President Donald Trump is conducting a risky experiment on the U.S. economy: He's allowing the government to run large budget deficits — some of the largest ever outside wartime or recession — in the hopes that this will somehow put growth on a higher trajectory. Irresponsible as that might sound, it actually makes some sense.

In the long run, economic growth is a function of two variables: population and productivity. For decades, America had plenty of both. Birthrates were ample, and any additional labor could be attracted from elsewhere. From 1947 to 2007, workers' output per hour grew at an average annual rate of 2.3 percent. So for the most part, American presidents could focus on improving rather than reviving growth.

But since the last recession, the picture has changed. Labor-force growth is slowing as baby boomers retire. For a variety of reasons, some understood and some not, productivity has decelerated as well. The Obama administration largely accepted the new reality: In a 2016 report, it projected inflation-adjusted gross-domestic-product growth of just 2.2 percent for the next decade, and it offered fairly traditional ideas such as immigration reform, more cross-border trade, infrastructure spending and education investments.

Trump has taken a very different approach, aiming for annual growth of 3 percent over the next decade. This certainly won't come from population, particularly given his administration's attitude toward immigration. That leaves productivity, which some of his policies don't do much to encourage, either. Tariffs on imports such as steel and aluminum will serve largely to make output more expensive. Tax cuts might prompt companies to make more productivity-enhancing investments, but the effect will likely be modest given uncertainty about how long the cuts will remain in place.

So the whole game becomes a big bet that deficits — created by the government's tax cuts and spending plans — will boost productivity growth. Treasury Secretary Steven Mnuchin suggested as much last week when he said that the Trump administration's policies could lead to wage growth without inflation and that people shouldn't worry about the forthcoming deficits. Ironically enough, this policy was espoused by the Bernie Sanders campaign (as my colleague Noah Smith has noted). The idea is that by running the economy hot and making labor more expensive, the government can induce businesses to do more investment than they would in a normal economy. Ever since the financial crisis, a weak economy has discouraged businesses from investing, leading to weaker productivity growth — so why not try the opposite? It's a theory that hasn't been tested in recent decades, but an intriguing one.

What are the potential risks and rewards? Sticking with the status quo promises more of the same underperformance — annual real GDP growth of about 2 percent. The deficit experiment has two possible outcomes. In the best case, the U.S. gets some form of productivity miracle. In the other, rising inflation forces the Fed to raise interest rates to cool off the economy, triggering a recession.

Mnuchin

Most policymakers, economists and investors aren't worried about a period of inflation like what the world experienced in the 1970s. In advanced economies, central banks have the tools they need to fight it. Slow productivity growth, by contrast, has become a real concern, especially as countries seek the resources to take care of aging populations and still invest in their futures.

Republicans and Democrats may disagree on the best way to create deficits, whether it be tax cuts and military spending or investments in infrastructure and education. But the balance of risks leans toward trying this experiment. Be it the Trump administration or the next, someone was eventually going to take the gamble.