This report shows that extreme conditions and volatility in markets are much more likely to result from systematic policy errors in gauging and responding to inflationary pressures in an economy than from unfortunate random shocks.

We describe a simple model that incorporates the key features of the policy control process. We use two versions of the model to define two hypothetical economies, one where inflation responds linearly to the state of excess demand and one that introduces an asymmetry, with excess demand having faster and stronger effects on inflation than does excess supply. Using stochastic simulations of the two economies, we study the consequences of errors in the model that is used by the monetary authority in formulating policy to keep inflation close to a target level. For each economy, we consider two cases: one where the monetary authority knows the true structure; the other where it mistakenly assumes that the other version of the model describes the economy.

The results indicate that, when a monetary authority cannot know the true structure of the economy, it minimizes risks of cumulative errors and volatility in markets by assuming it faces the more difficult task of controlling inflation in a non-linear environment. If this assumption is wrong, there are costs—for example, output is slightly lower than it could have been, on average. However, the costs of incorrectly assuming linearity are much greater, because this error tends to permit outbursts of inflation, which are followed by relatively severe corrections.