How long can America create jobs without unemployment falling?

You would expect strong job growth to be accompanied by falling unemployment, but America is proving that one does not always entail the other. Over the past year, employment is up by fully 3 million people, but the unemployment rate has stayed around 5 percent.

In fact, a few more workers are unemployed than a year ago. The reason is that more Americans are seeking jobs. Over the past 12 months, the labor-force participation rate of so-called "prime-age" workers — those between 25 and 54 — is up by just under 1 percentage point, the fastest growth recorded since January 1989.

The recent surge in prime-age participation follows a long decline from its peak, 84.6 percent, scaled in January 1999. Between then and September 2015, it tumbled by an average of about a fifth of a percentage point a year.

Hence, the refrain of some that low unemployment is a mirage: Stronger economic growth, they say, could draw plenty of folk back into work. Some 1.8 million people say they want a job, but are not technically in the labor force because they have not recently looked for one.

Restoring the employment-to-population ratio among prime-age workers to its pre-crisis high would require getting all these hands to work and then some: A total of at least 2.9 million new jobs would be needed.

This reservoir of potential workers remains untapped. Rising participation does not reflect more Americans getting off the couch to seek work. Rather, the long-term unemployed are less likely to stop looking for a job. The probability of a worker who has been unemployed for 53 weeks or more leaving the workforce in a given month is about 25 percent today, down from over 30 percent at the end of 2015, according to Goldman Sachs' Zach Pandl and David Mericle.

This wealth of potential workers suggests that participation could rise much further.

A new paper by Alan Krueger of Princeton University finds that participation is only loosely connected to short-term swings in the economy. Instead, Krueger identifies several underlying factors holding it down.

For instance, even within the prime-age bracket, the population has been aging. Those aged 45 to 54 have long been less likely to work than those aged 25 to 34. And prime-age men outside the workforce are in startlingly bad health. Nearly half of them take painkillers daily; 34 percent report at least one disability; and 25 percent receive disability benefits. A vast majority say that their disability is a barrier to employment.