Of course it’s hard to know for sure, but many economists would argue that a one percent increase in the number of hours Americans work would increase potential gross domestic product — potential GDP; potential output; the amount of stuff the economy could produce — by about seven-tenths of a percentage point. Extrapolating from this, if we could increase hours of work by 10 percent we’d increase the potential size of the economy by 7 percent.

Some economists would quibble with that calculation. But all economists would agree that the more Americans work, the more stuff the economy can produce. Jobs are an extremely important component — indeed, the key component over the short run — to economic growth.

And of course the opposite is true. If we have fewer jobs and if fewer people are participating in the workforce, then the ability of the economy to produce goods and services will shrink quite a bit.

With that in mind, let’s review briefly the trends in workforce participation.

First, some terminology is required. For the purposes of this paper, let’s consider only people between the ages of twenty-five and fifty-four. I focus on that age range because it contains people who are both too old to be in school and too young to retire — people in their prime working years. Each such person is classified by the Bureau of Labor Statistics (BLS) as either employed, unemployed, or not in the labor force. Workers are employed if they work for pay. Workers are unemployed if they don’t have a job, are trying to find a job, and are available for work. The labor force is the sum of employed and unemployed workers. Persons who are neither employed nor unemployed are classified as not in the labor force. The labor force (or workforce) participation rate is the share of the population which is either employed or unemployed.

The labor force participation rate for prime-age workers stood at 64.2 percent at the beginning of 1948. It rose steadily in the fifties and sixties, picked up steam in the nineteen-seventies, and continued to rise until about 1990, when it leveled off in the 83-84 percent range. For about the last fifteen years, the share of the prime-age population in the workforce has been declining. Currently, the rate is lower than it’s been in three decades.

It may surprise you to learn that over this entire period — from the end of the Second World War to the beginning of the Great Recession to the present day — the rate at which men have been participating in the workforce has been falling. The four-decade increase in workforce participation I just described is due to women entering the labor market.

In fact, the average monthly workforce participation rate for men in each decade dropped relative to the previous decade. From a peak of 97.9 prime-age men out of every 100 participating in the workforce in September 1954, the participation rate stood at 90.9 percent when the Great Recession began, bottomed out at 87.9 percent in October 2013, and has equaled about 88 percent this year. This is a staggering decline. And the trajectory of male workforce participation directly threatens the prospects for long-term growth, the health of society, and the ability of men to lead flourishing lives.

My purpose here is to propose two actions that could move us towards halting and, hopefully, reversing the downward trend in prime-age male workforce participation in an effort to shore up the prospects for long-term growth.

The first thing I would do — and I list this first because it is seldom discussed despite its clear importance — is increase the capacity of the federal government to collect social and economic data.

In order to (1) better understand why fewer men are working, and (2) understand how public policy can best attempt to reverse that trend, we simply need to know more than we know today about workers, potential workers, and the tasks firms want them to do.

The truth is that we don’t know enough about any of these. And much of what we do “know” is often inferred rather than measured. True, there is some great research out there that attempts to explain why male labor force participation has fallen over the decades, and there are some good ideas about how to pull it back up. But it would be an overstatement to say that there is a broad consensus among economists, and I don’t know of any practicing labor economist who would argue that better data aren’t needed to fully understand what’s driving the trends and how to reverse them.

We want to know why men have been participating in the workforce at lower rates over time, and what it would take to get nonparticipants back in. But the monthly Current Population Survey (CPS) — the source of our monthly information about labor force participation — doesn’t ask the obvious question to nonparticipants: what would it take to get you to come back to the workforce? The answers to this question could generate a wealth of information about why some people are not in the labor force, and how we can reincorporate them into it — and ergo how we could increase the prospects for long-term growth by increasing workforce participation.

We know how many workers are working “part time for economic reasons.” But, more broadly, the monthly CPS doesn’t ask whether workers — both part- and full-time — are fundamentally satisfied with their jobs, think their job is a good fit, or consider their current job part of their long-term career plan. We don’t know whether full-time workers get as many hours as they would like. These sorts of “satisfaction questions” could help us understand not only how many workers choose to leave the workforce, but why they do so.

We want to know why some “observationally equivalent” men work and others don’t. But we don’t really understand how people get jobs. If the CPS asked newly employed workers detailed questions about how they found their jobs, we would learn so much more about the process of job search and acquisition. We should ask questions about labor market networks, job referrals, and other issues of vital importance. This would help us to design policies to make finding jobs easier. And if finding jobs is easier, then more men will work, and long-term growth prospects will improve.

And the monthly CPS should evolve as the labor market evolves. My understanding is that the BLS is planning on introducing questions about licenses and credentials, in addition to questions about levels of education. This is important, because licenses and credentials are growing in importance relative to traditional degrees. We know very little about the “underground economy” — unreported jobs, odd jobs, little bits of work picked up here or there in the new “sharing economy,” and illegal activity. At least the “sharing economy” — Uber, Airbnb, etc. — is expected to grow over the medium term, but it’s hard to imagine that if the trends continue the other categories won’t grow as well. The CPS should be ready to collect and interpret data on twenty-first century styles and modes of work.

In addition to augmenting the CPS, there are other steps we should take. To administer the unemployment insurance (UI) program, the government already collects records on nearly all American workers. But those records contain very little information. If we were to add questions about the hours an employee works, the employee’s occupation, and the actual tasks the employee performs, then we would know a lot more about what is happening in the labor market. Since the UI records can be linked to data on workers and firms, this additional information would help us tremendously in our efforts to understand why some men are working and others aren’t, to design policies to pull more men back into the workforce, and to strengthen long-term growth.

The press pays a lot of attention to the monthly jobs report, which tells us how many new jobs were added to the economy on net. The BLS also reports data on gross labor market flows — on the total number of workers who quit or were laid off from their job, and who were hired into new jobs, among other measures. These statistics are available at very broad industry and regional levels. But they are not available at the state and MSA level, they do not have detailed industry breakdowns, and they do not break down by occupation or by job task.

Flows data on occupation, hours, and job tasks in specific geographical areas would help us understand some of the most important questions regarding the workforce participation of men. Why are middle-class jobs disappearing? Why are wages falling or stagnating? Which occupations and tasks allow workers to climb the income ladder when they lose a middle-class job, and what types fall down into lower-wage jobs? These answers doubtless have a lot to do with what workers are doing at work — with (from the economist’s perspective) the black boxes that are firms’ production functions — and adding information on hours, occupation, and job tasks to flows data would greatly improve our understanding, and help us design policies to shore-up male workforce participation in the face of powerful twenty-first century labor market trends.

Critically, we also need better “longitudinal” data — data that track individuals every year (or even more frequently) for a long period of time. The CPS gives us snapshots in time (and can be used to construct short panels), but even better would be to follow individuals for decades. Some workers leave the labor force and never return. Some leave and eventually return. Some leave, return, leave, and return. Some never leave. If we had detailed information about workers’ work history and socioeconomic background, then we could better understand which workers leave, which never return, and how to design policies to help those who are on the margin of returning to reenter. And the more who reenter, the greater the economy’s productive capacity.

We have the basic infrastructure for better data in place. All we have to do is up our game. The major federal statistical agencies need larger budgets to collect the data we need to design policies to increase workforce participation and to strength future growth. And in the context of labor market information and statistics, this applies especially to the BLS. Data is a public good, as well as an investment that will pay dividends in the future.

Having said that, it is a safe bet that one reason fewer men are working is that real wages for male workers without a college degree have been stagnant or falling for decades. So my second policy suggestion is to expand the Earned Income Tax Credit (EITC) for childless workers, many of whom are low-income men. The EITC is a federal earnings subsidy: if you work, and if you earn less than a certain amount, then the government will supplement your earnings with a transfer payment. The EITC offers very little support for childless workers, with a maximum credit of only about five-hundred dollars. This amount should be significantly expanded, as both President Obama and Rep. Paul Ryan have suggested. Previous expansions of the EITC have lifted millions out of poverty, and are designed to incentivize nonparticipants to return to the workforce. When they do, everyone wins — the economy has more workers and can produce more goods and services, and the new participants can earn their own success in the labor market, leading flourishing lives that include the dignity only work can provide.