Fiatmap

Unemployment vs Inflation: How the Two Really Relate

2026-10-19

Economic curve sketches frame article title

In the short run, unemployment and inflation usually move in opposite directions: when unemployment is low, inflation tends to run higher, and when unemployment is high, inflation tends to cool. That is the Phillips curve. The pattern is a tendency, not a law. It has weakened in the United States, it can break entirely when supply shocks hit, and it disappears in the long run. This guide explains what the relationship is, why it fails, and how to compare the two numbers across countries without fooling yourself.

Fiatmap
Compare Inflation Across Countries
Fiatmap brings official economic data from 128 countries together, helping you compare inflation and understand different macroeconomic conditions.
Explore economic data

TL;DR

Table of Contents

What each number measures

The unemployment rate is the share of the labor force that has no job and is actively looking for one. We explain the definition and its quirks in our glossary entry on the unemployment rate. The inflation rate is the yearly change in a price index, usually consumer prices. Both are headline numbers built from official surveys, and both hide detail.

As a current U.S. reference point, the Bureau of Labor Statistics put the unemployment rate at 4.2 percent in September 2026, with payroll employment changing little. On the price side, BLS reported that the Consumer Price Index rose 0.6 percent in September 2026 on a seasonally adjusted monthly basis, with gasoline a driver. A one-month price jump is not a trend, but it shows why you should never read one reading in isolation.

The Phillips curve: the classic trade-off

The curve that links the two is named after the economist A.W. Phillips, who studied U.K. unemployment and wage changes over a long historical span and found that wages rose faster when unemployment was low. In its traditional short-run form the curve slopes downward:

Labor market Typical wage pressure Typical inflation effect
Low unemployment, tight Workers have bargaining power, wages rise faster Inflation tends to be higher
High unemployment, slack Wage growth slows Inflation tends to be lower

The Philadelphia Fed notes that modern versions use an unemployment gap, meaning unemployment relative to a slow-moving trend, rather than simply the headline rate. That detail matters: the same 4.2 percent can be tight in one decade and loose in another, depending on what the economy’s sustainable level is.

The economic logic is intuitive. When firms struggle to hire, they pay more, and they pass costs on to customers. When jobs are scarce, workers accept smaller raises and spending softens.

How big is the effect?

Economists have tried to put a number on it, with heavy caveats. A Federal Reserve staff analysis using regional U.S. data found that, under a NAIRU framework, a 1-percentage-point rise in unemployment was associated with roughly a 0.3-point decline in inflation over the following year. The same paper warns that aggregate data alone may not reveal a stable relationship, because central banks react to the economy and blur the pattern.

So treat 0.3 as an illustration of the order of magnitude, tied to one dataset, one period and one model. It is not an exchange rate you can apply to another country or another decade.

Why the trade-off disappears in the long run

In the long run, people build expected inflation into wage deals and prices. If a central bank tries to hold unemployment below its natural rate by letting inflation run, workers and firms adjust their expectations, and the unemployment gain evaporates while the higher inflation stays. N. Gregory Mankiw’s NBER review of monetary policy describes the consensus pattern: a contractionary shock raises unemployment temporarily, and inflation then declines with a delay and gradually.

That is why central banks talk about expectations so much. A credible commitment to low inflation can keep the curve flat, because nobody expects prices to run away.

For roughly a decade after the Great Recession, the textbook link looked broken. St. Louis Fed economist Kevin Kliesen explains that unemployment fell to levels not seen in decades while inflation stayed low, and the Fed chair of the time described the relationship as a “faint heartbeat.” The practical lesson for investors: low unemployment does not automatically predict high inflation.

Three reasons are commonly given:

  1. Anchored expectations. When households and firms believe inflation will stay near target, tight labor markets feed through to prices more slowly.
  2. Policy reaction. Central banks that tighten as unemployment falls erase the very correlation researchers try to measure.
  3. Measurement. The headline rate can miss people who left the labor force or want more hours.

That third point is global. The ILO reported a 4.9 percent global unemployment rate in 2025, projected to stay there through 2027, alongside a projected jobs gap of 408 million in 2026. A calm headline number can sit on top of a lot of underutilization, and slower labor-force growth explains part of the apparent stability.

When supply shocks break the pattern

The Phillips curve describes demand-driven swings. Supply shocks work differently: they can raise prices and weaken activity at once.

The 2020 pandemic is the clearest case. U.S. unemployment was 7.9 percent in September 2020, 4.4 percentage points above September 2019, and every state plus Washington, D.C. had a higher rate than a year earlier. That was a shutdown shock, not an ordinary business cycle, and it should not be read as a standard point on the curve.

The inflation that followed had different roots. The IMF’s October 2024 outlook attributes the 2021 to 2022 global surge to pandemic supply disruptions, strong demand and commodity spikes after Russia’s invasion of Ukraine. Global headline inflation peaked at 9.4 percent year over year in the third quarter of 2022, and as supply constraints eased and policy tightened it came down without a major global recession.

The lesson: before you read inflation against unemployment, ask what is driving prices. Demand-driven inflation and supply-driven inflation call for different policy responses and carry different labor-market costs. That distinction is especially important for anyone trying to anticipate mortgage rates.

When supply shocks break the pattern — overview diagram

The post-pandemic twist: steeper, shifted, and measured differently

Research since 2020 has shown the curve can change shape. IMF researchers found that pandemic-recovery Phillips curves shifted and steepened across countries and sectors, which explains why old coefficients can fail after a structural break: the same unemployment gap can map to a different inflation outcome once supply is disrupted.

Another IMF paper by Ball, Leigh and Mishra argues that the vacancy-to-unemployment ratio is a better gauge of labor-market tightness than unemployment alone in the recent episode. They report that the U.S. ratio peaked at 2.0 in March 2022, against an average of 0.6 for 1985 to 2019, and stood at 1.1 in February 2025. Unemployment barely moved at the peak of tightness, but openings per jobseeker told the story.

U.S. vacancy-to-unemployment ratios over time

The New York Fed offers a related finding. Its model of the pandemic period found that underlying inflation fell from a 4 percent peak to 3.2 percent at the end of 2023 without a notable rise in unemployment, partly because expectations anticipated the unemployment gap closing. Disinflation without a job-loss cost is exactly what the old textbook rule said was hard.

Expectations are the missing variable

If expectations drive the curve, then watching them is as useful as watching unemployment. Surveys of households and professionals are one source. Another, for readers who follow forecasting markets, is the odds traded on economic outcomes; this overview of what prediction markets can show that traditional terminals miss is a useful starting point for understanding how those signals are collected. Treat them as one input among several, not as a forecast you can bank on.

How to compare unemployment and inflation across countries

If you are an analyst, expat or student building your own comparison, follow a short checklist.

  1. Match the definitions. Unemployment follows ILO-style definitions in most countries, but sampling, age limits and seasonal adjustment differ. Inflation may be CPI in one country and the harmonized index (see our HICP explainer) in another.
  2. Match the frequency. Do not set a monthly inflation print against a quarterly jobless figure without aligning them.
  3. Keep the lag. Unemployment often responds late to monetary tightening, and inflation declines gradually. We cover the delay in our piece on inflation and interest rates.
  4. Look past the headline. Add job openings, participation and expectations where official data exist.
  5. Separate the type of inflation. Producer prices often lead consumer prices, which we discuss in PPI vs CPI for analysts.
  6. Check the reference month. State and regional rankings change by month and by whether the series is seasonally adjusted. The BLS publishes separate adjusted and unadjusted state tables, and mixing them can change who ranks first.

For reproducible U.S. series, the St. Louis Fed’s FRED database provides downloadable unemployment, CPI and GDP data. For cross-country work, you need harmonized official statistics from one place.

What the pairing means for your money

The interplay matters well beyond economics class.

Our perspective

We build Fiatmap around official statistics from 128 countries, and the most common mistake we see in comparisons is treating unemployment and inflation as a pair of dials that move one another. They do not. They are two outputs of a larger system shaped by policy, expectations, supply chains and demographics.

When you lay the two maps side by side, you will find countries with low unemployment and low inflation, countries with both high, and countries with the opposite combinations. That scatter is the honest answer to the question in the keyword: there is a tendency in the short run, and a great deal of exceptions. Our advice is to start with the world unemployment map and the inflation ranking, then ask why a country sits where it does before drawing any conclusion.

We also believe a number is only as good as its source. That is why we use official statistics only, and why we show the definitions next to the data.

— Torstein

See both indicators side by side on Fiatmap

You can explore the unemployment rate by country and the inflation rate by country in free, sortable rankings, then compare them with GDP growth and mortgage rates. If you want the raw numbers for your own model, the Fiatmap Global Money Dataset is free, sourced and licensed under CC BY 4.0. Start from the Fiatmap home page to browse the full atlas.

FAQ

Does unemployment go up or down with inflation?

In the short run, the usual prediction is inverse: inflation tends to be higher when unemployment is lower, and lower when unemployment is higher. In the long run the trade-off fades as expectations adjust. Supply shocks can break the pattern, pushing both up at once.

Which curve shows the relationship between inflation and unemployment?

The Phillips curve. In its traditional form it slopes downward, linking low unemployment to higher inflation or wage growth. Modern versions use an unemployment gap rather than the raw rate and add inflation expectations.

What year was COVID unemployment?

The major shock came in 2020. U.S. unemployment was 7.9 percent in September 2020, 4.4 percentage points above a year earlier, and every state and D.C. had a higher rate. It was an unusual event and a poor guide to ordinary Phillips-curve behavior.

What state is number one in unemployment?

It depends on the month and on whether you use seasonally adjusted or unadjusted data. Check the latest BLS state table for the month you care about, and do not treat any ranking as permanent.

Sources

The figures on this site are collected from official statistics and each carries its own source and date — see /sources and the dataset. Where this article states a number, the country and ranking pages are what it should agree with.