On Memorial Day weekend in 1988, George Herbert Walker Bush emerged from his family compound in Kennebunkport, Maine, to deliver a warning. Genteel, dapper, blue-blooded, and careful, the presidential candidate cut a decidedly un-Trumpian figure; “he had always seemed a little like Scott Fitzgerald made him up,” an acquaintance once remarked. The setting that weekend sharpened the contrast: In place of the swampy, chandeliered indulgence of Donald Trump’s Mar-a-Lago resort, the Bush-family retreat was quintessential New England. But the content of Bush’s message presages Trump, and his subsequent actions suggest how one aspect of Trump’s presidency might evolve.

Bush’s target was the Federal Reserve, which he feared might strangle the economy and, should he win the election, weigh down his presidency. “As a word of caution: I wouldn’t want to see them step over some [line] that would ratchet down” economic growth, Bush told reporters, masking his warning with patrician courtesy. By today’s standards, it was a milquetoast protest; in last year’s campaign, Trump crudely accused the Fed of keeping interest rates low to get Hillary Clinton elected. But Bush’s meaning was evident. He was using his platform as the presumptive Republican presidential nominee to question the Fed’s competence in setting interest rates.

The Kennebunkport warning gave way to a full-blown attack after Bush’s election. Administration officials took to TV to urge low interest rates; Bush filled vacancies on the Fed’s board with political allies. His Treasury secretary, Nicholas F. Brady, tried to punish the hawkish Fed chair, Alan Greenspan, by excluding him from parties. There would be no more invitations—none!—Brady decreed. “Whoosh! Boom! Stop!” he sputtered. The administration’s budget director, Richard Darman, put the word out that there was something creepy about Greenspan, a then-unmarried 65-year-old who called his mother every day. Perhaps he was a bit like Norman Bates, the mother-fixated figure in Alfred Hitchcock’s Psycho?

Bush ultimately paid for his Fed-bashing populism. Greenspan stood up to the pressure and kept short-term interest rates high; meanwhile, the pressure itself—the open attempt to meddle with the Fed for political gain—caused investors to fear future inflation. As a result, long-term interest rates also remained high—banks, for example, wouldn’t lower the mortgage rates they offered, thinking that they might need high rates in order to make money on the loans if inflation picked up down the road. That, in turn, dampened economic growth, contributing to Bush’s loss in 1992. Bill Clinton and his successors learned a lesson from Bush’s self-inflicted injury: Don’t criticize the Fed; don’t even comment on it. Since then, Fed independence has come to seem like a given, as solid as the independence of newspapers or the courts.

Until now, that is. Under President Trump, it is possible, for the first time in a generation, to imagine a concerted attack on the central bank. Conceivably, the United States could repeat the story of the mid-1960s and ’70s, when a 15-year period of central-bank independence was brought to an end by presidential bullying. Back then, Lyndon B. Johnson summoned the Fed chairman, William McChesney Martin Jr., to his Texas ranch and shoved him around the living room while proclaiming that low interest rates were imperative in a time of war. “Boys are dying in Vietnam and Bill Martin doesn’t care!” he yelled. Martin ultimately delivered the looser money that Johnson wanted. Richard Nixon followed up by publicly smearing Martin’s successor, Arthur F. Burns, until he, too, complied. Because Martin and Burns, unlike Greenspan, buckled, the U.S. went through the most extreme bout of inflation in its peacetime history.

Might Trump repeat this pattern? Gary Cohn, the top White House economic adviser, and Steven Mnuchin, the Treasury secretary, are Wall Street pragmatists by background. They know that bald attacks upon the Fed can backfire. In an interview with The Wall Street Journal in April, Trump himself sounded a conciliatory note, saying of Janet Yellen, the Fed chair, “I respect her.” But Trump also demonstrated a willingness to go after the Fed during his campaign. He has bashed other institutions and experts, including the courts and the media. And Trump’s outsized expectations for economic growth make conflict between the Fed and the White House seem all too plausible.

All presidents face a dilemma regarding the economy. They are judged in large part according to how wages, jobs, and retirement nest eggs perform on their watch. Yet they have little control over that performance. Growth, ultimately, is determined by long-term (and in some cases mysterious) factors: demographic trends, business innovation and technological progress, the education level of the workforce. Under the right circumstances, some measures—tax cuts, government spending—can boost growth, at least for a while. But for the most part, presidents cannot quickly influence the deeper elements that govern growth.

There is, however, one lever that seems temptingly close to their grasp. If the Fed can be persuaded to hold down interest rates, cheap loans can boost home purchases, car purchases, and business’s spending on factories and machines, pumping up demand and juicing the economy. The Fed’s power is especially tantalizing because of the technocratic tidiness of its decisions—a single committee of experts sets the short-term interest rate as it pleases, with no need to run the gantlet of lobbyists, advocates, and congressional committees. In the long run, of course, lower interest rates are not a magic elixir. The extra demand may run ahead of the economy’s ability to supply things, causing scarcity that leads buyers to bid up prices, thus boosting inflation; and inflation, once permitted, can be tamed only by means of painful job losses. But in the short run, a Fed-created sugar high can transform a president’s fortunes.

Now, after years of rock-bottom interest rates following the 2008 crisis, the possibility of hiring bottlenecks and price pressures has reappeared on the horizon. Having recovered painfully and slowly from the crash, the U.S. economy is expected to grow by more than 2 percent this year—not very fast, but faster than the roughly 1.8 percent that the Fed considers to be sustainable without increasing the rate of inflation. The labor market, after all, is tight: Headline unemployment stands at 4.5 percent, considerably below its average of 6.2 percent since the start of 2000. The broader measure of unemployment—including workers who have given up looking for jobs and part-time workers who’d prefer to work full-time—tells a similar story. With workers now relatively scarce, companies must offer more to attract them. Higher wages, when matched by higher productivity, are a good thing. But if wages rise merely because of worker scarcity, companies may have to pass on the costs to consumers, stoking inflation.

Given these facts, the Fed has little choice but to hike the short-term interest rate from its current low level—if inflation is allowed to accelerate too much, workers will pay a terrible price later. Sure enough, the Fed has already started down that path, lifting borrowing costs in December and then again in March; two more hikes are expected before 2017 is over. In a normal political climate, this might feel routine. After all, the Fed is still paying people to borrow, in the sense that its lending rate is negative after accounting for inflation. But today’s political climate is far from normal. If Trump believes even part of his own rhetoric, his reaction to Fed tightening could well become aggressive.

Trump officially maintains that the economy can grow at an annual rate of 4 percent. Some of his advisers have tried to dial back this expectation: Mnuchin has said that growth of 3 percent is achievable. But even that is way above the Fed’s 1.8 percent estimate of sustainable growth. If the Fed, acting on its judgment of the safe speed limit, continues to raise interest rates, it will be announcing that the administration’s growth ambitions are delusional. The president, for his part, can be expected to believe that the monetary gurus are conspiring to frustrate his promises to voters.

Higher interest rates do not merely dampen growth; they do so through specific channels. Interest-rate-sensitive parts of the economy get squeezed first; the prime example is real estate, which may not be welcome news to this particular president. The tradable parts of the economy also suffer, because higher interest rates attract capital from abroad, putting upward pressure on the dollar and hence making it more expensive for foreigners to buy American goods. That will appeal even less to Trump, because the most tradable sector of all is manufacturing.

During his campaign, Trump pledged to protect blue-collar workers in the industrial swing states. If the Fed sustains a strong dollar, precisely those workers will suffer. Trump likewise pledged to cut the trade deficit. A strong dollar may cause its expansion. Even Trump’s election promises about immigration may be undone. The stronger the dollar, the greater the incentive for a Mexican worker to earn wages in the U.S. and send money home to relatives.

In sum, the White House and the Fed are likely to find themselves at loggerheads. The question is how the parties to this conflict will choose to behave. Trump may indulge his belligerent instincts, or he may listen to his pragmatic counselors. The Fed, for its part, may cave in to pressure, as it did under Martin and then Burns. Or it may resist, following the Greenspan model.

As a street-fighting defender of the Fed’s independence, Greenspan was a master. During his showdown with George H. W. Bush’s administration, the Treasury tried to get a bill through Congress that would have curbed the Fed’s regulatory power; Greenspan used his relationships with lawmakers to bury the initiative. When Bush’s lieutenants came after him, whispering slanders to the press, they got a taste of their own medicine: Greenspan was on friendly terms with journalists, and he could plant stories better than anyone. So skillfully did Greenspan manage his reputation that he proved impossible to unseat. The Bush team reluctantly appointed him to a second term, fearing that removing him might shake Wall Street’s confidence.

Janet Yellen will struggle to replicate some parts of the Greenspan model. Whereas Greenspan had strong ties to both Republicans and Democrats, Yellen lacks Republican allies—a vulnerability, given the makeup of today’s Congress. Whereas Greenspan operated in pre-Twitter Washington, Yellen faces a vicious media free-for-all. Yet there is one big historical lesson that Yellen can apply. And she holds an ace, if she is willing to use it.

The lesson is that it pays to manage the Fed’s board ruthlessly. In the 1980s, Ronald Reagan’s team undermined Greenspan’s predecessor, the redoubtable Paul A. Volcker, by appointing administration loyalists as Fed governors. Toward the end of his tenure, Volcker lost votes on three occasions; with at least four of the seven governors on the Fed’s board prepared to gang up against him, he no longer fully controlled his own institution. Greenspan applied the dark arts of bureaucratic politics to avoid this fate. When Clinton appointed a potential challenger as Fed vice chairman, Greenspan sidelined him so firmly that he eventually left (some possibly not-coincidental press criticism may have encouraged his departure). When Clinton tried to appoint a troublemaker in his stead, Greenspan used his Senate connections to block confirmation. Since Greenspan’s retirement in 2006, fashion has swung against his domineering style; Fed governors feel free to express their views in public, and the power of the interest-rate-setting Federal Open Market Committee is less concentrated in the chair. But now, with the Fed’s independence in peril, the pendulum must swing back. Three of the seven governorships currently stand empty. Trump will get the Fed he wants unless Yellen actively resists.

The ace that Yellen holds is that, although her term as chair expires in February, her appointment as a governor runs to 2024. Fed chairs usually resign from the board when their term expires, but they are not obligated to do so. If Trump refuses to keep her in the driver’s seat, she could remain on the Fed’s board and do some vigorous back-seat driving. The last chair to stay on—Marriner S. Eccles, in 1948—proved devastatingly effective. By force of character and intellect, he remained an influential voice, achieving his full revenge in 1951, when he helped lead a Fed revolt against the president who had demoted him.

If, despite recent conciliatory signals, Trump were to drop Chair Yellen, a back-seat-driving Governor Yellen could be formidable. Her public pronouncements might sway markets more than those of the new chair; she could lead a posse within the interest-rate-setting committee, and her backers might include the heads of the regional Feds, whose appointments are largely free of presidential influence. The mere prospect that Yellen might do this could be enough to cause the administration to back down. The Fed’s independence is not enshrined in law, but a determined central banker with the stomach for a fight can find ways to sustain it.