Why Is Inflation Global?

Explaining a strange co-occurence

Inflation commonly comes in global waves, which is extremely strange. Inflation is, after all, simply the change in the price level of a given currency. Why should it be common across countries? Yet, during the 1970s and early 2020s, most developed countries saw similar bouts of inflation, and inflation somewhere is substantially correlated with inflation everywhere. What explains this?

I should make sure that we’re all on the same page with inflation. We start with a very basic equation, MV = PY, where M is money, V is velocity, P is price, and Y is output. The “velocity” of money is simply how many times a given dollar is spent – if someone gives a dollar bill to a butcher in exchange for meat, who then gives it to a grocer in exchange for cigarettes, who then pays their employee, it’s as if there were four dollar bills in circulation, each of which was used once. Inflation is simply the change in price – strictly speaking, the first derivative.

We can therefore see that there are two basic ways in which we can have excess inflation. Either something on the left side of the equation, M or V, can increase, or output, Y, can fall. The former is referred to as demand-pull inflation, because it is the result of excessive demand (money) to buy products with. The latter is referred to as cost-push inflation, and is the result of insufficient supply. If the inflation was due to a shock to a commonly used good, like oil, then we can easily explain the co-occurrence of inflation.

But, what is striking is that the excess inflation of the 70s and early 2020s was not entirely due to supply shocks. We can tell this by looking at the growth rate of nominal GDP, as compared to real GDP. Real GDP is the real output of the economy, while nominal GDP is equivalent to the dollar values of the final goods. Suppose that the central bank wants NGDP (or, the left side of the equation) to grow by 4%, and expects the real economy to grow by 2%. The remaining 2% must be made up by a change in the price level. If real output growth is unexpectedly 0%, then inflation would be 4%. If inflation is higher than the difference between the implicit NGDP target and real GDP, it must be due to excess money printing, which is peculiar to each country. Since this was the case, we need a deeper explanation for why inflation is global.

“Global Price Shocks and International Monetary Coordination”, by Guerrieri, Lorenzoni, and Werning (2025) have a really interesting answer why. There is an economy with two inputs, one local to the country (labor) and one shared across all countries, which we can think of as oil. Labor can’t move, but oil can. There are also nominal wage rigidities, which is necessary to make inflation matter. (If wages and prices could perfectly adjust, then the price level could be anything, and it wouldn’t matter in the slightest). This gives rise to a Phillips Curve, or a tradeoff between inflation and employment/output. If prices grow by less than expected, people are left unutilized because the real value of their wages goes up. If prices are higher than expected, then output increases because more people are employed.

Given this, they analyze two scenarios. In the first, monetary policy is uncoordinated. Each country takes the price of oil as given, and sets their optimal monetary policy. In the other, we set the optimal monetary policy for the entire world. The key difference is that their policies affect other countries through the price of oil. Higher output increases demand for oil, which increases the price, which affects other countries. Each individual country faces less of a trade-off than the world does. We can think of this intuitively as countries facing a less steep Phillips curve than the world does – it takes less of an increase in inflation to generate the same rise in output.

Their results diverge from prior literature, which finds that uncoordinated policy actually has a contractionary bias, because they explicitly model that monetary policy has an effect on the output of the input good “oil”. The prior literature argued that each country has a unilateral incentive to contract, because it makes their currency relatively more valuable and thus makes imports cheaper. Because it’s only the relative strength of the currencies which matters, though, it’s a zero-sum game. Everyone doing this would result in excessive contraction. When you make the price of oil endogenous, however, you create multiple channels for the central bank to affect other countries, and under plausible parameter values, it is the effect on the price of oil which prevails.

Is this the best explanation for why inflation was higher, though? An alternative explanation is that supply shocks can lead to a change in inflation expectations. Suppose we’re going along at a low rate of inflation. A supply shock happens – prices rise by more than expected. The central bank is not credible, and unions demand large wage increase to keep pace with the change in prices. The central bank knows that if it is to try and change inflation expectations back, they must cause an increase in unemployment. Being unwilling to do this, inflation expectations shift permanently onto a higher course.

To make this model stronger, suppose that there are two unions bargaining over wages. If the central bank is targeting inflation, then the wages that any union sets affects the other union through the effect on the aggregate price level. This is analogous to the coordination problem that countries face. This is not something I am certain of. It is, nevertheless, something I am curious about, and I would like your comments on whether it works.