As U.S. inflation rates spike to levels not seen for decades, the problem of inflation is again at the center of public debate. Some who raise alarm gesture at the unhappy experiences of the 1970s. When inflation hit 12 percent in 1974, President Ford addressed the nation before a joint session of Congress and called for measures to “Whip Inflation Now” (he even donned a “WIN” button for good measure)—yet inflation remained a vexing problem throughout the decade. Today’s inflation-averse hawks call for aggressive measures such as interest rates hikes, and, perhaps motivated more by politics than economics, urge greater caution with purportedly inflation-inducing government spending. In contrast, others argue that the inflationary surge is transient and best waited out. The arguments on each side have merit, but the important question is not whether somewhat higher inflation is likely to linger, but what should be done if it does.

The important question is not whether higher inflation will linger, but what should be done if it does.

The answer to this question might seem obvious. Inflation is bad and, if it increases, government policy should prioritize bringing it down. But that is only half right. Indisputably, all other factors held constant, inflation is bad. It erodes the real value of money, reducing purchasing power. The same money buys less than it did before, including things like food and shelter.

But all other factors are not constant. “Nominal” values—such as the dollar cost of items—are less important than “real” values such as actual purchasing power. (In the Great Depression, prices fell, but it didn’t matter because no one had money to buy things.) An economy in which nominal aggregate economic growth—or median household income, a better measure of a society’s economic wellbeing—is 6 percent with an inflation rate of 4 percent is doing much better than an economy with nominal growth of 2 percent and an inflation rate of 2 percent. The former is experiencing real economic growth; the latter is stagnant. For the United States, the difference between those two scenarios is enormous, measured in the trillions.

Moreover, all inflation is not the same. It can have different sources, flows, contours, and expressions. All inflation is also not inherently “bad”—in fact, the “optimal” level of inflation for an economy is higher than zero. As recent experiences around the world remind us, deflation —a falling price level—is in almost all cases a more serious economic problem than inflation. More complicating still, even when the inflation rate is clearly “too high,” bad, and costly, introducing forceful measures to bring it down might not be the right response. Counter-inflationary measures such as interest rate increases (and contractionary adjustments to fiscal policy, which in practice offer a long and uncertain trip to price stability), also impose real costs on the economy. Thus, even in those cases where the inflation rate is obviously “too high,” it is essential that the cure not inflict more damage than the disease.

Weighing the relative costs and benefits of inflation-fighting is problematic, because both the economics and the politics of inflation are generally not well understood. Much of what people think they know about inflation is wrong. In particular, while inflation will often acutely impinge upon tight family budgets, the “costs” of inflation for the economy as a whole are very difficult to show. Setting aside hyperinflationary episodes, reams of studies—often undertaken by scholars eager to demonstrate the perilous dangers of inflation—have failed to show any real economic cost to inflation below levels of 20 percent per year, if not more. Nor are very low inflation rates associated with superior overall economic performance.

Moreover, the “optimal” level of inflation is unknowable and context dependent—although we do know that it should be higher than zero. In the abstract, the optimal rate of inflation is probably about 3 percent. There are three reasons for this. First, with or without inflation, relative prices in an economy are constantly changing—prices for certain things rise, while others drop. But some nominal prices are sluggish to adjust downward. An inflation rate of 3 percent allows some prices to rise, say, 6 percent, while others remain flat, allowing for an adjustment of relative prices without requiring nominal price drops. Second, and perhaps more importantly, interest rate adjustments are the most called upon tool of economic policy and can be highly effective. If the economy is overheating, interest rates can be raised, discouraging economic activity; if the economy is slowing down, they can be lowered to encourage it. But the ability to deploy this essential instrument is markedly circumscribed when the nominal interest rate approaches zero, as recent history has once again illustrated. Finally, an average inflation rate of zero is just that—average—and thus risks slipping below zero. We know that modest levels of inflation don’t hurt the economy, but deflationary environments do. Flirting with deflation courts disaster.

Inflation is not inherently “bad.” In fact, the “optimal” level of inflation is higher than zero.

Our theories of inflation and the relationship between the money supply and the inflation rate also do not perform as well as we might hope or assume. Milton Friedman famously proclaimed inflation is “always and everywhere a monetary phenomenon,” which sounds eminently plausible in theory and may indeed capture descriptive correlations in the extremely long run. But as a practical matter, history has demonstrated that there is no simple relationship between “money” and “inflation.” The money supply is hard to measure and even harder to control. Moreover, the relationship between changes in the money supply and changes in the inflation rate has not followed the expectations of economic models. Although it would be surprising if a large expansion of the money supply did not to lead to more inflation, we have been surprised many times, especially in the recent past. We can speak with even less confidence about fiscal policy, despite the (again deductively plausible) argument that higher levels of government spending tend to have an inflationary effect. Again, in practice this relationship is highly contingent, complex, and context dependent. Recall the conservative pundits’ confidence that President Obama’s stimulus measures would lead to runaway inflation. In fact, from 2009 to 2016, inflation averaged about 1.4 percent.

We do know that some low to moderate rate of inflation allows for the maximum possible levels of employment and economic growth. We also know that, although some inflation is permissive of those ideal prospects, it is necessary to guard against the fallacy that inducing inflation might lead to greater economic growth. It will not, except perhaps fleetingly by creating a short-term “money illusion” when people confuse nominal with real changes to economic signals. Ultimately, inducing inflation will lead to little more than more inflation. To deploy the Keynesian metaphor, “you can’t push on a string.” Giving the leash more slack will not prompt an old dog to move at a faster pace, but pulling on the leash will stop her in her tracks. Tight monetary policy can indeed strangle the economy.

Inducing disinflation (reductions in the inflation rate) may sometimes be necessary, but it is always costly. More restrictive macroeconomic policies largely succeed in taming inflation by reducing demand and depressing real economic activity, not via some theoretically imagined smooth and costless adjustment of expectations about future prices. Disinflationary policies are unambiguously associated with output loses. Indeed, inflation hawks’ strongest argument depends on this fact. Inflation can’t be allowed to “get out of hand,” they insist, because the costs of reducing it are so high. A common point of reference, and often formative experience, for such hawks is the severe recession of the 1980s that was then viewed as necessary to bring down the rampant inflation of the 1970s. They thus urge aggressive, proactive measures to keep inflation extremely low as a way to avoid that horror. But that choice is akin to Bismarck’s characterization of preventive war as “suicide from fear of death.”

The larger point is that understanding the specific cause and context of an inflationary episode is more important than simply aiming to avoid all inflation. Inflationary pressures come from numerous sources. At times, overly lax monetary policies were important drivers of inflation, such as during the late Johnson and early Nixon administrations. In the latter, Fed Chair Arthur Burns was strong-armed by Nixon—who attributed his failed 1960 presidential run to a sluggish economy—into opening the monetary spigots in an attempt to goose the economy in anticipation of his 1972 reelection campaign.

If reckless or politically motivated monetary expansions are at the root of the problem, a strong case can be made for taking the steps necessary to reel them in. But an economy heating up and growing apace presents a different problem. The cost of labor might be bid higher, nudging inflation. Indeed, economic growth can generally be associated with inflationary pressures. Here suppressing the economy to quell inflation may still be the right choice, but it may not. Whatever policy is chosen—whether active or passive—will create winners and losers. Monetary policy is never neutral. Setting a target for the inflation rate is always and everywhere a political phenomenon.

Policies designed to lower inflation increase unemployment and hold down wage pressures, disproportionately burdening those on the margins of the labor market.

The rate of inflation at most levels may have very little effect on the performance of the overall economy, but all levels of inflation have varied effects on different economic activities and interests. The economic costs of falling prices or soaring hyperinflation can be substantial. However, in almost all circumstances, the distributive effects of different inflation rates—and, crucially, of different policies chosen to influence inflation—will vastly outweigh their effects on the aggregate economy.

This is uncontroversial. No one would dispute that unanticipated inflation tends to benefit debtors at the expense of creditors (one reason why the financial community is so averse to inflation). Friedrich von Hayek spoke for legions of conservatives with his assertion that inflation is little more than theft—engineered by governments to transfer wealth from private to public hands. More generally, such effects are ubiquitous. Inflation is costly to those outside the workforce on fixed, non-inflation adjusted incomes. In contrast, policies designed to lower inflation increase unemployment and hold down wage pressures, disproportionately burdening those on the margins of the labor market.

These distributional effects tend to be underappreciated because macroeconomists, appropriately, tend to focus on the aggregate economy. But economists of all ideological stripes know these things to be true. John Maynard Keynes understood that inflation is “not spread evenly or proportionately” across the economy. Rather, it will affect the “distribution of purchasing power” with “social and economic consequences.” Committed anti-Keynesian Charles Rist knew the same truth: “Price stability, just as much as price instability, gives rise to inequalities.”

Various actors in and sectors of the economy will inevitably be affected differently by the policies chosen. Some sectors of the economy will be more than happy to have aggregate economic growth slow to lower the inflation rate—the overall economy may slow, but their narrow self-interests might be well served. A deceleration of economic growth will not be felt uniformly; it will vary across actors, sectors, and regions. But everybody in the economy gets the same interest rate policy. At best it is naive to fail to recognize that every possible choice will privilege some interests over others. And it is too easy to wrap preferred policy choices in the righteous cloak of economic efficiency when, in reality, no such singular path beckons.

Ultimately, even when guided by the best-intentioned expert, efforts to manage money and inflation will invariably involve guesswork, instinct, and mistakes. The theories called upon are imperfect and always evolving, the relevant data lagged and noisy, causes and effects slippery and unclear. Even if an abstract platonic ideal could be imagined, in practice, macroeconomic and monetary policies will almost always miss that target and end up either “too loose” or “too tight.”

Deflation —a falling price level—is in almost all cases a more serious economic problem than inflation.

The choice of which mistake to make depends on the primary drivers of an inflationary episode. Dialing back inflation induced by overly permissive monetary policy makes sense; tightening monetary policy into a supply shock is a mistake. It slows the economy, suppresses wages, and contributes to unemployment all in the service of treating the symptom of an unrelated problem in the real economy. (Econ 101 rightly tells us that the best policy solution is the one that comes closest to the source of the problem.)

Today’s inflation has four principal sources, which come largely from the supply side; none are monetary in nature. One is the crisis of affordable housing in many major cities—the expense that most people spend most of their money on. The others are related to the COVID-19 pandemic: severe supply chain interruptions and bottlenecks; pent up consumer demand set to be unleashed as economic activity returns to normal (similar to the post World War II surge in demand for long unavailable durable goods); and the heating up of the economy as businesses (especially, for example, the travel, hospitality, dining and entertainment sectors) reemerge and seek to expand their payrolls after staggering through an artificial, pandemic-induced slowdown.

Of these, only the latter suggests prioritizing policies designed to fight inflation, such as tighter monetary policy (or, much less directly, greater caution with government spending). Raising interest rates, for example, will not solve the problem of supply chain disruptions—it will simply make them less salient by visiting a different form of misery, tamping down aggregate demand for products struggling to find their way to the showroom floor. Solutions for the bulk of today’s inflationary problems require addressing problems on the supply side of the equation.

This means that, for now, we ought to err on the side of inaction, letting this inflationary wave pass through the economy. That is no small thing and will be felt acutely, especially by those living at the limits of their budget constraints. But there is no pain-free solution to this problem. Present circumstances suggest that the best of the largely unpalatable policy options available would be to reach for fiscal and regulatory measures to address supply side bottlenecks, provide temporary government support for those facing rising energy and housing costs, and otherwise tolerate, if through gritted teeth, some period of somewhat higher inflation.

This approach is, regrettably, extremely unlikely to be embraced. A powerful, psychological component to inflation will almost certainly force the hand of government to take more aggressive anti-inflationary measures than it should. People don’t like inflation. They don’t like slower economic growth or rising unemployment either—but they really don’t like inflation. And they notice it very quickly because, unlike most other economic maladies, they are confronted with it every day.

Many years ago, I asked a senior Federal Reserve Bank officer why the Fed was so aggressively suppressing any hint of inflation, when the costs of moderate inflation to the economy were impossible to show, but the costs of their disinflationary disposition were so significant and clear. “Because people don’t like inflation,” he explained. That answer was not grounded in high economic theory, but it was probably right.

For now, we should let this inflationary wave pass through the economy.

People’s visceral aversion to increasing inflation is disproportionate to its real effects on their economic wellbeing. In some ways, rising inflation rates are like rising crime rates—they provoke feelings of anxiety, uncertainty, and a loss of control over one’s destiny. Such feelings of insecurity are extremely powerful motivators. When crime is rising, people often demand that authorities “do something” about it—anything—even if that “something” is unwise, harmful, and barely relevant to solving the problem. Similarly, the anxiety provoked by moderate levels of inflation will lead people to support, and even demand, that urgent steps be taken to suppress it—even if, when all the math is done, those measures leave them worse off.