Conclusion
In this chapter, we have delved deeper into policies that the monetary authorities can implement to control inflation and stabilize the real economy. Four conclusions have emerged.
The first is that monetary policy should be determined by politically independent central banks on the basis of rules rather than discretion. Political independence of the central bank and well-defined objectives result in a more credible monetary policy. It helps avoid both the high inflation that could result if the central bank has to fund government deficits through seigniorage and the inflationary bias that would result if the central bank was under political pressure to reduce unemployment below the natural rate. Central bank constitutions should constrain them to pursue a low inflation target, as there is general agreement that the optimal rate of inflation is low.
The second conclusion is that in the presence of stochastic disturbances, policy rules should depend on the state of the economy. Such rules are called contingent. The central bank must be able to react to shocks that cause deviations of real variables from their natural rates, even if this results in temporary deviations of inflation from its medium term target, as is the case with supply shocks. For demand shocks, the optimal reaction of monetary policy does not imply a trade-off between employment and inflation.
A third conclusion is that the appropriate policy instrument of monetary policy is the short-term nominal interest rate, which can be controlled through open market operations. Control of the money supply (to the degree that an appropriate measure of the money supply can be defined) has always been difficult, and it becomes even more difficult if there are financial innovations and money demand shocks. Thus for some years, the focus of monetary policy analysis has been the appropriate contingent rule for short-term nominal interest rates.
The Taylor rule has provided the appropriate framework, because it allows for reactions of the central bank policy rate to the natural rate of interest, deviations of inflation from the low-inflation target, and deviations of output or employment from their natural rates. It provides flexibility and allows the central bank to pursue a low-inflation objective over the business cycle, which also results in low nominal interest rates.A final conclusion is that with a low-inflation objective, there is always the risk that a big enough negative shock to aggregate demand will force monetary policy to hit the zero lower bound of interest rates. In such cases, and especially in the presence of financial frictions, the short-run nominal interest rate ceases to be an effective instrument of monetary policy. Then central banks must follow unconventional monetary policies, such as forward guidance, quantitative easing, and asset purchases.
1. For example, in chapter 6, we investigated the effects of fiscal policy variables, such as government expenditure, taxes, and government debt, on economic growth. In chapter 7, we investigated the effects of the money supply and monetary growth on economic growth and nominal variables, such as the price level, inflation, and nominal interest rates. In chapter 12, we investigated the implications of monetary growth and central bank interest rate policy on prices and inflation. In chapters 14–17, we investigated the effects of alternative rules for the money supply and nominal interest rates for aggregate fluctuations in real and nominal variables. In chapter 18, we investigated the effects of unemployment benefit systems on equilibrium unemployment. In most cases, the policy variables (or the parameters of the policy rules determining them) were treated as exogenous.
2. See chapter 15 for a discussion of the rules versus discretion debate in the context of the Phillips curve. Fischer [1990] surveys the literature on rules versus discretion for monetary policy.
3. Chapters 12 and 15 introduced this discussion. The classic analysis of the appropriate choice of monetary instruments in a Keynesian model was Poole [1970], who conducted his analysis on the basis of the Tinbergen [1952]–Theil [1954] framework, which stressed the distinction between policy instruments and policy targets. Sargent and Wallace [1975] demonstrated that under rational expectations, a noncontingent interest rate target leads to price level indeterminacy and instability (see chapter 12). However, it is now accepted that this problem does not arise in the case of contingent interest rate rules that allow for the nominal interest rate to depend on the price level (McCallum [1981]), or to be a sufficiently sensitive positive function of inflation. See Woodford [2003a], chapter 1 and chapter 12, for the relevant arguments. In any case, central banks have been consistently using interest rates as their main monetary policy instrument. As noted by Bernanke [2006, p. 2], “In practice, the difficulty has been that, deregulation, financial innovation, and other factors have led to recurrent instability in the relationships between various monetary aggregates and other nominal variables.”
4. See Clarida et al. [1999], Woodford [2003a, 2011] Gali [2008, 2011a,b], and Taylor and Williams [2011] for surveys of the findings of this literature. Woodford [2003a] and Orphanides [2003c, 2008] survey the analysis of interest rate rules from Wicksell [1898] to Taylor [1993].
5. One could of course use the staggered pricing model of chapter 16 to analyze optimal monetary policy. For analyses of monetary policy based on this type of new Keynesian model, see Woodford [2003a] and Gali [2011b]. Central banks increasingly use empirically estimated generalized new Keynesian models with both price and wage rigidities to evaluate policy. Such models have been calibrated, estimated, and empirically implemented by Erceg et al. [2000], Smets and Wouters [2003, 2007], Christiano et al.
[2005], and others. However, not all of these models are analytically tractable. Simpler models, such as the ones we utilize, are better for highlighting the theoretical basis of the mechanisms at work.6. Equation (20.2) is the expectations-augmented Phillips curve derived in chapter 17, with the additional restriction of lack of persistence (i.e., by imposing δ = 0).
7. As demonstrated in chapter 15, the same outcome would be a feature of an ad hoc model of an expectations-augmented Phillips curve under adaptive expectations. The only difference is that the process of convergence to the equilibrium with high inflation will take more time when expectations are formed adaptively. For a collection of early papers on the issue of credibility of monetary and fiscal policy, see Persson and Tabellini [1994a].
8. As noted by Friedman [1968, p. 13]: “Our economic system will work best when producers and consumers, employers and employees, can proceed with full confidence that the average level of prices will behave in a known way in the future—preferably that it will be highly stable.”
9. For example, the Mortensen-Pissarides model that we examined in chapter 18 can be used to suggest some such policies. See Pissarides [2000].
10. The analysis of the significance of reputational mechanisms in repeated games is due to Kreps and Wilson [1982a, b].
11. See chapter 12, or Woodford [2003a, chapter 1] for the relevant arguments.
12. See Capie et al. [1994]. As noted by Bernanke [2006, p. 2], “In practice, the difficulty has been that, deregulation, financial innovation, and other factors have led to recurrent instability in the relationships between various monetary aggregates and other nominal variables.”
13. Obviously, these rules can always be expressed as contingent rules for the money supply by substituting for the nominal interest rate in a money demand function and solving for the money supply.
14. Taylor [1993, 1999] proposed a rule in which the natural real rate of interest was constant at 2% and the inflation target was equal to 2% as well.
Let us term this rule the constant-intercept Taylor rule. However, in the analysis that follows (as also done in the analysis of chapter 17), we stick to the spirit of the Taylor rule but treat the natural real rate of interest as a stochastic endogenous variable and not a constant.15. This point was first highlighted by Woodford [2001].
16. See Orphanides and Williams [2002] and Orphanides [2003a, b] for theoretical and empirical analyses of the case of such estimation errors.
17. Clarida et al. [1999] analyze monetary policy in a new Keynesian model based on staggered pricing and reach overall conclusions similar to those of this chapter, which uses a model of periodic wage contracts.
18. Alogoskoufis [2018] analyzes monetary policy when deviations of unemployment from its natural rate display persistence, as in the model of chapter 17.
19. Eggerstsson and Woodford [2003], who used a model with staggered pricing, suggested that this can be achieved by announcing a target price-level path, instead of a higher inflation path.
20. See McGough et al. [2005] for a theoretical analysis of this type of unconventional monetary policy.
21. Williams [2014] provides an assessment of the practical difficulties in pursuing unconventional monetary policy at the zero lower bound. Krugman [2009a, 2012] is the main proponent of a return to a more intensive use of fiscal policy for stabilization purposes.