Abstract:

I study the relationship of input-output linkages between firms and the conduct of monetary policy by modern central banks. First I examine the implications for monetary policy of having sticky prices in both final goods and intermediate input goods. I show that the dynamics of an economy with sticky prices in both sectors is substantially different from standard one-sector economies. I model monetary policy as "constrained discretion," the minimization under discretion of a simple loss function rather than the implementation of ad hoc simple instrument rules. I find that targeting inflation of both a consumer price index and an intermediate input price index performs significantly better than targeting a single price index, and has superior robustness properties. Price-level targeting of both consumer and intermediate goods prices performs even better.

I next examine the implications for optimal monetary policy under commitment when inflation persistence in intermediate goods prices differs from that in final goods prices. I incorporate inflation persistence into a New Keynesian input-output model by allowing for "near rational" behavior by a fraction of the firms in each sector due to information processing costs. I show that accounting for inflation persistence changes the optimal policy actions taken by the central bank. Additionally, I show that optimal policy actions depend crucially on the relative degrees of inflation persistence in the two sectors. Finally, I show that, unlike one-sector models, price-level targeting outperform inflations targeting unless inflation persistence is high in both sectors.

Finally, I examine inflation persistence for price indices at different stages of production. I find that in general inflation persistence is higher for goods that are more highly processed. However, this result must be modified when relative degrees of persistence are examined in different monetary policy regimes. I split the post-WWII period into a pre-Volcker era (until 1979) and a post-Volcker era (after 1982). I find that inflation persistence in crude materials and intermediate inputs is unchanged across these regimes, while CPI inflation is less persistent in the more recent era. Furthermore, in the post-Volcker era CPI inflation persistence is essentially equal to that of intermediate inputs inflation.