by Charemza, Wojciech
and Makarova, Svetlana
Published in Romanian Journal of Economic Forecasting, 2009, volume 10 issue 2, 5-22
| Requires a PDF viewer such as Xpdf
The paper focuses on the decomposition of inflation persistence into the linear and nonlinear components. The hypothesis is that the nonlinear component of inflation persistence results from a technological shock and might positively contribute to economic growth. The microfoundations are derived from an assumption of Calvo pricing and sticky-information Keynesian Phillips curve. The hypothesis is evaluated empirically with the use of monthly data series for inflation for 119 countries. Linear and nonlinear (bilinear) inflation persistence measures have been estimated with the use of a bilinear autoregressive moving average model in a state space form. Further on, correlation analysis has been performed for these countries to detect a relationship between economic growth and linear and nonlinear persistence. The paper concludes that the nonlinear inflation persistence contributes positively to economic growth after 2000.
inflation, growth, bilinear processes