WASHINGTON — Say it ain’t so, Doug.

From 2009 to early 2015, Douglas Elmendorf was the widely respected director of the Congressional Budget Office, where he repeatedly warned lawmakers that chronic deficits are unsustainable, pose long-term dangers — and that shrinking them sooner rather than later would make the task easier.

Now, economist Elmendorf, who has become dean of the Kennedy School of Government at Harvard, has changed his tune. In a long paper co-authored with economist Louise Sheiner of the Brookings Institution, they conclude that low interest rates mean that “spending cuts and tax increases” — which they concede are inevitable — “should be delayed and [be] smaller in size than widely believed.”

Instead, the government should borrow more at today’s low rates and “invest” in the future — more spending on federal research and development, infrastructure and early-childhood education. “My view has changed,” says Elmendorf. “It’s really cheap for the government to borrow.”

Indeed, it is. In the early 2000s, rates on 10-year Treasury notes averaged about 4.5 percent, notes the Elmendorf-Sheiner paper. The rate today hovers around 2 percent.

Especially in an election year, Elmendorf’s approach may appeal to candidates in both parties eager to avoid cuts in popular programs or middle-class tax increases. But it’s bad policy. It simply postpones, yet again, hard decisions.

As a share of the economy (gross domestic product), the publicly held federal debt is climbing. Elmendorf and Sheiner estimate that, with present policies, it will grow from today’s roughly 75 percent of GDP to more than 120 percent in 2040. This would exceed the debt at the end of World War II (106 percent of GDP in 1946) and triple the average of the last 50 years (38 percent of GDP). These projections could be optimistic; another Brookings study estimates the debt in 2040 at 152 percent of GDP.

How dangerous are these large debt loads? The honest answer is: No one knows. Some highly respected economists think that, in the current situation, the U.S. government’s ability to borrow at reasonable interest rates is virtually unlimited. Demand for “safe” U.S. Treasury bonds is strong, while business demand to borrow is weak (that’s one reason interest rates are low). They could be right.

But they also could be wrong. We shouldn’t gamble. The dangers that Elmendorf cited when he headed the CBO — and repeats in the paper — remain. High debt could make it hard to borrow to meet some future national emergency: a war, economic collapse or pandemic. High debt might trigger a financial crisis, as investors become sated with U.S. Treasury securities. Or it might crowd out productive private investment.

As Elmendorf and Sheiner say, federal debt can’t grow faster than GDP forever. Servicing the debt ultimately becomes economically and politically oppressive. Sooner or later, they note, taxes need to be raised and/or benefits cut. That’s the only way to bridge the gap between rising government spending, driven heavily by higher Social Security and Medicare outlays for an older society, and the tax base.

Low interest rates make this easier by holding down today’s deficits. Elmendorf’s alternative of “investing” to speed up economic growth sounds appealing, but there’s no guarantee that the added funds will achieve their stated goals. Some or all of the “investment” may be wasted or prove unequal to the task. Meanwhile, the politically charged issues of whose programs to cut or whose taxes to raise would be evaded. Similarly, the opportunity to introduce changes gradually — a desirable approach — would be undermined.

Politicians usually prefer to spend more and tax less. The practical effect of Elmendorf’s plan is to cloak political expediency in scholarly respectability. Unfortunately, there’s history here. Waves of economists of varying ideologies have rationalized deficit spending in the name of economic theories promising faster economic growth. In the 1960s, liberal Keynesian economists pioneered this approach; in the 1980s, conservative “supply-side” economists made similar arguments. Elmendorf now joins this dubious tradition.

Robert Samuelson is a columnist for The Washington Post.