Abstract
The central element of a monetary strategy is the target at which monetary policy is purportedly aimed in order to achieve its desired final objectives. Such a target can help reduce inconsistencies in the implementation and communication of central bank policies. As Friedman put it, “we are in danger of assigning to monetary policy a larger role than it can perform, in danger of asking it to accomplish tasks that it cannot achieve, and, as a result, in danger of preventing it from making the contribution that it is capable of making. ... By setting itself a steady course and keeping to it, the monetary authority could make a major contribution to promoting economic stability.”1
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
M. Friedman (1968, pp. 5 and 17).
B. Friedman (1994) interprets intermediate targeting in this strict sense, describing it as the adoption of a rigid rule on how the central bank reacts to new information. He contrasts this with an ‘information variables’ approach, which is broad-based and explicitly includes judgmental elements. He does not, however, consider the role of intermediate targets in precommitting monetary policy, communicating with the public and establishing central bank accountability. It comes as no surprise that-having disregarded these political economy considerations — he advises against intermediate targeting.
In a sense, the adjective ‘intermediate’ may be viewed as covering all targeting alternatives. Even strategies that focus on variables at the end of the transmission process, such as nominal income and inflation targeting, can be considered variants of the intermediate targeting framework. In both these cases, the forecasts for the target variable are the implicit intermediate targets towards the outturns under the end objectives.
The term “Just do it”-strategy was coined by Mishkin (1997), then Research Director at the Federal Reserve Bank of New York, in an article where the strategy’s disadvantages take up four times the space of the advantages.
Whether a reserve money aggregate or a short-term interest rate should be used to control money growth depends on the interest rate sensitivity and relative stability of the money demand and supply functions. For a discussion of this issue, see Goodhart (1989) and Wenninger (1990).
McCallum (1985) addresses these factors in the context of money targeting.
For a discussion, see Issing (1992) and Kloten (1992). Another viewpoint, set out for example by Kole and Meade (1995), is that the stability of German money demand is also, at least in part, the result of more limited competition and slower financial innovation in German financial markets.
Virtually all IMF programs are centred around credit ceilings, although recently a combination of performance criteria for credit and indicative targets for reserve money has become increasingly common. The design of monetary policy in IMF programs is discussed in Guitián (1994a) and Polak (1997).
This lies at the heart of the monetary approach to the balance of payments; for original essays see Guitián (1972) and Polak and Argy (1971).
A vivid illustration of this elementary distinction between ceilings on domestic credit and on a monetary aggregate is provided in Polak (1991, p. 35). In an IMF consultation with a European country suffering repeated balance-of-payments difficulties (later disclosed by Polak as having been the United Kingdom), the finance minister observes with pride that monetary policy could not be the origin of the problems, as money growth had been kept moderate. He was explained by the IMF staff that “whenever money leaked out through the balance of payments, it was promptly replaced through new credit so it could leak out again”.
This view is set out in Blinder and Stiglitz (1983); see also Alexander and Caramazza (1994).
For a discussion of the evidence for the US, see Radecki (1990) and early work by Fackler and Silver (1983).
A proposal to use total net credit as an intermediate target for US monetary policy is set out in B. Friedman (1983).
In an empirical analysis comparing monetary policy under fixed exchange rates in Austria, Belgium and the Netherlands during 1973–1992, Kool (1995) analyses the rewards of explicitly targeting domestic credit and the exchange rate, which was only done in the Netherlands. His findings suggest there may have been benefits in terms of exchange rate and interest rate stability, but that these were marginal at best.
Steindel (1990) sets out the historic experience with interest rate targeting, especially in the anchor currency of the system, the United States.
The option of targeting the slope of the yield curve should be rejected on the same grounds; Steindel (1990, p 296) discusses ineffective past attempts at targeting the spread between short-and long-term rates.
A detailed analysis of the difficulties of measuring real interest rates is provided by Brown and Santoni (1981).
Mote (1988) illustrates this tendency with historical examples of episodes when central banks targeted interest rates and did not move these rates sufficiently actively to achieve stabilisation objectives.
Sargent and Wallace (1975).
See for example McCallum (1981) and Barro (1989).
Exchange rate targeting should be read as setting a nominal exchange rate objective. Targeting the real exchange rate is inadvisable on account of the limited controllability of this variable. Moreover, a real exchange rate objective leaves monetary policy without a nominal anchor, automatically accommodates domestic price changes, and sets in motion either an inflationary or deflationary spiral depending on whether the underlying real exchange rate is more depreciated or appreciated, respectively, than the target. See also Adams and Gros (1986).
Giavazzi and Pagano (1988) elaborate on this disciplinary influence of ERM membership.
The fact that the ultimate decision on adjustments to exchange rate parities generally rests with the government rather than the central bank further underscores the importance of monetary policy being consistently supported by other macroeconomic policies, if only to make the exchange rate target credible.
Based on developments in German and Dutch consumer prices during the period 1961–91, Berk and Winder (1994) show that the inflation rates in these two countries move in line and are actually co-integrated, implying that the exchange rate peg has anchored the Dutch price level to that in Germany. Besides this, they find that two-thirds of the reduction in the Dutch inflation rate between the periods 1961–1979Q3 and 1979Q4–1991 can be attributed to the policy shift of pegging the guilder to the Deutsche mark within the framework of the European Exchange Rate Mechanism.
The impact of exchange rate instability on expenditure in the context of the Common Agricultural Policy is reviewed by Giavazzi and Giovannini (1989).
As stated by Icard (1994b, p. 241), at the time Director General of Research at the Banque de France.
In the literature, a distinction is made between first-generation (or fundamentalist) and second-generation models of speculative attacks. In the former, pioneered by Krugman (1979), currency crises are precipitated by the behaviour of rational investors anticipating an exchange rate collapse as a result of fiscal profligacy. In the latter, currency crises are triggered by self-fulfilling forces that boost the costs of adhering to an exchange rate commitment. As set out for instance in Obstfeld (1995), such crises reflect the government’s endogenous response to market expectations and do not require deteriorating fundamentals. In an analysis of the 1992 ERM exchange rate crisis, Eichengreen and Wyplosz (1993) present evidence suggesting that self-fulfilling forces played a role.
Total gross official reserves are sufficient to cover the monetary base in most industrialised countries, even without accounting for borrowing capacity; see Obstfeld and Rogoff (1995) who also discuss the technical feasibility of maintaining an exchange rate target in the face of a domestic bank run going beyond conversion of the monetary base.
Mishkin(1997).
The sharp September monetary tightening had largely exhausted support for interest rate increases in defence of the exchange rate. In late November, when sterilised intervention proved unsuccessful in the face of bandwagon effects, the krona was floated. See Hörngren and Lindberg (1993).
In analysing devaluation risk in the ERM, Thomas (1994) finds that interest rate differentials are only weakly influenced by movements in fundamental variables such as inflation, unemployment, relative unit labour costs, government debt and foreign exchange reserves. Based on research of the credibility of the ERM as a system, Knot, Sturm and De Haan (1998) find that inflation rates, budget deficits, unemployment and current account outcomes have generally had an impact on the system’s interest rate differentials; however, they also show that such differentials experienced a sustained decline from 1983 onwards and had all but disappeared by 1991.
In the original analysis of the validity of this hypothesis for the main industrialised countries, Balassa (1964) highlights the potentially misleading signals that price developments (and purchasing power calculations) may provide on the need for exchange rate adjustments. On the basis of an analysis of inflation in eight EU countries during 1975–93, Alberala and Tyrväinen (1998) simulate that inflation differentials of up to 2 percentage points may be attributed to the Balassa-Samuelson effect. In a separate report, the Banco de Portugal (March 1997 Economic Bulletin, p. 8) finds that inflation in the non-tradables sector was 2 to 4 percentage points higher than in the tradables sector during the early 1990s.
For a discussion of the related problems in the Bretton Woods era, see Giavazzi and Giovannini (1989).
This follows the definition of a fixed exchange rate that prevailed under Bretton Woods, as set out in Obstfeld and Rogoff (1995) and Rogoff (1998).
In this broad analysis, the expression ‘nominal income’ can be used interchangeably with ‘nominal GNP’ or ‘nominal GDP’. The subtle differences between these variables do not materially affect the thrust of the arguments set out in general terms for nominal income targeting. Proposals have also been made to target final sales, since this has the technical advantage of circumventing the impact of inventory changes; such a target may, however, be less well understood and thus less useful in communicating the prime considerations driving monetary policy.
The case for nominal income targeting has been made, amongst others, by Tobin (1983), Taylor (1985) and McCallum (1987 and 1996). Reviews of the debate are provided by Hilton and Moorthy (1990) and Clark (1994).
In a more loosely defined framework, Frankel (1990) similarly proposes the use of nominal income targeting as an international policy co-ordination device between the G-7 countries.
For a formal exposition of the stabilising properties of a nominal GNP rule, see Frankel and Chinn (1995). Using a wide variety of empirically based open-economy models, Flood and Mussa (1994) similarly illustrate that monetary policy rules directed at nominal income perform better in terms of output stability and inflation stability than rules focused on the exchange rate or the money supply.
Meltzer (1987) argues that forecast errors for output growth are generally large and that discretionary policies based on these forecasts are unlikely to stabilise the economy. Based on the US experience, Kahn (1988) finds inevitable imprecision and at times large errors in nominal GNP forecasts.
Taylor (1985, 1993) presents a policy rule that can be adjusted according to the relative welfare effects of output and inflation fluctuations. As an alternative approach, Hall (1985) proposes an ‘elastic price standard’ under which the central bank is instructed to stabilise prices at a particular level, but is allowed to depart from this level when unemployment deviates from target. The magnitude of the permissible price level departure is equal to the deviation of unemployment multiplied by a predetermined elasticity. This elasticity effectively determines the short-run trade-off between inflation and unemployment; Hall suggests it be set between 5 and 8.
Mishkin and Posen (1997).
In the more distant past, Sweden explicitly pursued a policy of price-level stabilisation in the 1930s, making it the first country on record to have aimed monetary policy openly at a price target; it is also the only country to date to have targeted the price level stability rather than an inflation rate. This early Swedish experience is set out in Jonung (1979) and Berg and Jonung(1999).
Leiderman and Svensson (1996).
M. Friedman (1968, p. 15) rejected this assumption, although he acknowledged that: “Perhaps, as our understanding of monetary phenomena advances, the situation will change.”
Haldane (1995a) draws a parallel between the target choice and decisions on an optimal asset portfolio, where the precept applies that all eggs should not be placed in one basket.
In a formal comparison between inflation targets and monetary targets, Cukierman (1995) assumes that visibility is the main advantage of the former.
In New Zealand, the governor’s employment contract can be discontinued if the inflation target is breached. See Debelle (1997).
Masson, Savastano and Sharma (1997). The Bank of England was an exception, however, as it did not have instrument independence during the United Kingdom’s first 4½ years of inflation targeting.
This was originally not the case in Sweden, as the Government initially had a different inflation objective and only officially accepted the Riksbank’s target after 2½ years of inflation targeting.
In this context, Cecchetti (1997) presents data suggesting that the inflation-output variability trade-off is extremely steep, as small declines in inflation variability are associated with steep increases in output variability. By contrast, using simulation techniques, Haldane and Batini(1998) show that an inflation forecast-based rule with a judiciously chosen targeting horizon naturally embodies the desired degree of output-smoothing. Indeed, Goodhart (1999) summarises the evidence that-for the United Kingdom at least — the trade-off is not large, as long as the central bank chooses an appropriate time horizon to meet its inflation target.
Green (1996).
Svensson (1997).
Woodford (1994) criticises the use of indicators that forecast future inflation, rather than variables that are proximate causes of inflation, as a guide to the conduct of monetary policy. In particular, he emphasises the danger of self-fulfilling prophecies when a policy rule is adopted that feeds back from an indicator that is itself sensitive to policy expectations. Bernanke and Woodford (1997) elaborate on the risks of using private sector inflation forecasts as a guide to monetary policy. While such forecasts can provide useful information, they find that the central bank must ultimately rely on an explicit structural model of the economy to guide its policy decisions.
With an element of jest, Blinder’s Law of Speculative Markets states that markets usually get the sign right, but tend to exaggerate by a factor between three and ten; see Blinder (1997).
Goodhart (1995) echoes the call by Alchian and Klein (1973) that the orientation of monetary policy should be broadened beyond consumer prices to include asset prices. In his view, this would focus policy on a more relevant concept of inflation (since asset price increases are a direct source of future consumer inflation), while also countering the procyclical influence of monetary policy through the credit channel. However, this approach has numerous problems, prominent amongst which are the difficulty of correctly identifying the inflationary component of asset price increases and the instrument volatility this would imply. For a discussion of the monetary policy implications of asset inflation, see Capei and Houben(1998).
The Curriculum Vitae of M. King, Executive Director and Chief Economist of the Bank of England, is said to make reference to the fact that he used to play tennis until the United Kingdom adopted inflation targetry.
Of course, it is still early to come to a firm judgement on the efficacy of direct inflation targeting. Strictly, the jury remains out until there is a broad-based experience across a full cycle (or two) and under a wide variety of shocks. In this respect, Groeneveld, Koedijk and Kool (1998) juxtapose the performance of three inflation targeting countries (the United Kingdom, Canada and New Zealand) with three comparator countries (Germany, the United States and Australia, respectively) and do not find strong evidence of a structural break in the relationship between inflation and interest rates at the time countries switched to inflation targeting. This is also the conclusion reached by Bernanke et al. (1999). However, based on a comparison between seven inflation targeting countries and four control groups, Almeida and Goodhart (1998) present tentative evidence suggesting that the adoption of inflation targeting reduces the sacrifice ratio of disinflation in terms of unemployment and especially foregone output growth.
See Fischer (1994 and 1996), Goodhart and Vinals (1994), McCallum (1996) and Summers (1991).
On the first point, Akerlof, Dickens and Perry (1996) present evidence that nominal wages are downward rigid. In a zero inflation environment, this reduces the scope for real wage adjustments and leads to an unnecessarily higher rate of unemployment. However, their findings are not uncontested (see Yates (1998) for an overview of the inconclusive empirical literature). Also, in a context of generally increasing labour productivity, there will usually be scope for some adjustment in real labour costs even when nominal wages remain stable. On the second point, Fuhrer and Madigan (1997) show that the adjustment of the economy to adverse shocks is significantly slower in a zero-inflation climate, since nominal interest rates cannot be set below zero in practice, thereby constraining the central bank in its ability to establish negative real interest rates and to cushion the output decline. Their conclusions are corroborated by Orphanides and Wieland (1998), who simulate the impact of typical stochastic shocks on the US economy and find that monetary policy effectiveness is significantly reduced when inflation is targeted between 0 and 1 per cent. As a result, recessions become more frequent and longer lasting. The consequences of the zero nominal bound are found to be negligible for inflation targets of 2 percent or more. See also Fischer (1996) and, for the original essay on this issue, Summers (1991).
While the prominent study by Boskin et al. (1996) concludes that the US Consumer Price Index exaggerates inflation by 1.1 percentage point per year, subsequent studies for European countries suggest a somewhat smaller measurement bias. For instance, the Bundesbank study carried out by Hoffmann (1998) indicates that the overall bias in Germany may be in the order of 0.75 percentage point per year.
The deviations under the inflation assumptions built into Germany’s money target are presented in Appendix VI. It has been assumed that deviations above 1.0 percentage point (respectively 1.5 percentage points) would have fallen outside a hypothetical bandwidth of 2 percentage points (respectively 3 percentage points). Over the past twenty-four years, the average absolute deviation has been 0.9 percentage point, with a standard deviation of 1.1.
The only country in Europe to have explicitly pursued more than two targets is Italy, which combined targets for money, credit and the exchange rate in the 1980s. Outside Europe, a notable example of the mixed results to be expected from multiple targeting of nominal anchors is provided by the Philippines, where base money and inflation targets were combined with in practice overriding exchange rate targets in the mid-1990s; see Houben (1997). In principle, the basic arguments against strategies with two targets hold a fortiori against strategies with three targets.
For example, Clarida, Gali and Gertier (1997) find that since 1979 each of the G3 countries-as well as the United Kingdom after it relinquished its exchange rate commitment in 1992-has followed a remarkably similar forward-looking policy rule, whereby nominal interest rates are adjusted in response to changes in expected inflation and the output gap. Thus, notwithstanding different declared monetary policy strategies, each of these countries has effectively pursued an inflation target with some allowance for output stabilisation-a strategy which they coin “soft-hearted inflation targeting”. Chadha and Janssen (1997) corroborate these findings with evidence that changes in official interest rates in each of the G3 countries plus the United Kingdom have been prompted by a similar set of variables through time, regardless of the announced policy rule. These authors venture an interpretation that central banks combine a stated policy rule with discretion and, in practice, pursue something like a state contingent inflation target or a nominal income rule. In a similar vein, Bernanke and Mihov (1997) find that the most ardent advocate of money targeting, the German Bundesbank, in fact hardly reacts to changes in forecasted money growth and “is much better described as an inflation targeter than as a money targeter.”
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2000 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Houben, A.C.F.J. (2000). Targeting Options for Monetary Policy. In: The Evolution of Monetary Policy Strategies in Europe. Financial and Monetary Policy Studies, vol 34. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-4471-5_3
Download citation
DOI: https://doi.org/10.1007/978-1-4615-4471-5_3
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4613-7014-7
Online ISBN: 978-1-4615-4471-5
eBook Packages: Springer Book Archive