On the Generality of the New Keynesian Phillips Curves

The New Keynesian Phillips curve is widely used in macroeconomics and monetary policy analysis. It is explicitly based on micro-foundations, monopolistically competitive firms and sticky prices. In its original form the New Keynesian Phillips curve is purely forward-looking model of inflation dynamics in the closed economy context. It is based on time-contingent price setting, which can be derived using Taylor’s overlapping contracts model (Taylor 1980), Rotemberg’s model of quadratic costs of price adjustment (Rotemberg 1982) or the Calvo (1983) model with random price adjustment. The alternative New Keynesian Hybrid Phillips curve includes elements of both forward- and backward-looking prices setting (Galí and Gertler, 1999). In the hybrid model only some price setters behave optimally when adjusting prices while the rest use rules of thumb or indexation, which is based on recent history of aggregate prices. When the New Keynesian Phillips curve is extended into open economy framework, inflation dynamics become more complicated, as new channels arise due to exchange rate changes and the effects of foreign shocks. In this approach, not only domestic demand and supply, but also foreign economic conditions influence domestic inflation. Imported goods can be modelled as intermediate goods (McCallun and Nelson, 1999, 2000; Kara and Nelson, 2003; Allsopp, Kara and Nelson, 2006), or as final consumption goods (Galí and Monacelli, 2005). Also more complicated models have been investigated (Batini et al, 2005; Leith and Malley, 2007; Rumler, 2007). The exchange rate pass-through is assumed to be full when final consumption goods model is analysed, but incomplete when imported goods are treated as intermediate goods. In this study we examine the empirical relevance of the New Keynesian Phillips curve relationship. Using pooled data for the euro area since the late 1980s, we compare the empirical fit of alternative Phillips curve specifications. We investigate both purely forward-looking models and hybrid models, which include both forwardand backward-looking elements of expectations. In the open economy context, we make the assumption that all imports are intermediate goods. Possible persistence in expectations is taken into account by using direct proxies i.e. Consensus Economics survey data for inflation expectations. The empirical analysis provides the strongest support for the open economy New Keynesian hybrid model. The Wald test of coefficient restrictions suggests that compared with the purely forward-looking specification, euro area inflation dynamics are better captured by the hybrid Phillips curve. Moreover, the empirical performance of the hybrid specification is improved, if the model is extended into open economy context. Robustness analysis indicates that the same open economy hybrid model is appropriate for countries with low and with high output gap volatility. Moreover, the inflation process in the four biggest and in the rest of the countries can be modelled using the same model parameters. Inflation dynamics are a central issue in monetary policy analysis. When conducting monetary policy, the inflation process and the effects of foreign shocks (for example energy and food price shocks) on domestic inflation must be carefully analysed. It is also important to examine how persistent the effects of shocks on inflation are and how the exchange rate and inflation are related. Overall, monetary policy analysis must be based on structural models, which capture expectations dynamics and the open economy aspects of the inflation process accurately. Recently, due to sharply weakening conditions in the world economy and highly volatile commodity prices, maintaining a deep understanding of inflation dynamics in the open economy context has become even more important for central banks.