The Agenda

NRO’s domestic-policy blog, by Reihan Salam.

Will Obamacare Raise Low-Skill Wages? Not If Businesses Hire Robots Instead


As Patrick Brennan noted yesterday, the new CBO analysis of the Affordable Care Act suggests that new health insurance subsidies will tend to reduce work effort. That is, the availability of health insurance subsidies will lead some workers who are only working to retain their health benefits to cut back their hours, or exit the labor force entirely, and the means-tested nature of the subsidies will create a situation in which some workers will find that working longer hours (and thus earning more income) will lead them to lose enough in health insurance subsidies to make it a losing financial proposition. The silver lining, according to some observers, is that by depressing labor supply, Obamacare might force employers to raise wages.

Josh Barro of Business Insider walks us through this logic. Assuming wage levels are stable, Obamacare will complicate things for employers. If workers really do choose to work less, as seems plausible, and if firms demand exactly the same amount of labor as they did before the new health insurance subsidies started to influence labor supply, firms will have to raise wages to attract the same number (and quality) of workers. Barro suggests that this positive wage effect will disproportionately benefit less-skilled workers, as these are the workers who will be most likely to reduce their work effort in response to means-tested health insurance subsidies. Moreover, he posits that by improving the bargaining position of workers, who will no longer be desperate for jobs that provide health insurance as they can acquire health insurance without jobs, Obamacare might over time increase the labor share of GDP. This is an appealingly simple Econ 101 model of how the economy might adjust to new health insurance subsidies. But there is a catch.

Among students of immigration, the notion that an influx of new immigrant workers will necessarily drive down the wages of native workers is a source of immense frustration. The reason is that it is not just the raw number of workers that matters. Rather, it is whether the immigrant workers in question are competing with native workers or complementing them. If immigrant workers are perfect substitutes for native workers, then an increase in the number of immigrant workers may well drive down wages for native workers. But if immigrant workers have skill sets that are different from native workers, and that make it easier for native workers to do their jobs well, or to work longer hours, than the arrival of immigrant workers might actually boost wages and incomes for native workers. The bulk of the evidence suggests that new immigrant workers don’t lower the wages of native workers; they lower the wages of earlier immigrants, the workers who are most substitutable for new immigrants. (Dylan Matthews, an immigration enthusiast, summarized some of this evidence last June.)

So to understand the impact of reduced work effort on wages, we don’t just want to know the raw number of workers who will cut their hours, or who will exit the labor force entirely. We want to know more about the composition of this slice of the workforce, and whether it will actually be worth it to employers to draw them back into the workforce. If we assume that Barro is right and that it is less-skilled workers who will reduce their work effort most of all, we need to consider whether the less-skilled workers in question are in occupations that are, for example, extremely difficult to automate, or if they are in occupations where capital can substitute for labor, or where offshoring is a viable option. For example, if the workers who exit the workforce are concentrated in food service and child care, we’ll find that high-skilled workers who depend on such workers to outsource household production and increase their work hours might be forced to reduce their work hours as well, or spend more to employ the workers who remain in the sector (and spend less on other goods and services). If the workers who exit the workforce are concentrated in sectors that are already shrinking, perhaps they’re just hastening the inevitable, and wages for the workers left behind won’t actually increase — we’ll just see voluntary separations take the place of involuntary separations. 

Back in 2010, Tyler Cowen floated the provocative idea that some share of unemployed workers were best understood as “zero marginal product workers,” or workers who weren’t contributing to the bottom line of their employers, and so who could be safely, and profitably, shed. Under healthy economic conditions, employers are reluctant to fire workers, as restructuring a business model is painful. But necessity is the mother of invention.

Let’s say that many workers choose to reduce their hours and firms respond by investing in business models that will reduce their reliance on less-skilled labor. Instead of having ten less-skilled workers, an employer could hire five or six workers with stronger skills (perhaps stronger non-cognitive skills) who could work more effectively with machines, and with each other, than in a more labor-intensive model. This shift, repeated across many firms, could well result in higher wages for the smaller number of workers who will actually be employed. The list of countries ranked by GDP per hour worked looks different from the list of countries ranked by GDP per capita. Workers in some countries, like South Korea, work extremely long hours while workers in other countries, like the Netherlands, work relatively few hours. The U.S. has been climbing the charts when it comes to GDP per hour worked, in part because we have yet to see much of a labor market recovery among the less-skilled, the workers who would tend to drag down GDP per hour worked.

This leads us to the kernel of the left-right conflict. There are at least some thinkers on the left who believe that the best thing public policy can do is give workers, and particularly poor workers, better alternatives to working in the market, so they can have more leverage in negotiating with employers, and also so they can lead their lives as they see fit. But if you believe that it is good and valuable for people to make a meaningful contribution to their own sustenance, because, say, it will give them pathways out of their families and neighborhoods and into the wider society, where they can foster a wider array of “weak ties,” the prospect of a world in which labor force participation declines as people choose not to work thanks to the availability of health insurance and other subsidies elicits mixed feelings (at best).

I see labor-saving business model innovation as a good thing. Indeed, this is why I’m often baffled by claims made by advocates of a sharp increase in less-skilled immigration, who insist that disaster would strike if we didn’t have enough low-wage workers willing to pick lettuce — we already have lettuce bots, which will get cheaper and more effective over time, and in Denmark, convenience stores and fast food restaurants operate with less labor-intensive business models than their U.S. counterparts. Yet I also want to see less-skilled natives in the labor force, to help them build the networks and the assets they need to escape entrenched poverty. Any system of means-tested health insurance subsidies will depress work effort, including health-system reforms that I think would represent a net plus. But when you consider Obamacare as a whole, with its innovation-inhibiting regulations and how it might exacerbate provider concentration in health care markets, you see that its labor market effects are just part of the problem. It’s hard not to conclude that we have to return to the drawing board.


Sign up for free NRO e-mails today:

Subscribe to National Review