On the relevance of equilibrium unemployment and aggregate demand – some thoughts Steinar Holden Department of Economics http://folk.uio.no/sholden/ Summer course, JVI, Vienna, 2005

Outline The macroeconomics of equilibrium unemployment theory Norway from 1980 Changes in eq. unemployment, or Persistent effect of aggregate demand? Equilibrating mechanisms Experiences from other selected countries Reflections and thoughts

Equilibrium unemployment Price curve Un U Real wage Wage curve Equilibrium unemployment Price curve Shocks may take U away from Un. If U < Un, increased wage and price growth will push U towards Un.

Persistent fluctuations in unemployment Under equilibrium unemployment theory, fluctuations are caused by expectational errors and nominal rigidities Unlikely that these will persist for many years =>Persistent fluctuations usually viewed as due to changes in Un But is this always the correct interpretation? Other sources of fluctuations are plausible Equilibrating mechanisms are weak

Norway, from 1980 to 2004 Liberalization of credit markets and expansionary fiscal policy lead to boom in 1984-85. Huge imbalance was created, with over-investment and high household debt Then the economy turned, investment fell sharply Households had to reduce consumption oil prices fell and fiscal policy had to be tightened The start of a long-lasting downturn

Persistent fluctuations in investment, saving and unemployment, Norway

Unemployment & investment/GDP, Norway, deviation from mean

Persistent increase in unemployment Sluggish demand or increased equilibrium unemployment? Increased eq unemployment consistent with OECD’s NAWRU-estimate (Non-Accelerating Wage Rate of Unemployment) Constructed to be the rate of unemployment for which wage growth is constant But can this explain what happened?

Unemployment (U) and NAWRU-estimates for the Nordic countries (from Holden og Nymoen, SJE, 2002)

Difference between actual and eq Difference between actual and eq. unempl (U – NAWRU) (horisontally) as predictor of the change in wage growth (vertically) (Holden og Nymoen, SJE, 2002)

Changes in equilibrium unemployment unlikely explanation NAWRU-estimates poor predictive power At the same time no changes in labour market institutions that could explain the increase in equilibrium unemployment More complete wage equation indicated stable wage setting (Holden&Nymoen)

Alternative explanation – persistent deviation from equilibrium unemployment Fluctuations caused by real shocks, related to consumption, investment and fiscal policy The equilibrating mechanisms are weak, and cannot be expected to restore equilibrium quickly

How powerful are the equilibrating mechanisms ? (I) Forward-looking elements Consumption smoothing – but what do households know about the future? If income falls, what happens to consumption? transitory shock ? => C falls less persistent shock ? => C falls in proportion persistent shock to the growth rate ? => C falls more Likewise for investment – firms may postpone investments if significant risk that demand remains sluggish If large shocks affects the expected future path, the effect on consumption and investment can be large

How powerful are the equilibrating mechanisms ? (II) Wage and price growth Pigou-effect is negligible Usually strong effect of tight labour market Dampening effect of slack labour market is often weaker Lower (zero) bound to nominal wage growth Euro-zone: rigidities + country-spec shock Weak expansionary effect of lower wage and price growth Income distribution effect opposite direction

How powerful are the equilibrating mechanisms ? (III) International competitiveness Important, but traded sector only part of overall economy Low relative costs may be viewed as temporary Available resources Do high unemployment really stimulate investment and job creation? Monetary policy potentially important equilibrating mechanism But Norway had fixed exchange rate target at the time

Equilibrating mechanisms – summing up Equilibrating mechanisms in the economy are not strong probably stronger down (if activity too high) than up (if activity is too low) Negative, large or persistent shocks may have persistent effects on unemployment Most fluctuations probably due to transitory shocks => economy will return when shock vanishes Investment ↓=> GDP↓ ;Investment ↑=> GDP ↑ But some shocks are permanent Large shocks may affect expected future path of the economy

Is Norway unique ? No, in fact very similar development in other Nordic countries, and also other European countries Finland and Sweden experienced very similar consumption and investment fluctuations as Norway But also important differences Norway had oil money, and could pursue expansionary fiscal policy without concern for high future public debt In Finland, the downturn was strengthened by the breakdown of the Soviet Union, which was an important export market

Persistent effect of downturn in Finland in the early 1990s

Unemployment & investment/GDP, Finland, deviation from mean

Persistent effect of downturn in Sweden, too

Unemployment & investment/GDP, Sweden, deviation from mean

And the UK

Unemployment & investment/GDP, the UK, deviation from mean

And in Euro 12

Unemployment & investment/GDP, Euro area, deviation from mean

Long-lasting effect on employment rates (15-64): Finland and Sweden

And in Norway, the UK and the Euro Area

Tentative conclusions (in the wait for better evidence) Long-lasting fluctuations in investment, saving and unemployment Clear indication of a role for aggregate demand Changes in structural variables in the labour market (wage setting, labour market institutions, benefits, etc) cannot by themselves explain these correlated fluctuations Explanations based on the combination of shocks and institutions more promising But these explanations allow role of demand