Donald Trump won a narrow victory in the Rust Belt by making a lot of promises he probably can’t keep: that he will bring millions of manufacturing jobs back from Mexico and China; that he will revive coal mining; that he will restore the steel industry in Pittsburgh.

But to “jump-start America,” as Trump pledged in a speech this summer, he won’t need to do any of that. He can deliver for his blue-collar base by following a much simpler path. This set of policies would probably create millions fewer jobs than he promised, widen income inequality and almost certainly fail to produce an industrial revival. Yet it would still render meaningful economic improvement, creating jobs and lifting median wages for white workers without college degrees in Ohio, Pennsylvania, Michigan and Wisconsin. It builds on trends already at work in the economy. And it was pioneered by a Clinton.

If Trump can simply find a way to keep the economy growing even modestly for several years, and if the unemployment rate remains low, he has a good chance of presiding over a period of sustained wage growth similar to what America saw in the mid- to late 1990s, when Bill Clinton was president. Those gains will lift all workers, from elite engineers in Silicon Valley to janitors in Maine, if history is a guide. They might not bring back many factory or mining jobs, but they will boost the paychecks of the men and women who lost production jobs years ago and remain angry about it today.

Best of all for Trump, those gains have already begun. For all Trump’s railing against the “terrible” economy under President Obama, census data released in September shows that typical workers at every level, from the very poor to the very rich, experienced income growth in 2015. More in-depth analyses that track individual workers over time, such as one conducted by former Clinton economist Robert Shapiro, suggest that Americans without college degrees started to see rising incomes as far back as 2013.

There are lots of questions about Trump’s economic plans and how they might affect the economy — more so than with a typical new president, because Trump declined to fill in key details of his proposals on the campaign trail. His platform included an unorthodox mix of traditional conservative and liberal-populist proposals. He would cut taxes, particularly for corporations and the rich; roll back government regulations; open more land for energy drilling; spend big on infrastructure; and reshape America’s trading relationships with the world (this one is hard to forecast, because GOP leaders in Congress disagree with him on it). Contained in those proposals is an easy-to-sketch scenario for how Trump and a Republican Congress could plausibly stoke job growth for years to come. It’s the scenario that many Wall Street economists and traders in financial markets appear to be betting on: Trump and Congress agree quickly on tax cuts, deregulation and an infrastructure push, and major trade changes are delayed or nixed.

You can see the path for Trump in the Wage Growth Tracker tool developed by economists at the Federal Reserve Bank of Atlanta. It documents nominal wage gains — which is to say, the growth of workers’ pay, not adjusted for inflation — across several demographic groups, one of which is workers without college degrees. It shows that wage growth is accelerating for those non-degree-holding workers, from a 3 percent clip at the beginning of 2015 to nearly 4 percent growth this October. At that pace, wages for that group will soon be growing at a rate similar to that in the early years of Clinton’s second term — still a percentage point less than the red-hot growth of the very late ’90s, but nevertheless reminiscent of a period remembered fondly in many economically beleaguered parts of the Midwest and Northeast.

That period, as today, was characterized by low unemployment and modest inflation. In the best-case scenario for Trump, 2017 would look a lot like 1997, which began with unemployment at 5.3 percent, just a few tenths of a percentage point higher than where it is now. Arguably, the greatest success of Clinton’s second term was simply that he avoided stepping on the job growth that continued for the next several years and that eventually drove nominal wage growth above 5 percent per year for non-college workers.

If unemployment keeps falling now like it did then, workers can expect a “slow and steady increase” in their inflation-adjusted incomes over the next few years, says Jonathan Willis, a vice president and economist at the Federal Reserve Bank of Kansas City. To reach the rapid wage gains of the late ’90s, he added, “we would need another three or four years of solid job growth.”

That’s plausible for Trump. He could sign a big tax-cut bill, negotiated with and passed by House and Senate Republicans. Such a bill would almost certainly tilt its benefits, both in total dollars and in percentage terms, to business owners, shareholders and the highest-income earners. The big question, based on the differences between the House GOP blueprint and Trump’s campaign plan, is how much of a break middle-class and lower-income taxpayers would see.

Economists generally expect that those tax cuts, whatever form they take, would stoke job growth by increasing consumer demand. (Even many liberal economists concede that tax cuts for the rich are a form of economic stimulus, though not a very efficient one in their models.) Another boost could come from a deficit-financed plan to spend hundreds of billions of dollars on infrastructure. Conservative economists, such as John Cochrane at Stanford University’s Hoover Institution, contend that Trump could further fuel job growth by repealing the Affordable Care Act and rolling back Obama-era regulations on the financial sector and other industries. Their prediction is that those changes would give businesses more freedom and incentive to invest in America and hire American workers. The new jobs would very likely be in service sectors, such as home health care and retail, often requiring relatively few skills. But thanks to wage pressures, those types of jobs would pay much better than they did in the years after the recession.

There are economic and political risks for Trump in each of those policies, of course. The tax bill and, to a lesser extent, the infrastructure push have the potential to balloon the federal budget deficit. Rolling back Wall Street regulations could raise the risk of another financial crisis and sit poorly with Trump’s populist base, which remains furious at bankers over the last crisis; scrapping environmental rules would hurt the fight to avert catastrophic climate change.

And then there are the Trump proposals Wall Street is discounting right now: his promises to deport millions of immigrants quickly and to threaten tariffs on trading partners such as China and Mexico in an effort to revive the millions of manufacturing jobs lost in the Rust Belt since the turn of the century. Trump’s advisers believe that those policies would stimulate job growth and bring millions of working-age Americans who are currently out of the labor force back to job sites. Most forecasters say the policies would, instead, dampen job growth by reducing the labor force and raising prices on imported consumer goods, and would possibly push America into recession.

Avoiding recession is a must for Trump if he wants to maintain his Rust Belt appeal. Since the 1970s, recessions have followed a damaging pattern for workers in traditionally male, blue-collar industries, including mining, manufacturing and construction. When downturns hit, those industries shed jobs. When good times return, they rarely make up any of the lost ground. (In 1970, a quarter of the American labor force worked in one of those blue-collar industries. Today, it’s about one-eighth.)

Crucially for Trump, his tax cut and infrastructure plans alone could end up triggering recession. There’s a good chance — which markets are already pricing in — that those plans could stoke higher inflation. The Federal Reserve could respond by raising interest rates more quickly than expected, which could pump the brakes on growth and possibly cause a downturn.

This is the scenario that should most worry Trump and his team as they craft policy. It’s possible to avoid, though there aren’t many examples in American history of the economy humming along at low unemployment for several years without eventually sliding downhill. What Trump needs is just enough policy fuel to continue growth, without provoking a shock — from a trade war, interest rate hikes or anything else — that would send the economy tumbling.

He’ll need to walk a narrow path to pull it off. Sound familiar?