Barrington Lecture 2007/08: Wage Inflation and Structural Unemployment in Ireland

What the paper says

1. INTRODUCTION The notion of a structural rate of unemployment is central to dominant economic theories explaining labour market behaviour and the relationship between unemployment and inflation. Structural unemployment estimates for a small open economy engaged in a common currency and having undergone significant buoyant economic growth, are particularly problematic and controversial. At the same time, foreign prices due to our open economy position play the major role in the measurement of our price inflation. We will assume that domestically-generated inflationary pressures for Ireland show up mainly in wages. Wages, representing the majority share of production costs, are widely regarded as the principal factor determining the competitiveness and export share of a small open economy. (2) It is timely to study the effects and determinants of wage inflation in Ireland and examine specifically its response to aggregate demand pressures--measured by the slack in the labour market or the gap between observed unemployment and the structural rate of unemployment over the period 1980 to 2005. (3) Central Banks, in particular, find it helpful to make empirical estimates of the structural rate of unemployment measured by the non-accelerating inflation rate of unemployment (the NAIRU). It serves as an input into various assessments of inflationary pressures, fiscal positions and structural issues. This paper employs a Kalman filter estimate of unobserved states to measure a quarterly NAWRU--the rate of unemployment at which wage inflation is stable. The NAWRU indicator is based on the idea that actual unemployment below structural unemployment is associated with inflation above its expected value. This method, similar to that employed by Richardson et al. (2000) and Laubach (2001) is used to identify how the Irish economy since 1980 has successfully achieved a reduction in the NAWRU. The method has the advantage of considering the unobserved structural rate of unemployment as a time-varying variable. Methodological extensions to previous studies include: inserting productivity as an cointegrating variable with the unemployment gap for Ireland, testing a Stock and Watson (1998) procedure to estimate the signal-to-noise ratio instead of taking standardised, rule-of-thumb measures; as well as making use of Irish quarterly data up to the end of 2005. The novel approach of applying the Kalman filter technique with an explicit exogenous variable, such as aggregate productivity, allows us to quantify how changes in this variable change with the unemployment gap. The significant presence of total factor productivity (TFP) gains implies that a buoyant economy that causes unemployment to remain low for a prolonged period is thus associated with a decline in the unobserved structural rate of unemployment. As unemployment falls, a shift in the composition of employment associated with higher effective demand (narrowing labour market slack) is significantly aided by aggregate productivity. The improvement in the skill set of the labour force and more effective use of capital changes the structure of the labour market such that there could be an actual reduction in the inflationary pressures associated with any given level of unemployment in the long run. The paper is structured as follows: the next section outlines indicators of wage inflation and unemployment since 1980 which suggest that a priori that labour market structural conditions have changed over the period implying that the NAWRU could not have been constant for the duration to 2005. Section 3 outlines alternative approaches to estimating a structural rate of unemployment. We conclude that a semi-structural approach modelled as an unobserved components model with a Kalman filter, allowing for productivity movements should generate a good representation of the time-varying NAIRU in an Irish context. …