The expectations-augmented Phillips curve (Friedman 1968, Phelps 1967) argues that the long-run Phillips curve is VERTICAL at the natural rate of unemployment U*, so any level of inflation is compatible with U = U*. On this view, demand-management policy aimed at lowering U below U* yields only accelerating inflation, not sustained employment gains. The claim has strong theoretical and empirical support but also significant qualifications.
The case FOR the claim — the expectations-augmented argument. The short-run Phillips curve (SRPC) is drawn for a given level of expected inflation πᵉ. If demand-management pushes U below U*, actual inflation exceeds πᵉ. Workers briefly accept lower real wages, firms hire more, and U falls — moving along SRPC₁ from point A to a boom point B. But workers eventually realise that actual inflation exceeds expected, they demand compensating nominal wage rises, real wages return to equilibrium, and the SRPC shifts up to SRPC₂. The economy returns to U = U* at point C — same unemployment, higher inflation. If the government persistently tries to run below U*, each round of stimulus delivers only a short-term employment gain and a permanent ratchet up in inflation. The long-run locus of equilibrium points traces out the vertical LRPC at U*. There is no long-run trade-off.
Empirical support. The 1970s stagflation broke the apparent stable trade-off: inflation and unemployment rose together in a way the traditional Phillips curve could not explain. The US and UK experience since 1980 shows episodes of low unemployment combined with low inflation (late 1990s, mid-2010s) and inflationary surges that unwound. The accelerationist prediction — that holding U below U* yields accelerating inflation — matches the stagflation experience.
Rational expectations — a stronger version. Lucas and Sargent argued that if agents form rational expectations, they anticipate the stimulus and adjust nominal wages and prices IMMEDIATELY. The short-run trade-off then disappears for ANTICIPATED policy; only UNEXPECTED policy can shift U below U*. Even that gain is brief. The strict version of the no-long-run-trade-off claim then becomes a no-short-run-trade-off claim for predictable policy.
The case AGAINST — hysteresis. Blanchard and Summers argued that long-term unemployment damages skills and makes the long-term unemployed harder to re-employ. Prolonged slack therefore RAISES U* itself. This hysteresis effect works in favour of some demand stimulus to prevent recessions from permanently raising U*, and against the strict view that U* is fixed. Hysteresis does NOT restore the exploitable long-run trade-off (there is still no 'menu'), but it means that passive acceptance of high U is costly in the long run.
Other challenges. Downward nominal wage rigidity, menu costs and sticky expectations can make the short-run Phillips curve persist longer than Friedman implied — the trade-off is not instant, so even transitory demand-management may deliver real benefits during slow adjustment. Behavioural economics suggests expectations are not always rational. And in a crisis (demand collapse, zero lower bound), fiscal and monetary stimulus can prevent the economy from falling deep below U* for prolonged periods — again not an argument for running below U*, but for stabilising around it.
Inter-relatedness with other macroeconomic problems. Even if a short-run trade-off exists, exploiting it generates OTHER costs through the inter-related links of Topic 10.2. Higher inflation erodes the internal value of money, pulls the external value down (currency depreciation), worsens the current account via loss of competitiveness, and compounds with demand-led growth to accelerate the import bill further. The real cost of 'modestly higher inflation' is therefore spread across all four macroeconomic problems, which argues further against exploiting any apparent trade-off.
Policy implications. First, modern central banks target inflation (an inflation anchor) and allow U to settle at U* in the long run — the strict Friedman-Phelps implication. Second, lowering U* itself requires SUPPLY-SIDE policies — training, education, flexible labour markets, incentives — not demand-management. Third, demand-management has a legitimate role: to STABILISE output and employment AROUND U* and around the inflation target. The claim should therefore not be read as 'demand-management is useless' but as 'demand-management cannot sustainably push below U*'.
Reasoned judgement. The view that there is no long-run trade-off is CORRECT in its essentials: the LRPC is vertical, the stagflation evidence confirms it, and attempts to run below U* generate only accelerating inflation. BUT the policy conclusion 'no demand-management' is too strong. Demand-management is still needed to stabilise the economy around U*, to prevent recessions from raising U* via hysteresis, and to deal with crises. The right formulation is: demand-management should be used to stabilise, inflation targets should anchor expectations, and supply-side policy should lower U* itself. The no-long-run-trade-off view is a critique of ATTEMPTS TO RUN BELOW U*, not of demand-management per se. On that qualified reading, the claim is correct and remains the foundation of modern macroeconomic policy.