This is a review article of two books: Financial Crises and What to Do about Them by Barry Eichengreen (Oxford University Press 2002) and Financial Crises, Liquidity, and the International Monetary System by Jean Tirole (Princeton University Press 2002).

The paper shows first that real exchange rate appreciation typically reduces, but may increase, the output costs of disinflation. The direct favorable effects of cheaper imports on consumer prices, on the prices of imported inputs, on wage demands, and on the demand for money may be outweighed by the unemployment caused by the trade deficit. Second, the effects of wage indexation on the costs of disinflation are reexamined. Ex ante indexation speeds up disinflation. Ex post indexation, the type observed in practice, automatically increases the real wage at the start of a disinflation and may therefore increase the sacrifice ratio.

[It is well known that in the Sidrauski monetary intertemporal optimizing model the steady state capital stock is invariant to the rate of inflation. This paper shows that the rate of accumulation of capital is not invariant to the rate of inflation and that, for the constant relative risk aversion family of utility functions (except logarithmic), the rate of capital accumulation is faster the higher the growth rate of money. Money is thus not neutral on transition paths.]

[The relationship between the contingent commodity and asset approaches to consumer behavior under uncertainty is used in examining Slutsky equations for assets. The menu of assets describes an "attainable set" of contingent commodities; it is shown that a number of types of changes in the distributions of returns on assets leave this attainable set unaltered and that the effects of such changes in distributions on asset demands are simply related to the effects of a change in the asset price on demands.]

Much has happened in the world of central banking in the past decade. In this paper, I focus on three issues associated with the zero lower bound (ZLB) on short-term nominal interest rates and the nexus between monetary policy and financial stability: 1) whether we are moving toward a permanently lower long-run equilibrium real interest rate; 2) what steps can be taken to mitigate the constraints imposed by the ZLB; and 3) whether and how financial stability considerations should be incorporated in the conduct of monetary policy. These important topics deserve the attention of both academic and government professionals.

I stand here with deeply conflicting emotions. I am honored to be delivering this prestigious lecture. I am profoundly sad that Rudi Dornbusch, who should have delivered the Ely Lecture, died in July last year and that I am here in his place. So I would like to start by talking about Rudi. Rudi was born and grew up in Krefeld, Germany. He was an undergraduate at the University of Geneva, and completed his Ph.D. at the University of Chicago in 1971, which is where we met. He was a student of Robert Mundell, and both the subject matter – the development of the Mundell-Fleming model – and the elegance and insights of his early work reflected Mundell’s influence. He taught at the University of Rochester and at the University of Chicago before accepting an offer from MIT in 1975. In 1976, soon after coming to MIT, Rudi wrote his most famous and influential theoretical article, “Expectations and Exchange Rate Dynamics”. As Ken Rogoff said in his celebratory lecture on the 25 th anniversary of its publication, “The ‘overshooting’ paper … marks the birth of modern international macroeconomics.” From the late 1970s, Rudi became increasingly interested in policy issues. Within

The case for central-bank independence (CBI), while not a new one, has been strengthened by a growing body of empirical evidence, by recent developments in economic theory, and by the temper of the times. The case is a strong one,1 which is becoming part of the Washington orthodoxy. The purposes of this paper are both to make more precise the type of CBI that is likely to enhance economic performance, and to point to some remaining open issues and anomalies. I start with the two theoretical approaches to CBI.