Why high inflation makes central banks raise rates
The answer
Raising interest rates does not make anything cheaper. It makes borrowing expensive, which reduces how much people and businesses spend, which reduces the number of buyers competing for the same goods. Prices then rise more slowly. A central bank raises rates because it cannot increase supply, so it reduces demand to meet it.
Why this costs you money
The mistake is not misunderstanding the theory. It is reading the wrong number.
In 2021 the US federal funds rate was near zero and inflation was rising past 5%. Many people looked at the low rate and concluded that policy was normal, or even that the Fed had already done a lot by ending emergency purchases. The opposite was true. When the policy rate is near zero and inflation is 5%, money is being lent at a real cost of about minus 5%. Borrowing was cheaper in real terms than it had been at almost any point in modern history.
What did that cost people? Anybody who bought long-dated bonds or long-duration bond funds in that period, believing rates were already low and could not rise much, took a large loss when the policy rate went from near zero to above 5% within 16 months. The signal that it could happen was sitting in plain sight: the real policy rate was deeply negative, which is a description of a spring under compression.
The same error appears in reverse at the other end of a cycle. When the policy rate is 3 percentage points above inflation and people describe the central bank as "still fighting inflation", policy is in fact very tight and something usually breaks.
How it works
A rate rise reaches prices through 5 channels. All 5 work by reducing demand.
1. Borrowing becomes expensive. Home loans, car loans and business loans cost more, so fewer are taken. Fewer houses are bought, fewer factories are built, fewer people are hired to build them.
2. Saving becomes attractive. Higher deposit rates make holding money more appealing than spending it today.
3. Existing borrowers have less to spend. A floating-rate borrower whose payment rises has less money left for everything else, starting at the next reset date.
4. The currency tends to strengthen. Higher rates attract foreign money looking for yield. A stronger currency makes imports cheaper, which lowers the imported part of inflation directly. For a country that imports most of its oil, this channel is large.
5. Expectations change. If firms believe the central bank will keep raising until inflation falls, they price next year's contracts more modestly, and workers ask for smaller increases. This channel works before any of the others and it is the reason central banks talk so much.
The trade-off is the mechanism, not a side effect
Every one of those 5 channels works by making somebody spend less. Less spending means fewer orders, slower hiring and sometimes job losses. A central bank raising rates into an inflation problem is deliberately slowing the economy. This is not an unfortunate consequence of the policy. It is the policy.
Anybody who describes a rate rise as costless has not stated the mechanism.
The number that actually matters
Real policy rate = policy interest rate − inflation rate
This is the single most useful number in monetary policy and it is almost never in a headline.
- Real policy rate well below zero: money is cheap in real terms. Policy is loose, regardless of how many times rates have been raised.
- Real policy rate around zero to 1%: roughly neutral, in most estimates.
- Real policy rate 2 to 3 percentage points above zero: policy is tight. Interest-sensitive parts of the economy start to contract.
The middle band is the uncertain one. The genuinely neutral real rate — the level that neither speeds up nor slows the economy — cannot be observed. It is estimated, the estimates disagree, and they are revised afterwards. Treat the bands above as rough, not precise.
What it tells you, and what it does not
A rate rise tells you the central bank believes demand is running ahead of supply, or that inflation expectations are drifting.
It does not tell you inflation will fall. If inflation comes from a supply shock — a war closing a pipeline, a monsoon failure destroying a crop — higher rates do nothing to the shock. They only reduce demand until it fits the smaller supply.
It does not affect all prices equally. Rate rises hit housing, cars, capital goods and anything bought with credit. They barely touch the price of onions. This is why a central bank facing a food price spike often does nothing, and is then accused of ignoring inflation.
It works with a long delay of roughly 4 to 8 quarters, so a central bank is always acting on a forecast and will always be criticised at the moment it acts.
The decision rule
Judge policy by the real policy rate, not the nominal one.
If the policy rate is below inflation, policy is loose and rates have further to rise, however many increases have already happened.
If the policy rate is 2 or more percentage points above inflation, policy is tight, and the risk has shifted from inflation to whatever breaks first — usually the most indebted borrowers, then property, then employment.
The exception: if the inflation is a supply shock and the shock is passing, the same nominal rate becomes tighter every month without anybody doing anything, because the inflation term is falling.
That last line is worth reading twice. A central bank that holds rates steady while inflation falls is tightening policy, not pausing it.
Try this now
Five minutes with 2 web pages. You will finish knowing whether policy in your country is loose or tight, which is a judgement most commentary skips.
- Look up your central bank's current policy rate. In India, the repo rate is on the front page of rbi.org.in. In the United States, the federal funds target range is in the most recent FOMC statement on federalreserve.gov. Write it down with the date of the decision.
- Look up the latest inflation print. In India, the CPI press release on mospi.gov.in. In the United States, the CPI release on bls.gov, and the PCE release on bea.gov. Write down both the headline and the core figure.
- Subtract. Policy rate − headline inflation = the real policy rate. Do it a second time using core inflation.
- Now do the same subtraction on yourself. Take your own loan rate and subtract your personal inflation rate from the first article in this cluster. That is your real borrowing cost.
- Take your deposit rate, subtract tax, subtract your personal inflation rate. That is your real saving rate.
What you should see. The 2 official answers in step 3 will often differ by more than a percentage point, and which one you use changes the conclusion. That disagreement is the actual policy debate, reproduced in 30 seconds.
Steps 4 and 5 give you something no headline will. If your real borrowing cost is negative, borrowing is cheap for you whatever the news says. If your real saving rate is negative, holding cash is costing you, whatever the deposit rate looks like.
Three real cases
1. The Federal Reserve, March 2022 to July 2023 — the full swing, in 16 months The Fed began raising in March 2022 from a target range of 0% to 0.25%. US CPI inflation was already above 8%, so the real policy rate was roughly minus 8%. Over 11 increases the range reached 5.25% to 5.50% in July 2023. By then headline CPI had fallen to about 3%, so the real policy rate was about plus 2.4%. The nominal rate rose by around 5.25 percentage points. The real rate moved by more than 10. The second number is the one that did the work.
2. The Reserve Bank of India, 4 May 2022 — an unscheduled meeting India's CPI inflation for April 2022 came in at 7.79%, above the top of the 2% to 6% tolerance band. The MPC met between scheduled meetings and raised the repo rate by 0.40 percentage points to 4.40%, and raised the Cash Reserve Ratio at the same time. The announcement was not on the calendar, and Indian equity indices fell sharply on the day. By February 2023 the repo rate had reached 6.50% across a sequence of increases. The lesson in the timing: an off-cycle move is a statement about urgency, and the market reads it that way.
3. Türkiye, 2021 to 2024 — the natural experiment Between 2021 and 2023 the Turkish central bank cut its policy rate while inflation rose, on the argument that high interest rates cause inflation rather than restrain it. Annual inflation passed 80% in October 2022 and the lira lost most of its value against the dollar. In mid-2023 policy reversed, and by March 2024 the policy rate had been raised to 50%. This is the closest thing economics has to a controlled experiment on the question, and it was run at the expense of a country's savers.
The question that resolves it
A novice sees a rate rise and asks: how high will they go?
An expert asks: where is the real rate now, and where does it need to be?
The first question has no answer, because the destination depends on data that has not arrived. The second question has an answer today, from 2 published numbers and one subtraction.
What would make this wrong
If raising rates did not slow demand, the whole framework would fail. The clearest disconfirming case would be a country that raised rates sharply and saw neither borrowing nor inflation respond. Türkiye ran the opposite experiment and got the predicted result.
Three honest limits.
Supply shocks. When inflation comes entirely from supply, rate rises reduce inflation only by reducing output. Reasonable people disagree about whether to apply the cure.
The neutral rate is unobservable. Everything above depends on knowing what "neutral" is. Nobody does, and estimates are revised by as much as a full percentage point after the fact.
Distribution. A rate rise falls on people with floating-rate debt, on small businesses that borrow, and on first-time home buyers. A household with no debt and large deposits gains. The aggregate number hides a transfer, and the transfer is the part that reaches individual lives.
In India
The MPC's mandate is written in law: 4% CPI inflation with a tolerance band of 2% to 6%. If average inflation is outside that band for 3 consecutive quarters, the RBI must write to the government explaining the failure, the reasons, and the remedial actions.
That accountability shapes behaviour. The RBI reacts to the headline number, including food, even though rate rises cannot affect vegetable prices. It also publishes a quarterly inflation forecast, which is the thing to read if you want to anticipate the next move rather than react to it.
The RBI has a second lever the Fed no longer uses: the Cash Reserve Ratio. Raising the CRR removes money from the banking system without changing the policy rate. In May 2022 the RBI used both at once.
India also imports the large majority of its crude oil. A rate rise that supports the rupee therefore has a direct and fast effect on domestic fuel and transport costs. This channel is much stronger in India than in the United States.
In the United States
The Federal Reserve has a dual mandate written by Congress: maximum employment and stable prices. When the 2 conflict, the FOMC must state how it is balancing them.
That mandate makes US policy slower to tighten than a pure inflation-targeting mandate would be. A rate rise expected to raise unemployment is a cost the Fed is legally required to weigh, not an externality it can ignore.
The Fed targets core PCE inflation in practice, which strips out food and energy. So a US oil price spike does not push the Fed to act in the way an Indian food price spike pushes the RBI.
The Fed has a second instrument: the size of its balance sheet. Selling bonds, or letting them mature without reinvestment, raises long-term interest rates without touching the overnight rate. Between 2022 and 2024 the Fed did both at once, so the tightening was larger than the fed funds range alone suggests.
Where they differ, and what that tells you
The mandate. The RBI has one target: inflation, with a band, in law. The Fed has 2, in law, with no numerical employment target. So a weak jobs report is a reason for the Fed to hold or cut. In India, a weak growth number is a consideration but not a mandate.
The measure. India targets headline CPI, which is 36.75% food. The United States watches core PCE, which excludes food entirely. The same global food price shock produces a policy response in one country and nothing in the other.
The external constraint. India imports most of its oil and runs a current account deficit in most years. If the RBI holds rates far below the Fed's, money leaves, the rupee falls, and imported inflation rises — which forces the RBI's hand anyway. The Fed has no equivalent constraint, because the dollar is what everybody else is trying to hold.
What that tells you. The RBI's decisions are more constrained and more mechanical to anticipate. Watch 3 things and you will usually predict it: the headline CPI print against the 6% upper band, the gap between the repo rate and the fed funds rate, and the rupee. The Fed has more discretion, so its decisions depend more on the language in its own statements. The next article is about how to read those.
Carry this
- Rate rises do not lower prices. They lower demand until it fits supply.
- Real policy rate = policy rate − inflation. That is the number that says loose or tight.
- Holding rates steady while inflation falls is tightening, not pausing.