Robert Pollin is right to see full employment as a desirable goal. But I am unconvinced by much that he says about how to achieve it.

First, Pollin notes, “unemployment would likely have risen to nearly 17 percent in the absence of the stimulus.” But according to the best current estimates, the fiscal stimulus proved far less effective, and for reasons that are instructive for employment policy.

The 2009 stimulus bill included a variety of temporary tax provisions aimed at low-to-moderate income households—Making Work Pay, the expanded Earned Income Tax Credit and Child Tax Credit, among others. The idea was that less affluent Americans were more likely to spend the proceeds than more affluent Americans, who were likely to save. But economists Matthew Shapiro, Claudia Sahm, and Joel Slemrod estimate that only 13 percent of households increased their spending in response to the temporary tax cuts (25 percent of households increased spending in response to the temporary tax cut of 2008). Shapiro, Sahm, and Slemrod found that more pessimistic households—ones that anticipated a decline in future income of 10 percent or more—were more likely to save the additional income than households that had a more sanguine view of their economic prospects. So it seems that in the case of temporary tax cuts, at least, less affluent households are as attuned to their future prospects as more affluent households.

If these observations are generally applicable, they have striking implications for public policy. Stimulus plans that rely on deficit spending can be expected to lead to a heavier tax burden on future workers. But that expected burden will presumably lead low-income families to save, not spend, and that will dampen the prospects for a sustainable jobs recovery.

Second, it is worth thinking about what exactly we mean by full employment. It could be that different groups choose different combinations of household and market labor, and that some groups are more inclined to take part in the informal labor market. This last point strikes me as most salient in the American case.

Long before the Great Recession, the male labor-force participation rate was in sharp decline. Between 1979 and 2009, it had fallen by an extraordinary five percentage points, with the decline concentrated among less-skilled workers. Pollin might characterize this as a consequence of the “neoliberal revolution,” but that terminology is at best imprecise.

There are other reasons why men have been dropping out of work. One of them is mass incarceration. In Reconnecting Disadvantaged Young Men, policy experts Peter Edelman, Harry J. Holzer, and Paul Offner outline the labor-market challenges faced by less-skilled men, particularly ex-offenders. Prison has badly damaged the employment prospects of millions of young Americans. Indeed, I’d suggest that the explosion in the prison population has proven a far more powerful driver of the decline in male labor-force participation than the decline in marginal tax rates, the deregulation of trucking, airlines, and telecommunications, and other elements of neoliberalism. One could creatively attribute rising incarceration rates to rising wage dispersion. The trouble is that we’ve seen rising wage dispersion across virtually all of the advanced market democracies, yet the United States is unique in the scale of its incarceration crisis, which has its origins in the Rockefeller drug laws and other measures that preceded the neoliberal revolution.

Moreover, state agencies have grown more effective at garnishing the wages of non-custodial parents, usually fathers. So consider the choice faced by a young man who is an ex-offender: if he takes part in the formal labor market, he faces a punishingly high effective tax rate and intense labor market discrimination. If he instead enters the informal labor market, he faces great uncertainty but is able to draw on familiar social networks while retaining earnings and avoiding interaction with a state that he may not trust.

We need lower barriers to entry, not a government jobs program.

To address these problems, we need to create a more inclusive labor market by lowering barriers to entry. We might, for example, adopt the economist Edmund Phelps’s proposal for wage subsidies, rather than imposing onerous mandates on employers. New mandates will exacerbate the problem of labor-market exclusivity. (And note that while wages have been fairly stagnant for many U.S. workers, compensation, particularly in the form of health benefits, has been increasing roughly in line with productivity gains.)

Productivity gains have been uneven across sectors and across firms. Within a sector, firms differ dramatically in output per worker, in part due to variation in organizational strategies. As Erik Brynjolfsson and Adam Saunders argue in Wired for Innovation, the 2000s saw the rise of “digital organizations,” which use incentive systems and decentralized decision-making to drive productivity gains. Inevitably, these incentive systems contribute to wage dispersion, as workaholics separate themselves from the pack. Digital organizations tend to be homogeneous in terms of skill levels. While public-sector productivity in the United States lags far behind levels achieved in Singapore, Sweden, Canada, Australia, and other advanced market democracies, our digital organizations flourish.

Crude, economy-wide productivity measures fail to capture the fact that many sectors of the U.S. economy are productivity laggards. But many workers in these lagging sectors have nevertheless enjoyed healthy compensation gains thanks to their political power. This may be a function of the American left’s increasing focus on the interests of relatively privileged public-sector workers as opposed to those of workers on the margins of the economic mainstream.

The truly difficult question is how more workers and more firms can achieve the productivity gains we’ve seen in digital organizations. It seems that, as the economist Dirk Pilat observed in 2004, a key driver of productivity gains is “a process of search and experimentation, where some firms succeed and grow and others fail and disappear.” A more heavily regulated labor market, in which political prerogatives play a more prominent role, will move us in a very different direction. But that seems to be the policy direction that Pollin prefers.

The danger of moving in that direction is that, again, we will create a more exclusive labor market, not unlike labor markets in Brazil and South Africa, where cosseted formal sectors enjoy a wide array of privileges while a large and growing number of workers are left to the vagaries of the informal market. That is not Pollin’s goal. But his preferred policies will, I believe, do more to protect labor-market privileges than to create the more open and inclusive labor market that we need.