The U.S. economy continues to operate way below estimates of its potential that were made prior to the onset of financial crisis in 2007, with a shortfall of gross domestic product now in excess of $1.5 trillion — or $20,000 per family of four. Just as disturbing, an average economic growth rate of less than 2 percent since that time has caused output to fall further and further below those estimates of potential. Almost a year ago, I invoked the concept of “secular stagnation” in response to the observation that, five years after the financial hemorrhaging had been stanched, the business cycle was not cycling back to what had been previously thought of as normal levels of output.

Secular stagnation, in my version, has emphasized the difficulty in maintaining sufficient economic demand to permit normal levels of output. Given a high propensity to save, a low propensity to invest and low inflation, it has been impossible for real interest rates to fall far enough to spur the economy to its full potential, since nominal interest rates cannot fall below zero.

Given the various factors — rising inequality, the lower capital costs needed to enter dynamic businesses, slowing population growth, increasing foreign reserves and greater spreads between borrowing and lending cost — operating to reduce natural interest rates, it continues to seem unlikely to me that, as currently structured, the U.S. economy is capable of demanding 10 percent more output with interest rates that are consistent with financial stability. So demand-side secular stagnation remains an important economic problem.

But as the work of Northwestern University economist Robert J. Gordon in particular points up, it may well be that now supply-side barriers threaten to hold back the economy before constraints on the ability to create demand start to bind. Two ways of looking at the situation point up the difficulty.

First, while I have emphasized that U.S. GDP is still far short of what pre-crisis trends predicted, the unemployment rate, now at 6.1 percent, has reverted most of the way back to even relatively optimistic estimates of its normal level. In other words, even as growth has been poor, it appears that demand has been advancing rapidly enough to substantially reduce slack in the labor market. As Gordon rightly emphasizes, weak growth along with significant decreases in labor slack suggest a major slowing of the growth of potential output.

To be fair, one can quarrel with the use of the headline unemployment rate as a measure of slack in the labor market. But the degree to which the labor market appears to be normalizing is even greater if one looks at measures of job openings and vacancies, new unemployment insurance claims or short-term unemployment.

Second, with Friday’s relatively weak employment figures, monthly job growth has averaged more than 225,000 since February. If this trend continued, what would happen to unemployment? This, of course, depends on what happens to labor force participation, which has been trending down in recent years because of the aging of the population and long-term structural trends. Assume, for simplicity’s sake, that participation rates remain constant (an optimistic assumption) and that the economy keeps on creating 200,000 jobs a month. A simple calculation reveals that the unemployment rate would fall to the 4 percent range by the end of 2016.

While such a low unemployment rate is conceivable, it seems more likely that employment growth will slow at some point, either because of employers having difficulty finding workers, rising wages or government policy decisions. In any of these cases, the economy would be held back not by a lack of demand but a lack of supply potential.

Why has supply potential declined so much? This will be hotly debated for years to come. Part of the answer lies in the effect of past economic weakness. Part of it is the brutal demographic realities of an aging population, the end of the trend toward increased women’s labor force participation and the exhaustion of the gains that could be won from an increasingly educated workforce. And part is the apparent slowing of productivity growth.

To achieve growth of even 2 percent over the next decade, active support for demand will be necessary but not be sufficient. In the United States, as in Europe and Japan, structural reform — to both increase the productivity of workers and capital and to increase the number of people able and willing to work productively — is essential. Infrastructure investment, immigration reform, policies to promote family-friendly work, development of energy resources and improvements to the business tax system will become ever more important.

Lawrence Summers is a professor at and past president of Harvard University. He was treasury secretary from 1999 to 2001 and economic adviser to President Obama from 2009 through 2010.