In December 1965, President Lyndon Johnson met with Federal Reserve Chairman William McChesney Martin at the president’s surprisingly modest ranch in the Texas Hill Country. Johnson was upset with Martin for tightening credit despite Johnson’s expressed preference for more accommodative policy. At one point, the president began pushing the Fed chairman around the room, haranguing him with one of his patented hard sells, “Martin, my boys are dying in Vietnam, and you won’t print the money I need.”

Martin resisted Johnson at the time, but eventually moved policy in the direction the White House demanded. Looking back years later, Martin, who was dubbed the “happy Puritan” by one journalist, lamented, “To my everlasting shame, I finally gave in to him.”

The idea that monetary policy should be made independent of political influence is widely (though hardly completely) accepted today. As former Fed Chairman Ben Bernanke (2006-14) noted, “Careful empirical studies support the view that more-independent central banks tend to deliver better inflation outcomes than less-independent central banks, without compromising economic growth.” But as the LBJ anecdote suggests, the Fed did not always enjoy the degree of independence it has today. Before the era that began with Paul Volcker (1979-87), political influence on monetary policy was the rule rather than the exception.

The independence of today’s Fed is supported by its unusual institutional framework. But the primary bulwark against interference is psychological, driven less by laws and regulations and more by the convictions of key players that the economy is better off with a central bank that can exercise broad discretion in monetary policy. Hence a president inclined to dismiss Bernanke’s “careful empirical studies” who is abetted by a Congress unwilling to resist meddling could reverse the four-decade precedent.

How it Works in Theory …

The independence of the Fed depends upon two key factors: the will of the governors to ignore threats from politicians unhappy with their policies, and the ability to pay its own bills without appropriations from Congress. First, though, let me put on my pedant’s hat to remind you of how the arcane system is actually cobbled together.

Start with the role of the leadership of the regional Federal Reserve banks. The presidents of these 12 public-private hybrid agencies are appointed to five-year (renewable) terms by the boards of directors of each Reserve Bank. The Fed board of governors in Washington can veto the appointments. But, importantly, the selection process is independent of the White House and Congress, and neither has the authority to remove regional presidents.

While the regional banks’ primary task is to do most of the routine non-policy work of the Fed, five of the regional bank presidents also serve on the Fed’s monetary policy making body — the Federal Open Market Committee — on a rotating basis. (Only the head of the New York Fed, which you might say is more equal among equals, is a permanent member.)

In addition to the five rotating regional bank presidents, the FOMC includes the seven members of the board of governors. The U.S. president appoints the governors, subject to Senate confirmation, to 14-year nonrenewable terms. The long term gives board members a perspective that extends beyond the president who appointed them, while the one-term limit removes their temptation to skew policy in the sitting president’s direction in order to curry reappointment.

To further distance monetary policy from the direct influence of the president, the governors’ appointments are staggered so that one governor’s term expires every other January. This is supposed to limit the number of appointments any single president can make, with the goal of preventing the White House from stacking the board with allies.

One of the seven governors is chosen by the president to be chairman or chairwoman of the board for a renewable four-year term. Although this assures the president’s appointee a high profile role in policy deliberations, the law doesn’t give the chairman or chairwoman any extra votes to influence the decisions of the FOMC. Indeed, the whole convoluted design is supposed to facilitate a process in which key stakeholders are represented in policymaking, but no single interest can control policy.

… And in Practice

A little — well, not a whole lot of — history belongs here. The Fed was created in 1913 in response to the well-founded conclusion that the financial system of a modern economy can’t run on autopilot. But the system we have today, with its focus on managing the supply of money rather than on assuring commercial bank solvency, is the product of reforms in 1935 made in response to the Fed’s disastrous performance in the early years of the Great Depression.

Marriner Eccles, the first Fed chairman under the new system, believed it was his job to support FDR’s economic recovery effort, so for all intents and purposes, White House policy was Fed policy. The Fed’s independence was thus not tested.

That changed after World War II. During the war, the Fed had pegged the interest rates paid on government bonds in order to contain the budgetary costs of borrowing for the war effort. This commitment effectively put the Fed out of the business of making monetary policy. Inflationary pressures driven by defense mobilization were managed with price controls and taxes rather than by squeezing private spending with tighter credit.

Once the war was over, though, Eccles saw the need to return to what we now think of as conventional monetary policy. However, when he tried to end the interest rate peg — and the deference the Fed had shown to Treasury edicts since 1935 — President Truman decided to bring in someone who would continue to let Treasury call the shots.

In January 1948, Truman replaced Eccles with his own pick: Thomas McCabe, the former CEO of the Scott Paper Company. However, Truman was in for two surprises. First, Eccles did not resign from the Federal Reserve Board, instead choosing to stay on as a regular board member. Second for the next three years he pursued his public crusade against pegging interest rates at the expense of inflation fighting.

The Treasury countered with its own public campaign. And the tense standoff finally led the warring parties to hammer out a deal known as the Fed-Treasury Accord of 1951. This fuzzy pact acknowledged the Fed’s need to respond to inflationary pressure, but also recognized the Treasury’s need for “stability” in government bond markets. Part of the accord — though it wasn’t in the actual text — was that both Eccles and McCabe would step down. Two Treasury insiders, James Robertson and William Martin, replaced them, with Martin taking over as chairman.

Truman assumed that Martin would continue the tradition of subordination to the Treasury. But he was surprised once more. Though Martin paid lip service to the interest rate peg, he made it clear that the peg wouldn’t last much longer. Thus, the Fed-Treasury Accord marked the beginning of the end of the Fed’s long honeymoon with the White House.

But there was still a long way to go before the Fed exhibited the degree of independence we view as par for the course today. Martin still believed the Fed had an obligation to support new Treasury bond issues — which translated as not allowing the financing of budget deficits to drive up interest rates at a pace determined by the market. As a result, despite Martin’s concerns about inflation, monetary policy remained relatively accommodating.

Before concluding that Martin lacked a backbone, it’s important to remember that Congress was threatening to force the Fed to support the interest rate peg if Martin didn’t fall into line. Since, in the end, the Federal Reserve is a creature of legislation, congressional threats inevitably carry weight with the board of governors.

Martin survived the Johnson years, but not the wrath of Richard Nixon. Nixon believed Martin’s management of the economy had cost him the 1960 election, and he wanted someone he could trust to facilitate his agenda.

That someone was Arthur Burns, a conservative economics professor at Columbia University who was appointed Fed chairman in February 1970. During his time in the hot seat, Burns talked the talk of Fed independence, but rarely walked the walk. Indeed, he politicized his office by keeping interest rates down to support Nixon’s re-election in 1972. The modest independence Martin had extracted in the Fed-Treasury Accord was all but gone. And, almost needless to say, the few times that Burns did deviate from the will of the White House, he was met with threats from Congress.

Burns also strove to concentrate power in the hands of the Fed chairman, and did what he could to strong-arm FOMC members who would not endorse his — or perhaps more accurately, Nixon’s — policies. Among other abuses, Burns asked Nixon to kick one problematic governor upstairs by giving him an ambassadorship, and he persuaded Nixon to write a letter to a board member demanding his cooperation.

To further consolidate power, Burns also tried to influence the selection of regional Reserve Bank presidents by threatening to veto appointments. This illustrates the reality that seeding the Fed’s infrastructure with checks and balances is not enough to guarantee independence. Even a rule that appears to insulate members of the FOMC can be short-circuited by a determined president.

The consolidation of power in the hands of the Fed chairman not only deviates from the intent of the law to give a wide variety of interests a voice in monetary policy, it also makes Fed independence more vulnerable to assaults from Congress and the White House. When monetary policy is set by the 12 members of the FOMC with one vote each, the control of the chair by an aggressive president does not mean that monetary policy will be dominated by the White House. When power is concentrated in the hands of the chairman through the exercise of bureaucratic or extra-legal initiative, such White House dominance is much more likely.

Independence can also be compromised if the president is able to make more appointments to the board of governors than the statute intends. Although board members are appointed to 14-year terms, few serve this long. The median term, excluding the chairmen and chairwoman, is only five years. As a result, most presidents have been able to appoint a majority of the governors — if not all of them — by the end of their second terms.

The reality that board members almost never stick around for their full terms has consequences that go beyond allowing presidents to stack the board. A loophole in the legislation creating the Fed allows a newly appointed governor to serve out the remainder of the term of a governor who has resigned, and then to be reappointed to his or her own 14-year term once the partial term is completed. It turns out that most appointments are of this type. Hence, at some point, most governors can be considered for reappointment, giving them incentives to court favor with the White House.

The final important change under Nixon relating to Fed independence was the decision in 1973 to abandon the gold standard. This had large unintended consequences for the Fed’s ability to fund itself.

When the Fed was created in 1913, it was freed from “the power of the purse.” Reserve banks funded themselves by providing various paid services to private banks and through earnings on loans to private banks to cover the latter’s required reserves. The Federal Reserve Board (which was then organized quite differently — for example, it was chaired by the Treasury secretary and its role was mostly to regulate commercial banks) covered the cost of its operations through an assessment on the Reserve banks.

This changed in 1935 when power was centralized within the board of governors and the Fed was given the authority to conduct “open market” operations — that is, to alter the money supply by buying and selling Treasury securities. When the Fed increases the money supply, it uses newly printed money (actually, electronic bank reserves) to buy government bonds. It then collects interest on those bonds — more than it needs to fund itself — so Fed is freed from the congressional appropriations process and the strings that often come attached.

Indeed, the Fed makes far more than it needs to fund the Federal Reserve System. It turns over the remainder — which was an astonishing $92 billion in 2016 — to the Treasury. In recent years, this has become a major source of federal revenue.

Congress was worried that this self-funding ability would be abused when the modern Fed was created in 1935, but was satisfied that the gold standard would limit its ability to self-fund through expansions in the money supply. (I won’t bore you with the reason.) Once the gold standard was abandoned, though, there were no constraints at all on the Fed’s capacity to create money — which was not what had been intended when authority to conduct open market operations was granted in 1935.

Since the financial crisis began in 2007, the Fed has used this authority to buy trillions of dollars worth of securities — yes, trillions — as a means of easing credit beyond the point possible by reducing interest rates. But for many in Congress, this novel extension of policymaking constituted more independence than they thought the Fed should have. Once again, the result was congressional threats to hobble the Fed.

As discussed above, the Fed was far from independent under Nixon. But by the time Nixon was replaced by Gerald Ford, reducing inflation became Priority 1. At this point, the Fed and the president were mostly on the same page, and the president had little incentive to intervene.

The next big challenges to Fed independence began with Jimmy Carter. During his first presidential campaign in 1976, Carter argued that the economy needed additional monetary stimulus to create jobs and expressed displeasure with Arthur Burns’s focus on inflation fighting. Burns made no secret that he returned the enmity. So it was not surprising when, in March 1978, President Carter replaced Burns with William Miller.

In fact, Miller, the former CEO of an industrial conglomerate (Textron) with little knowledge of central banking who might have been expected to follow the White House’s lead, did not fully cooperate. Carter’s economists wanted tighter money and Miller failed to deliver. After just under a year and a half, Carter recruited him to be secretary of the Treasury, replacing him with a financial technocrat, Paul Volcker.

When Volcker began his term in August 1979, the inflation rate was 13 percent and the unemployment rate was 6 percent — a miserable combination dubbed “stagflation.” Volcker made it clear before accepting Carter’s offer that reducing inflation would be his priority and that he would operate independently. Carter agreed, but his deeds did not always track his words. For example, he lobbied for looser monetary policy before his re-election bid, but Volcker paid no attention.

When Carter lost the election to Ronald Reagan, the new president reappointed Volcker. He was not Reagan’s first choice, but Reagan feared upsetting financial markets by replacing him. Inflation remained a big problem, both economically and politically, so the Reagan administration shared Volcker’s desire to bring it under control and chose not to interfere with Volcker’s shock-therapy approach.

The cost of fighting inflation was the deep 1981-82 recession. Some members of the Reagan administration did call for looser monetary policy at this time, and Congress weighed in with the now-familiar threat of legal action if Volcker persisted. But Volcker’s success in bringing down the inflation rate cemented the idea that an independent Fed could do what elected officials could not. Thereafter, it became conventional wisdom that the Fed should run free.

The Fed, Ascendant

The value of an independent Fed was reinforced by contemporary academic research. Finn Kydland and Edward Prescott showed that a “time inconsistency” problem — making promises to reduce inflation in the future and then finding it undesirable to do so when the time arrives — would cause higher than desirable inflation rates unless the monetary authority was able to credibly commit to a future inflation target. Elected officials are generally not credible when they make promises about future inflation; hence the value of an independent (and therefore highly credible) central bank.

Presidents do not often appreciate Fed chairmen who exercise independence, and in August 1987, Reagan replaced Volcker with the business economist Alan Greenspan. Thus began an 18-year run: Greenspan was reappointed by George H.W. Bush, reappointed twice by Bill Clinton, and reappointed once again by George W. Bush.

Though he certainly met frequently with the presidents he served under, such was his prestige that, if anything, influence ran from the Fed to the White House — particularly where the budget deficit was concerned. The economy had its ups and downs during the Greenspan years, but the fluctuations were mild. When financial crises and recessions threatened the economy, Greenspan was able to take effective action in concert with the Treasury. In general, the economy performed remarkably well.

Should Volcker and Greenspan — more relevant here, the degree of independence they exercised — be given credit for sustaining the 1982-2007 period of price and growth stability dubbed “the Great Moderation”? From the perspective of 2017, it’s not clear. Though Greenspan’s reputation would eventually be tarnished by his failure to contain the housing bubble that led to the Great Recession, contemporaries were inclined to celebrate both the Fed’s power to stabilize the economy and the success of the Fed governors when left alone to do their job.

When the Princeton economist Ben Bernanke replaced Greenspan in 2006, support for the idea that the Fed should have a free hand to manage the economy was at an all-time high. Presidents might take a run at jawboning the Fed into compliance, but the institution had built the political capital during the Great Moderation to act as it wished.

One important change Bernanke did make was to reverse the growing dominance of the Fed chairman under Volcker and Greenspan. During his term, and continuing with the tenure of Janet Yellen, policy decisions became a cooperative FOMC effort rather than being stage-managed from the top. This, by the way, is much more in accordance with how the law says the Fed is supposed to operate. And the decentralization, combined with ongoing efforts to seek consensus on the FOMC before it acts, probably serves to strengthen the Fed’s independence.

After the FOMC cut interest rates to almost nil during the Great Recession, the committee took advantage of its independence to boldly go where no Fed had gone before. With its “quantitative easing” program, the Fed purchased vast quantities of securities other than Treasury bonds, flooding banks with loanable deposits. President Obama respected the Fed’s independence, probably because it was pursuing policies that his advisers found appealing in light of their own frustrations with being unable to use more fiscal stimulus to speed the recovery. But the reaction from Republicans in Congress was a familiar one: threats to intervene.

This was partly inspired by genuine fears that the magnitude of the quantitative easing initiative would unleash inflation, and partly by a partisan inclination to attack an institution seen as allied with the Obama administration. It’s hard to know the extent to which pressure from Congress constrained the Fed’s actions. But the threats to audit the Fed, monitor policy decisions and strip it of its regulatory powers, combined with the public’s anger at policies that served the interests of Wall Street, certainly motivated Bernanke to mount an unusually public campaign defending the Fed’s independence.

The congressional challenge to the Fed’s independence has receded under Janet Yellen’s leadership. Appointed by President Obama in February 2014, Yellen’s job has been to guide the Fed’s exit from the policies put in place during the Great Recession without slowing — or inadvertently reversing — the recovery.

This means deflecting criticism from the left that the recovery is too slow and criticism from inflation hawks asserting that interest rates have been too low for too long. Even so, the sense that Congress was prepared to take a significant bite out of the Federal Reserve’s independence has faded as the economy has recovered.

Through a Glass Darkly

Faded, but not vanished. There is some congressional support for pressuring the Fed to follow a rule-based monetary policy that mechanistically ties short-term interest rates to inflation and output. The likely impact of such a move on the Fed’s independence is ambiguous. On the plus side, if the Fed is following a rule endorsed by Congress, it is much harder for politicians to blame the rule followers. On the minus side, any rule would function as a limitation on Fed discretion.

Currently, the Fed governors are very much opposed to this idea. But to the degree one can talk of a Republican monetary policy, it is strongly influenced by John Taylor of Stanford who is a strong proponent of rule-based intervention.

To date, President Trump has not directed his ire toward the Fed. But that could change when Janet Yellen’s term as chairwoman ends in February 2018. If Trump chooses to replace her (as seems likely), a big question will be whether he will appoint someone with the inclination and self-confidence to act independently. One must assume that if the economy turned south during President Trump’s tenure, he wouldn’t hesitate to try to redirect the blame toward the Fed.

An important question relating to how the Fed will evolve under Trump is the composition of the Federal Reserve Board. At present, there are three openings on the board because congressional Republicans refused to confirm Obama’s nominees. The names of possible Trump nominees floated to the date of this writing all apparently support faster interest rate increases, a rollback in financial regulation and adherence to a set rule for determining changes in the money supply.

If the attrition rate among current board members follows the historical pattern, one could expect more positions to open soon. Stanley Fischer has already resigned. However, the prospect of President Trump stacking the board with people eager to reverse existing board policies may motivate the other governors appointed by Obama (Lael Brainard, Jerome Powell and Janet Yellen) to serve until the end of Trump’s presidency. Whether they will in fact stay on, I suspect, rests in part on who is chosen to fill the open seats and who is appointed as the new chairman or chairwoman.

History shows us that independence cannot be guaranteed by legislation. It requires presidents to appoint qualified people and then to support their discretion. More generally, it depends upon the willingness of official Washington to put the nation’s long-term interests above partisanship. On that score, it’s hard to claim we’re making progress.