New IMFS Working Paper by Gerhard Illing on "Economic Theory and Central Bank Independence"
During the last decades, central bank independence proved to be the bedrock of credible and effective monetary policy. Both past and recent history provides plenty of evidence of disastrous outcomes with high inflation and economic stagnation, when central banks – lacking independence – were put under strong political influence. Game theoretic modeling provides valuable insights into the crucial role of independence. In a fiat money economy (with costs for producing additional pieces of paper money being close to zero) a commitment for price stability is key. In the absence of binding rules, it would not be dynamically consistent to stick to the promise to implement price stability. The model by Barro Gordon Model (1982) is the work horse model for game theoretic analysis. It builds on the seminal work of Kydland Prescott (1977) on dynamic inconsistency. The starting point of Barro Gordon is a short run Phillips curve, causing a trade-off between stabilizing inflation and stabilizing output. If there are structural distortions in the economy, the promise to stick to some target rate * not credible: There is an incentive to stimulate economy by raising above target rate of inflation * to raise output. Private agents anticipate this incentive; they raise expected inflation e ex ante. So the optimal policy is not dynamic consistent (in game theoretic terms: not subgame perfect). This leads to an inflation bias: Inflation being too high ( >*), without any gain in output and employment. So ex post, discretionary policy ends up in an inferior outcome.