 4.9 and 5 percent, about the rate many economists consider the lowest jobless rate consistent with stable inflation. Does that mean we’ve arrived at the critical destination for an economic expansion: full employment, meaning a very tight matchup between labor supply and labor demand?

No, because at 9.8 percent, underemployment is still too high. In earlier work, I showed that the underemployment rate commensurate with full employment is around 8.5 percent. Just in case you think that’s unrealistically low, note that underemployment bottomed out at 8 percent in the last expansion and 7 percent in the 1990s boom (see figure).

Thus, we’d better learn something about underemployment – what it is, why it’s still elevated, and what can be done to knock it down to where we need it to be.

Underemployment, called u6 by the Bureau of Labor Statistics, combines three groups: the unemployed, part-time workers who’d prefer full-time jobs (workers who are literally underemployed), and “marginally attached” workers (a small subset of people neither working nor actively looking for work, but who would be willing to work if there were more good opportunities available).

The big story in the rise of u6 was and remains the involuntary part-timers (IPTs). Without at all dismissing the uniquely stark psychological costs of unemployment (especially long-term joblessness), when computing labor market slack, it actually makes more sense to think in terms of desired hours worked. Someone who wants to work full-time can, in reality, work between zero hours (unemployed) and 40 hours (fully employed). This continuum, versus the 0-1 metric of unemployed/employed, provides a more nuanced and accurate measure of labor market slack.

As the figure shows, one reason it’s taking so long for u6 to come back down is because it went up so much in the recession, climbing about 3 percentage points more than unemployment (u6 rose about 8 points; unemployment, about 5 points). What’s important, albeit obvious, here is that what drove u6 up so high was the big, negative demand shock known as the Great Recession. That’s a cyclical increase, one that can be (and is being) reversed by strong job growth, as opposed to a structural increase, one we’d expect to persist even as the economy strengthened.

Some who disagree with that assessment blame the rise in IPTs on Obamacare. Since the law includes employer mandates, phased in over 2015-16, that apply to firms with over 50 full-time workers, some argue that the Affordable Care Act (health-care reform) is already causing employers to move full-time employees into part-time work to avoid the mandate.

Not so. The Bureau of Labor Statistics breaks data on part-time work into two categories: voluntary and involuntary. The data clearly show a decline in IPTs and an increase in voluntary part-timers. As I show here, the decline in IPTs is of a magnitude you’d expect given the pace of the recovery. As Dean Baker shows here, the increase in voluntary part-time work has been largest among young parents, which is an unequivocally positive development. These parents likely had to work full-time to obtain health insurance in the pre-Obamacare world, but now, as Dean explains, “It seems the ACA [is] allowing young parents the option of working part-time in order to spend more time with their kids” while still getting coverage.

If this trend proves to be lastingly true, unlocking the “job lock” described above will turn out to be an extremely worthy and heretofore underappreciated aspect of health-care reform. In the meantime, the evidence suggests that Obamacare is not a factor in the elevated numbers of IPTs (nor, by extension, with the elevated rate of u6). Which makes sense; the vast majority of firms and workers affected by the mandate already have coverage (well under 1 percent of employees work 30 to 34 hours a week and are employed by businesses affected by the employer mandate and do not have insurance).

That doesn’t mean we’ll necessarily be able to get back to the 7 percent of the late 1990s. There are three structural changes underway that actually may have had an impact on the elevated rate: a shift to service industries (like wholesale and retail trade or food services), which are more prone than others (like manufacturing) to using part-timers; demographic shifts to older workers, who, relative to younger workers, tend to prefer full-time work; and a shift toward “just-in-time” staffing, where employers micromanage workers’ schedules based on subtle shifts in demand.

While these structural changes add to IPTs in both bad times and good, they can’t account for much of their rise. It’s also important to remember that it is beyond economists’ skills to accurately peg unemployment and u6 rates consistent with full employment. We’ve long erred on the high side, meaning that in the name of tilting against an inflationary threat that was often a phantom menace, policymakers have accommodated too much slack.

The key point here is that, since the elevated u6 problem is largely a cyclical problem, it should continue to come down as the job market tightens. I estimate that if job growth continues at its current pace of around 200,000 per month, we’ll hit 8.5 percent by early 2017. Should job growth accelerate to, say, 250,000, we could hit that (alleged) threshold by late this year.

For the least advantaged workers to reap the benefits of tight labor markets, reaching full employment won’t be enough. We’ll need to stay there, which means we’ll need to avoid bubbles and preemptive tightening by an overanxious Fed. But that’s a different problem. For now, let’s just get there.