The short-run trade-off between inflation and unemployment, why the curve shifts when expected inflation changes, the vertical long-run Phillips curve at the natural rate (NAIRU), accelerating inflation, and what it means for demand and supply-side policy. Through UK data since 2000.
In 1958 A. W. Phillips, a New Zealander at the LSE, plotted a century of British wage inflation against unemployment and found a smooth downward curve: when unemployment was low, wages rose fast. Governments drew the obvious conclusion: they could buy lower unemployment with a little more inflation.
The UK since 2000 is a test. From 2008 to 2013 unemployment rose above 8% and inflation stayed around 3 to 5%. In 2022 unemployment was under 4%, the lowest since the 1970s, and inflation reached 11.1%. By September 2026 inflation was 3.5% with unemployment at 4.7%: both above the levels of 2019, when they were 1.8% and 3.8%.
Is there a trade-off, or not? The answer is: in the short run, and only until people catch on.
Source: Office for National Statistics, consumer price inflation (D7G7), 20 October 2026; labour market overview (MGSX), 19 October 2026.
The short-run Phillips curve (SRPC) shows an inverse relationship between inflation and unemployment for a given rate of expected inflation. Lower unemployment means a tighter labour market, faster wage growth and higher inflation: a movement along the curve.
The SRPC shifts when expected inflation changes. If workers and firms come to expect 4% inflation rather than 2%, they build 4% into wage claims and price lists, so every level of unemployment now comes with 2 percentage points more inflation.
The long-run Phillips curve (LRPC) is vertical at the natural rate of unemployment, also called the NAIRU (non-accelerating inflation rate of unemployment). In the long run, when expectations have caught up with actual inflation, there is no trade-off: unemployment returns to the natural rate whatever the inflation rate.
This is the expectations-augmented Phillips curve of Friedman and Phelps (1968). The natural rate is set by supply-side factors: skills, mobility, benefits, trade union power, minimum wages, hiring costs.
In October 2022 UK inflation was 11.1% and unemployment 3.9%. In September 2011 inflation was 5.2% and unemployment 8.4%.
A famous idea in economics says that low unemployment comes with high inflation, and high unemployment with low inflation: a trade-off.
Do the two years fit?
2022 looks like the trade-off: a tight labour market and high inflation. 2011 breaks it: inflation and unemployment both high. The reason is expectations and supply shocks, and it is why the original Phillips curve had to be rebuilt.
Start with the economy at **A**: unemployment 5%, inflation 2%, and people expect 2%. Along **SRPC1** lower unemployment comes with higher inflation and higher unemployment with lower inflation. The curve is steep at low unemployment: when almost everyone who wants a job has one, a little more demand mostly raises wages and prices.
Check: Moving down a short-run Phillips curve from left to right, as unemployment rises:
Along a given SRPC expectations are fixed; higher unemployment means weaker bargaining power and less demand-pull, so inflation falls.
The government wants unemployment lower and boosts demand: lower interest rates, a bigger deficit. The economy moves along SRPC1 to **B**: unemployment 3.3%, inflation 4%. Firms raise prices faster than wages, so real wages fall and hiring more pays. Workers accept the jobs because they still expect 2% inflation: they have been fooled by money illusion.
Check: The government boosts aggregate demand and unemployment falls from 5% to 3.3%. On SRPC1 inflation rises to 4%. Why does employment rise at all?
Firms raise prices faster than wages, so real wages fall and hiring more workers is profitable. Workers, still expecting 2% inflation, think the wage offers are better than they are.
The illusion fades. Workers see 4% inflation, demand 4% rises, and real wages return to where they were. Firms shed the extra workers. Expected inflation is now 4%, so the curve shifts up to **SRPC2** and the economy is at **C**: unemployment back at 5%, inflation stuck at 4%. Join A and C and you have the vertical long-run Phillips curve.
Check: Workers and firms now expect 4% inflation. What happens to the short-run Phillips curve?
Expected inflation is built into wage claims and price lists, so the whole curve moves up by the rise in expectations.
Suppose the government refuses to accept 5% and boosts demand again. Along SRPC2 to **D**: 3.3% unemployment, 6% inflation. Expectations rise to 6%, the curve shifts to SRPC3, the economy ends at **E**. Each attempt buys a temporary fall in unemployment at the price of permanently higher inflation: accelerating inflation. The 1970s, roughly.
Check: If the government keeps unemployment at 3.3% year after year, the expectations-augmented Phillips curve predicts:
Holding unemployment below the natural rate requires inflation to stay ahead of expectations, and expectations keep rising. That is why the natural rate is also called the non-accelerating inflation rate of unemployment.
The lesson for policy. Demand management cannot hold unemployment below the natural rate for long. To lower unemployment permanently you must lower the natural rate itself: supply-side policies such as training, better job matching, reforms to benefits and taxes, more flexible labour markets. That shifts the LRPC left. Meanwhile monetary policy’s job is to keep expected inflation anchored at the target.
Check: Which of these would shift the long-run Phillips curve to the left?
The LRPC sits at the natural rate, which is set by supply-side factors. Only policies that change frictional or structural unemployment move it.
Back to the opener. 2011 was a shift, not a movement: oil and food prices, a weaker pound and a VAT rise pushed the SRPC up, so inflation was high even with 8.4% unemployed. 2022 was both: a tight labour market and a huge supply shock from energy. The simple trade-off holds only while the curve stays put, which is while expectations and costs stay put.
Check: In 2011 UK inflation was 5.2% with unemployment at 8.4%. Which best explains this on the Phillips curve diagram?
Cost-push shocks raise inflation at every unemployment rate, shifting the SRPC up. The economy was far to the right on a curve that had itself moved.
SRPC2 is drawn below SRPC1. Higher expected inflation adds to actual inflation at every level of unemployment, so the new curve must be above the old one, crossing the LRPC at 4%.
Using a Phillips curve diagram, explain why a government cannot reduce unemployment below its natural rate in the long run by expanding aggregate demand, and evaluate the implications for policy.
The short-run Phillips curve shows an inverse relationship between inflation and unemployment for a given rate of expected inflation: lower unemployment means a tighter labour market, faster wage growth and higher inflation. The long-run Phillips curve is vertical at the natural rate of unemployment, the NAIRU, where inflation equals expected inflation and the labour market is in equilibrium.
Suppose the economy starts at A on SRPC1, with unemployment at the natural rate and inflation at the expected 2%. The government expands aggregate demand. Prices rise faster than wages, so real wages fall and firms hire more; workers accept because they still expect 2% inflation. The economy moves along SRPC1 to B, with lower unemployment and inflation of, say, 4%.
Workers then notice that inflation is 4%, build it into wage claims and restore real wages. Firms shed the extra workers. Expected inflation is now 4%, so the short-run curve shifts up to SRPC2, and the economy settles at C: unemployment back at the natural rate, inflation permanently higher. Repeating the policy produces D and E, with inflation accelerating each time, which is why the natural rate is called the non-accelerating inflation rate of unemployment.
The implication is that demand management can lower unemployment only temporarily. Lasting reductions require supply-side policies that lower the natural rate itself, such as training, better job matching and labour market reform, which shift the LRPC left. Monetary policy’s task is to keep expected inflation anchored, which is why the Bank of England has an independent 2% target.
**However,** the conclusion depends on how expectations form. With adaptive expectations the short-run gain may last long enough to matter politically; with rational expectations it may not appear at all. The natural rate cannot be observed, and estimates for the UK have varied between about 4% and 6%, so a central bank that misjudges it may keep unemployment needlessly high. And the short-run curve shifts for reasons other than policy: the supply shocks of 2011 and 2022 raised inflation while unemployment was high or falling. The Phillips curve is best read as a warning against using demand to chase an unemployment target, not as a precise map.
CPI inflation (D7G7, released 20 October 2026) and unemployment rate (MGSX, released 19 October 2026): Office for National Statistics. The curves on the diagram are for an invented economy, chosen so the numbers are clear. Diagrams drawn with Rough.js and perfect-freehand, both MIT licensed.