The Credibility Tax: Why Central Banks Overreact to Sins They Didn't Commit

When a government spends more than it collects in taxes, it has to find the money somewhere. It can borrow from the public by selling bonds. Or it can do something quieter and more dangerous: get the central bank to create new money and lend it directly to the treasury. Economists call this deficit monetization — covering a budget gap by printing currency rather than raising real resources.

Monetization is seductive because it feels free. No new taxes, no unpopular spending cuts, no bond investors demanding higher interest. The bill arrives later as inflation, because you have added money to the economy without adding goods for it to buy.

A recent NBER paper traces this across sixty years and dozens of countries — and finds something non-obvious at the end: the damage outlives the people who caused it by decades. Here is the chain, one link at a time.

The chain

Link one: left-wing populist governments lean on the central bank. Populist here means a government that frames politics as the pure people against a corrupt elite, and proves it by delivering immediate material benefits — subsidies, transfers, jobs. That costs money, and the fastest source is the central bank. In the data, when a left-leaning populist regime took power, central bank lending to the government rose sharply.

Link two: that lending produces inflation. When the central bank funds the deficit with new money, prices climb. This is not a subtle correlation; large jumps in central bank lending line up with large jumps in inflation.

So far this is the textbook story of Latin American hyperinflations: populism, printing, prices. The paper's real contribution is link three.

The scar

Here you need one piece of machinery. A monetary policy rule describes how a central bank sets its interest rate in response to conditions. The most famous version — the Taylor rule — says: raise the rate by more than one-for-one when inflation rises above target. The coefficient on inflation is the whole game. A bank with a coefficient of 1.5 raises rates 1.5 percentage points for every 1 point of excess inflation. A bank with a coefficient of 2.5 hits the brakes far harder for the same provocation.

Why does the coefficient matter so much? Because of inflation expectations — what people believe future inflation will be. Expectations are self-fulfilling: if workers and firms expect 8%, they demand 8% raises and set prices 8% higher, and the expectation comes true. A central bank's core job is to keep expectations "anchored" near its target, and the size of its policy coefficient is the signal that does the anchoring.

The finding: countries that suffered deficit monetization under past populist regimes now run higher policy coefficients. Faced with the same drift in expected inflation, they react harder.

The obvious objection is that these are just countries with high past inflation, and burned economies stay twitchy. The authors control for past inflation directly — they strip out the level of inflation a country actually experienced and ask whether the history of populist monetization specifically still predicts a fiercer reaction. It does. Two countries with identical inflation histories, but only one with a past of politically captured money-printing, behave differently today: the captured one reacts harder.

Be calibrated about this. This is observational cross-country data, not an experiment; the control for past inflation makes the "it's just high-inflation memory" story much less likely, but it cannot fully rule out that monetized countries differ in some other unmeasured way. Treat it as strong, suggestive evidence for a mechanism that already makes sense on its own terms — which is the next section.

Why the scar makes sense

The mechanism is about trust, and it is not mysterious.

A central bank's power over expectations depends on being believed. "We will keep inflation at 2%" only anchors expectations if people trust the bank to actually do it. In a country whose central bank once printed money on political command, that trust is damaged. The public has learned, from lived experience, that the presses can be captured.

Economists call this experienced learning: people weight events they personally lived through far more heavily than events they only read about. A generation that watched its savings evaporate does not un-learn that lesson from a press release.

So a central bank in the shadow of that history faces a harder problem. Its words are discounted, so it must move expectations using something words can't provide: costly, visible action. A high policy coefficient — slamming rates up hard and fast at the first sign of drift — is the substitute for the trust it doesn't have. When promises are cheap and disbelieved, you signal resolve by paying for it.

The portable model

Here is the idea worth keeping: institutional credibility, once broken, is not restored by fixing the original problem. It has to be over-earned by later actors who did nothing wrong — and the way you over-earn it is by converting cheap signals into expensive ones. Words and promises are cheap; when nobody believes them, you are forced to substitute demonstrable, costly action. That is the credibility tax, and it is paid by whoever inherits the distrust.

This generalises cleanly, and each case shows the same conversion of cheap signals into expensive ones:

• A firm that once shipped a dangerous product runs absurd safety theatre for a decade after its culture has genuinely changed — because "it's safe now" is no longer believed, so it must be shown.
• A person who broke a promise finds "trust me" stops working, and has to substitute visible, verifiable reliability instead.
• A regulator that missed one scandal over-polices the next ten years.

This buys you two specific abilities.

First, a better read on apparent overreaction. When an institution seems to be overreacting — a central bank hiking at the first whiff of inflation, a partner over-explaining, a company drowning in compliance — ask whether it is responding to the present situation or to a past betrayal it is still living down. Often the "overreaction" is exactly rational: it is the extra force required to move expectations when your word is discounted. You have been mispricing it as irrationality.

Second, a warning if you currently hold credibility. Spending it on a shortcut — the free money, the broken promise, the covered-up failure — does not just cost you the immediate blow-up. It levies a permanent tax on every future version of you and your institution, who will have to shout, and pay, to be believed where you once could whisper for free.

Distilled from NBER Working Papers