Assessing the impact of incomes policy: the Italian experience

Purpose - The Saint Valentine's Decree (1984) and the ensuing hard‐fought referendum (1985), which reduced the automatisms ofscala mobile, started a process of redefinition of wage fixing in Italy, which culminated with the final abolition ofscala mobile(1992) and the approval ofProtocollo d'intesa(1993). Since then, following new corporatist principles, a national system of centralised wage bargaining (concertazione) and so‐called “institutional indexation” have governed the determination of wages. Does incomes policy generate greater coordination in the process of wage formation? Does it cause greater co‐movement of wages, prices, labour productivity and unemployment? This paper aims to answer these questions with reference to one of the G8 economies. Design/methodology/approach - After testing for unit root each component by using the ADF, Phillips and Perron, DF‐GLS and Zivot and Andrews statistics, the paper tests for co‐integration the so‐called WPYE model using different methods. The Engle and Granger approach is used to assess the impact of incomes policy on the speed of adjustment of real wages, productivity (and unemployment) to their equilibrium value, while the Gregory and Hansen procedure serves as a means to endogenously detect the presence of a regime shift. The paper estimates coefficients before and after the structural break. Findings - Incomes policy based on the 1993 Protocol has caused a regime shift in the process of wage determination. The long‐run estimates of the WPYE model do not generate stationary residuals except when a dummy for 1993 is added. The share of wages over GDP reduces by about ten percentage points in the early 1990s and has stood at about 57 per cent since 1995. The link with productivity is close to one‐to‐one only before the break. The feedback mechanism, as measured by the coefficient of lagged residuals in short‐run estimates, is increased from −0.46 in the pre‐reform to −0.79 in the post‐reform period, suggesting that incomes policy has increased real wage flexibility indeed. In recent years the link between real wages and (very low) labour productivity growth has weakened. In a sense, incomes policy has introduced a new form of (upward) wage rigidity. Last but not least, incomes policy has changed the correlation with the unemployment rate from positive to not statistically significant. Research limitations/implications - Future developments will focus on disentangling the impact of incomes policyvis‐à‐visother policy interventions on WPYE and on unemployment. Practical implications - The analysis calls for a careful revision of the 1993 Protocol aimed at better protecting the purchasing power of real wages without losing control on inflation, and introducing growth‐generating mechanisms. Originality/value - The paper studies the impact of incomes policy on WPYE and the Phillips curve by means of co‐integration and structural break analysis. It proposes to interpret the effect of incomes policy on the Phillips curve as changing the coefficient of the error correction mechanism that leads real wages to their long‐run equilibrium value.