Old arguments, new realities, and why the debate keeps coming back
Almost everyone agrees that double-digit inflation wrecks growth. The argument starts once inflation falls into the single digits. After the inflation surge of 2021-2023, most industrial economies have now brought price growth back down to around 3% — roughly where it was in the 1960s. For some policymakers that is still too high. They argue the job isn’t finished until inflation is eliminated completely. Price stability, they say, is the cleanest foundation for long-term growth.
On the other side, a growing group of economists and market watchers say central banks are pushing too hard. Their view is that the problem today is not inflation, but recession risk, high debt, and weak demand. A “little dose of inflation,” they argue, helps cut the real burden of debt and keeps economies moving. So the core question remains: how low should we go? Is 5% acceptable? Is 3% better? Is zero the ideal? Governments have mostly settled on an answer in practice, with targets of 2% or less. New Zealand aimed for 0-2% by the end of 1993, Canada for 2% by 1995, Japan and the Bundesbank for no more than 2%, and Britain for 1-4% in the near term and 2% or less long term. The Bank of England’s governor even planned a speech to make the case for price stability. These targets would have sounded strange a generation earlier, because since the 1930s prices in industrial economies have risen almost every year — by about 1,000% in the US and 4,000% in Britain. Yet history shows that continuous inflation is the exception, not the rule. Between bursts in the Roman Empire, the Middle Ages, and Elizabethan England, prices were often flat for decades. In Britain, the average price level in the early 1930s was no higher than in the 1660s. Those who want a return to that kind of stability argue it will unlock the fastest growth. Their critics say a bit of inflation is healthy, or that the cost of getting to zero is greater than the cost of living with low inflation.
What’s wrong with inflation
If inflation were perfectly predictable, it would barely matter whether it was 0% or 5%. Wages, contracts, interest rates and taxes could all adjust in advance, and the economy would function much the same. The damage comes from the fact that inflation is never fully predictable.
Unforeseen inflation distorts the price mechanism. In a well-functioning market, a 20% rise in apple prices relative to other fruit should signal farmers to grow more apples and consumers to buy more plums. But when general inflation is running at 15%, neither the shopper nor the grower pays much attention to that relative move. The signal gets lost in the noise. Resources are misallocated and growth slows. Even at just 5% inflation, prices double every 14 years, which is enough to swamp most real changes in relative prices. With a stable price level, the market would allocate resources much more cleanly.
The second cost is uncertainty. If businesses don’t know what prices, and therefore real interest rates, will look like in five or ten years, they become cautious. Long-term investment gets postponed in favor of short-term projects. Inflation also pushes up real interest rates, because lenders demand a premium to protect themselves against an unknown future. Taken together, these effects suggest that the ideal inflation rate is the one that plays the smallest role in people’s decisions. In that logic, the answer must be zero. Anything higher introduces unnecessary uncertainty and inefficiency.
Yet some people don’t worry
Other economists push back. They argue that volatility matters more than the level. An inflation rate that averages 0% but swings between +5% and -5% is just as damaging as one that averages 10% and swings between 5% and 15%. From this view, the goal should not be to eliminate inflation, but to stabilize it so people can predict it.
In practice, the last 30 years show that low inflation and stable inflation tend to go together. Countries that kept inflation low also had the least volatile inflation, partly because low inflation creates a virtuous circle of low expectations.
Larry Summers made the most influential case for keeping a bit of inflation, around 2-3%. First, it preserves the ability to have negative real interest rates. In a deep recession, central banks need to push real rates below zero to stimulate demand. With zero inflation, nominal rates can’t fall below zero, so that tool disappears. Second, a little inflation acts as lubricant. Workers resist nominal wage cuts, but will accept a small raise that is below inflation. With zero inflation that flexibility disappears, and the result can be layoffs and more labor unrest.
There is a counter to this as well. When inflation is always present, wage bargaining happens every year and unions stay strong. With stable prices, pay rises would only be justified by productivity gains, which would weaken unions and reduce strikes. Data from that period showed that high-inflation countries were indeed more strike-prone.
The zero-inflation case also ignores a measurement problem. The Consumer Price Index tends to overstate inflation. It does not fully capture quality improvements — a 1992 car cost more than a 1982 car but was far better. And the weights used to build the index are often outdated, so they don’t reflect how people shift away from goods that become expensive. If apple prices rise twenty-fold, people stop buying apples, but apples still have a large weight in the index for years. In the US, the CPI weights in the early 1990s were still based on 1982-1984 spending. Economist Robert Gordon estimated that the US CPI overstated durables inflation by 1.5% a year. A reported 2.2% rise in goods prices could therefore mean goods prices were effectively flat. Other major economies were seeing goods inflation of only 2-3% as well.
Because of this bias, central banks define “price stability” as 0-2% CPI inflation, not zero. If they targeted literal zero, they would in practice be aiming for falling prices. That creates its own problems. Rising prices encourage people to buy now. Falling prices encourage them to wait. During inflation, higher nominal interest rates give people a reason to hold money. But when prices are falling, interest rates cannot go negative, so holding cash becomes more attractive than buying goods. Demand falls, and prices fall further.
That does not mean governments should ignore 2% inflation just because it may equal true price stability. The CPI itself drives wage and price setting. If it keeps rising, countries miss out on the benefits of stability. The lesson is that governments need better price indices, not that they should accept measured inflation at face value.
From theory to fact
Despite strong theoretical reasons to dislike inflation, the hard evidence is mixed. The global slowdown of the 1970s did coincide with high inflation. But the boom of the 1950s and 1960s also happened with relatively low inflation, and that could be explained by post-war reconstruction and trade liberalization just as easily.
Looking at 20 industrial economies, the relationship changed over time. In 1955-1973, when inflation was modest, higher inflation was actually associated with higher growth in GDP per head. Japan was the extreme case, with 5.8% inflation and 8.6% growth. Remove Japan and the link disappears.
After 1973, the pattern reversed. Countries with inflation below 6% averaged 2.1% per capita growth. Those with 6-10% averaged 1.9%. Those above 10% averaged 1.7%. There were exceptions. Ireland, Italy and Spain grew strongly despite inflation above 10% because they started from a low base and had room to catch up, just as Japan had earlier. Switzerland grew slowly at 1.1% a year despite having the second-lowest inflation, because it was already rich.
Another complication was expectations. In the 1950s to 1970s, people consistently underestimated future inflation. That kept real interest rates artificially low and boosted growth in high-inflation countries. From 1974-1983, real long-term rates averaged 2.1% in low-inflation countries, 1.2% in medium-inflation countries, and were negative in high-inflation countries. Markets learned. From 1984 onward, high-inflation countries paid higher real rates, and the penalty for inflation became much larger.
The link with unemployment is clearer. During 1974-1991, low-inflation countries had the lowest jobless rates. Cutting inflation often caused a temporary rise in unemployment, but over the full period low inflation was associated with more jobs, not fewer.
More rigorous studies try to isolate the effect. Research by two Bank of Canada economists covering 62 countries over 25 years found that a 1-point reduction in inflation raises annual growth by 0.1 points. An OECD study found something similar. It sounds small, but it compounds. Cutting inflation from 5% to 0% would leave output about 10% higher after 20 years.
Is it worth it?
If the benefits of moving from 3% to 0% are at least as big as moving from 6% to 3%, then the question is whether those long-term gains outweigh the short-term costs. In theory, price stability maximizes growth in the long run. In practice, getting there is painful. To bring inflation down, unemployment usually has to rise above its natural rate for a while, and output has to run below potential.
A common estimate is that you need to sacrifice 1 point of growth for 1 year to cut inflation by 1 point. Using the Bank of Canada estimate, it would then take 10 years before the economy is better off than if nothing had been done. If the sacrifice is smaller, it might take 5 years. Over 20 years the economy clearly gains, but governments only last 4 or 5 years, and voters care more about jobs today than prosperity decades away. That is why some governments, including Britain at the time, began to accept a bit of inflation.
The transition is costly because expectations adjust slowly and because the economy is built around inflation. Workers expect annual raises in line with last year’s inflation. Homeowners take on big mortgages expecting inflation to erode the debt. If inflation suddenly falls, they are left with a heavier real burden. Likewise, real interest rates tend to rise and stay high until investors believe that low inflation will last.
A study by Stephen King at James Capel argued that zero inflation was not feasible for all countries at that time. The short-term costs would be too great and would force policymakers to reverse course. He said price stability would only work if private debt was sustainable and labor markets were flexible. Of the six big industrial economies, he judged that only Germany and Japan met those conditions. Britain looked least able to cope, because of rigid markets, high debt, and entrenched inflation expectations.
Yes, if governments cut the cost
The conclusion is not to abandon the fight against inflation, but to reduce the cost of fighting it. Tight money and tight budgets are not enough. Governments also have to change the habits and institutions that keep inflation alive. That means credible policy, so people believe inflation will stay low and don’t build high expectations into wages and prices. The key step is central bank independence, which removes the political temptation to stimulate the economy in the short term.
There is evidence this works. Greater independence for the central banks in New Zealand and Canada, combined with explicit inflation targets, helped bring inflation down to around 1% and hold down wage demands. Both countries went through deep recessions to get there. New Zealand then reaped the rewards, with output growing almost 4% in the following year and strong forecasts ahead.
Zero inflation is not a miracle cure. Its benefits are long term. Unless governments also attack supply-side rigidities, the goal will stay out of reach. But the sooner policymakers can convince households and firms that they are committed to price stability, the sooner the economic rewards will arrive.
