Post-war macroeconomics: Part 1
For many advanced economies, the early post-war period was characterised by full employment, modest inflation and economic prosperity. Under the guidance of the Keynesian model, fiscal and monetary policy was used for economic stabilisation and to achieve low levels of unemployment on a sustained basis.1
During this time, newly available macroeconomic data and the development of applied econometric models, largely due to the work of Jan Tinbergen and Lawrence Klein, encouraged empirical analysis. Klein (1947) argued that the micro-foundations of Keynes’s aggregate relationships (e.g. the consumption function) should be established to underpin the large scale macro-models which were under construction. The search for micro-foundations combined with the Hicks-Hansen IS-LM representation of The General Theory defined the neoclassical synthesis which dominated post-war macroeconomic thought.2
Extensive empirical analysis of inflation and unemployment dynamics ensued. In particular, Phillips (1958) observed an inverse relationship between money wage growth and unemployment in the UK, known as the Phillips curve.3 Samuelson and Solow’s (1960) analysis of the Phillips curve as an exploitable trade-off between inflation and unemployment gave it credence for policymaking.4 During the 1960s it was believed that a stable trade-off existed, so policymakers had a clear set of choices. Unemployment could be reduced via expansionary Keynesian demand-side policy but would likely result in higher inflation. Contractionary fiscal policy was seen as a solution to a supply-side inflation shock since it could move the economy along the Phillips curve. The Keynesian model provided all the necessary levers for economic stabilisation.
Friedman (1948), however, was critical of the short-run nature of Keynesian analysis which he argued eschewed the long-run objectives of economic stabilisation policy. Friedman (1948:263) began to outline a ‘stable framework of fiscal and monetary action, [which] largely eliminates the uncertainty and undesirable political implications of discretionary action by governmental authorities’. Here, Friedman was critical of the long and variable lag time associated with fiscal policy in particular but of discretionary policy in general. Friedman’s framework only promised reasonable full employment and a reasonable degree of stability.
While fiscal policy was generally emphasised over monetary policy by post-war (neo-) Keynesians, Friedman’s (1956) restatement of the classical quantity theory of money challenged Keynesian policy prescriptions and re-asserted the role of money and monetary policy (i.e. Monetarism). While distinct from Fisher’s (1911) earlier transaction version, Friedman argued that controlling the money stock was the most effective means of controlling inflation, and rejected any role for macroeconomic policy in stabilising real variables, such as output and employment in the long-run. Tobin (1972) recalls that Friedman had set out his theoretical framework using the widely accepted Hicks-Hansen IS-LM model. In doing so, debate regarding the role of fiscal and monetary policy was largely reduced to an econometric debate about empirical magnitudes (see Friedman and Schwartz 1963, 1970, 1982).
In the 1970s, global supply shocks, notably the 1973 oil crisis and the 1979 energy crisis, resulted in many economies experiencing stagflation, characterised by high unemployment and high inflation, following contractionary macroeconomic policies designed to reduce inflation. Stagflation was inconsistent with the Phillips curve trade-off which was already under attack by economists, particularly Edmund Phelps and Milton Friedman.
Phelps (1967, 1968) and Friedman’s (1968, 1977) hypothesis distinguished between nominal and real wages, and the short- and long-run outcomes of an unanticipated change in nominal demand. The latter would lead to conflicting perceptions among employers and employees regarding real wage adjustments which would permit a temporary deviation in the unemployment rate from its so called natural rate. Once inflation expectations are fully incorporated, however, the initial employment effects disappear as the unemployment rate returns to its natural rate. Expectations were formed adaptively, a notion which had been revived by Friedman’s (1957) work on the consumption function.
The key implications of the Phelps-Friedman natural rate hypothesis were that unanticipated inflation, not inflation per se matters; there is no permanent or stable trade-off between inflation and unemployment; and, unemployment can be kept below the natural rate only by accelerating inflation or above it only by accelerating deflation (Friedman 1977). Hence, the long-run expectations-augmented Phillips curve was vertical and unemployment would tend towards its natural rate. The latter is consistent with real forces and accurate perceptions. Lowering the natural rate would require policies which increased the competitiveness and flexibility of labour markets, such as reducing the power of trade unions, enhancing labour mobility, and minimum wage reforms. Keynesian demand-side policy could neither temporarily nor permanently reduce the natural rate.
The natural rate of unemployment (NRU) was later replaced by the non-accelerating inflation rate of unemployment (NIRU/NAIRU, see Modigliani and Papademos 1975). Full employment is now largely associated with the unemployment rate consistent with the NAIRU. But, unlike the Beveridge (1944) full employment definition of the early post-war period, the NAIRU does not imply equality between the number of job vacancies and the number of unemployed persons.
1 The Beveridge Report (1942) in the UK, Australia’s White Paper on Full Employment (1945) and the US Employment Act (1946) highlighted the strength of Keynesian principles in influencing policymakers. Policymakers also benefited from Lerner’s (1943) work on functional finance.
2 Concerns raised by members of the Cambridge Circus (e.g. Kahn, Robinson, Sraffa) about the use of aggregate production functions in the modelling of growth and income distribution were ignored with the development of large-scale, though single-sector IS-LM models.
3 Another important empirical relationship was Okun’s (1962) ‘law’ which reports an inverse relationship between unemployment and output. Ball et al. (2013) suggests that the ‘law’ continues to offer a strong and stable heuristic for most countries.
4 The empirical work of Phillips, which focused on the long-run relationship between unemployment and inflation, was inappropriately interpreted as an expression of a short-run policy trade-off to compensate for the obliteration of Keynes’ own approach to price and wage dynamics. The latter was set out in Keynes’ Z and D curve analysis in The General Theory, but had been obscured by debates over the nature of respective elasticities in the IS-LM model.