The Evolution of Inflation Targeting, 1992–1998

  • Aerdt C. F. J. Houben
Part of the Financial and Monetary Policy Studies book series (FMPS, volume 34)


The rise of direct inflation targeting in Europe during the 1990s was spurred by conceptual advances in monetary policy strategy, technical progress in the analysis of inflation, as well as practical problems with prevailing strategies. On the first issue, the inflationary experiences of the 1970s and early 1980s had brought home the lesson that there is no durable positive trade-off between inflation and unemployment (at least, in case of high inflation) and that, if anything, this relationship is negative in the longer run. Moreover, the view gained ground that a monetary policy focus on real variables introduces an inflationary bias. As a result, it became increasingly accepted that monetary policy should prioritise the establishment and maintenance of price stability. In Europe, this shift was apparent in institutional changes that spelt out an overriding price stability objective for monetary policy and that granted central banks the related policy autonomy. This was most notably the case in the 1991 Maastricht Treaty setting out the creation of a European System of Central Banks (ESCB), constituted of independent, price stability oriented entities.


Monetary Policy Central Bank Inflation Target Central Bank Independence Inflation Forecast 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Berk (1998) provides a comprehensive overview of the learning process-and the remaining uncertainties — on the monetary policy transmission process, and assesses how this incomplete knowledge can nonetheless be put to good use by policy-makers.Google Scholar
  2. 2.
    Almeida and Goodhart (1998) suggest that raising the accountability of the monetary policy decision-making process was a prime reason for the adoption of direct inflation targeting. It is unclear, however, why a monetary authority in one of the European countries (the Government in the United Kingdom and the central banks in the other countries) would introduce inflation targeting to increase its own accountability, since this accountability generates potentially high costs to that authority. Rather, it seems to have been the underlying quest for policy credibility that drove these countries to this monetary strategy.Google Scholar
  3. 3.
    For an extensive discussion of this early Swedish experience of inflation targeting, see Jonung (1979) and Berg and Jonung (1999).Google Scholar
  4. 4.
    The position papers presented to the Swedish Governor in December 1992 after the krona had left the ERM and prior to the decision on the new monetary policy strategy, included a review of Sweden’s price level stabilisation policy in the 1930s; see Sveriges Riksbank (1992).Google Scholar
  5. 5.
    The sustainability of the framework was, as it were, made conditional on the support it could earn from the public and opposition parties; see Mishkin (1997).Google Scholar
  6. 6.
    The lack of a fall-back scenario is mentioned by Svensson (1995, p. 70). This lack further testifies to the sincerity of the authorities’ exchange rate commitment.Google Scholar
  7. 7.
    In an effort to stimulate an open debate on monetary strategy in the context of a floating exchange rate regime, the articles were promptly published in December 1992 as a Special Issue of the central bank’s Quarterly Review; see Sveriges Riksbank (1992).Google Scholar
  8. 8.
    Åkerholm and Brunila (1995).Google Scholar
  9. 9.
    Bank of Finland (1996, p. 3). It is noteworthy that the central bank referred to the increasing importance of exchange rate stability in the conduct of economic policy (and not specifically of monetary policy). In this context, Almeida and Goodhart (1998, p. 21) draw a questionable conclusion when they classify Finland as no longer being an inflation targeting country after it joined the ERM. Although the classification admittedly becomes murky when countries formally pursue multiple policy targets, Finland’s position seems to have been very close to that of Spain (which they classify as an inflation targeting country despite its ERM membership).Google Scholar
  10. 10.
    Vega (1994) and Ortega and Bonilla (1995).Google Scholar
  11. 11.
    Gutiérrez (1998) and Ayuso and Escrivá (1998).Google Scholar
  12. 12.
    Visco (1995) discusses the increasing role of inflation objectives in the Italian economy during 1980–94.Google Scholar
  13. 13.
    An eye-catching proposal was developed at the IMF by Lane, Prati and Griffiths (1995).Google Scholar
  14. 14.
    The text of the Governor’s concluding remarks is customarily reproduced in the central bank’s Annual Report.Google Scholar
  15. 15.
    Next to announcing a prospective inflation path, the Governor indicated that “If price trends over the next few months show a tendency to diverge from the pattern I have just described, we shall not hesitate to tighten credit conditions still further.” See Banca d’Italia, Annual Report 1994 (p. 173).Google Scholar
  16. 16.
    Von Hagen (1995) views the Bundesbank as a covert inflation targeter of sorts and classifies the monetary policy strategy of Germany as a case of combined ‘inflation and monetary targeting’. Similarly, Mishkin and Posen (1997) include Germany-next to Canada, New Zealand, and the United Kingdom — in a case study review of inflation targeting experiences.Google Scholar
  17. 17.
    The arguments for, and against, using a stripped rather than a headline inflation measure are also applicable to the use of caveats for pre-defined events.Google Scholar
  18. 18.
    In essence, the caveat for indirect taxes, subsidies and interest costs was incorporated — or so it seems — to compensate for initially having chosen the wrong target measure and not wanting to raise suspicions by adjusting this. Consideration was given to switching the targeted inflation measure from RPIX to RPIY in 1995, but the proposal was rejected on the grounds that frequent shifts in the target definition risked undermining the framework’s credibility; see Mishkin (1997, p. 73). In fact, to enhance credibility, the calculation and publication of the RPIX index was moved during this year from the Bank of England to the independent UK Office of National Statistics.Google Scholar
  19. 19.
    Heikestein and Vredin (1998) discuss the complex considerations that governed the Riksbank’s initial choice of target variable as well as the gradual change in emphasis that occurred over time.Google Scholar
  20. 20.
    Goodhart and Viñals (1994). However, a target that is judged to be too tight will not be credible to start with.Google Scholar
  21. 21.
    As stressed in the Chancellor of the Exchequer’s original Remit for the newly created Monetary Policy Committee (printed in the Bank of England’s August 1997 Inflation Report, p. 75), the requirement to send an open letter when inflation moves more than 1 percentage point from target is not to establish a tolerance range for deviations, but to enhance monetary policy transparency. In line with this instruction, King (1997, p. 441) emphasises that the United Kingdom’s inflation target is not a range of 1½ to 3½ per cent, but 2½ percent on average.Google Scholar
  22. 22.
    Brunila and Lahdenperä (1995, p. 129).Google Scholar
  23. 23.
    See the comments by Ortega in Haldane (1995b, p. 77).Google Scholar
  24. 24.
    The first European definition of price stability was spelt out in a report to the EU central bank governors by the Group of Experts chaired by Raymond (1990).Google Scholar
  25. 25.
    Along similar lines, Andersson and Berg (1995) draw a distinction between a monetary strategy’s political credibility, which reflects investors’ confidence that the strategy itself will be maintained (rather than be replaced by a more inflationary alternative), and its operational credibility, which relates to the perceived likelihood that the central bank’s policies will actually deliver the announced inflation target.Google Scholar
  26. 26.
    This can be inferred from Svensson (1995, pp. 73-76).Google Scholar
  27. 27.
    Brunila and Lahdenperä (1995, p. 130).Google Scholar
  28. 28.
    Ayuso and Escrivá (1998) and Gutiérrez (1998).Google Scholar
  29. 29.
    The Bank of England’s Inflation Report is actually produced in two editions, tailored to different groups of readers: while the standard edition is comprehensive, with considerable detail and analytical depth, there is also a popular version providing a broad-brush and readily accessible overview of developments in (prospective) inflation.Google Scholar
  30. 30.
    As the Chancellor of the Exchequer was then still in charge of monetary policy-making, his inflation forecast was actually the more relevant. With the transfer of policy autonomy a little more than a year later, this peculiar irregularity was ironed out.Google Scholar
  31. 31.
    The conditional nature of the inflation forecast in the fan charts leads to a probability distribution that widens as the time horizon is extended. In reality, of course, the controllability of inflation may be assumed to increase beyond the lags of the monetary transmission process, implying a narrowing of the probability distribution. But this requires policy adjustments to keep inflation on track. Moreover, the assumption of unchanged policy raises issues of consistency, since many financial variables that serve as inputs to the forecasts obviously do not assume a static policy. At a technical level it also open to debate whether the fan chart should centre on the mode of the distribution (the outcome considered most likely) as has been the common practice, rather than on the mean (the average outcome). For a discussion of the genesis and technical evolution of the Bank of England’s fan chart, see Britton, Fisher and Whitley (1998).Google Scholar
  32. 32.
    Just as in the United Kingdom, the progress of time has seen the Riksbank’s graphical inflation forecasts become fancier and more informative. Remarkable in the case of Sweden is that fan charts are now presented for the targeted CPI inflation measure as well as an underlying inflation measure. Berg and Gröttheim (1997, pp. 169-171) discuss the efforts of the Riksbank to advance market communication.Google Scholar
  33. 33.
    At a global rather than regional level, Debelle (1997) and McCallum (1996) similarly find evidence of improved central bank performance following the introduction of inflation targeting schemes. However, Almeida and Goodhart (1998) do not find the inflation experiences of countries with inflation targets to be statistically significantly different from those in a variety of control groups (although they do find the fall in inflation on average to have been sharper).Google Scholar
  34. 34.
    The inflation targeting record is evaluated on the basis of annual figures. While quarterly or monthly (or even higher frequency) figures could also be used, inflation targets have often been cast explicitly in an annual framework (see, for instance, Berg and Gröttheim, p. 170). The impact of inflation targets on inflation assumptions in other sectors (e.g., the budget and wage contracts) also generally follows an annual cycle. As a welcome by-product, an assessment using annual data annual data facilitates comparison with other monetary policy strategies (especially money targeting) where targets are set in annual terms.Google Scholar
  35. 35.
    Based on a comparison of disinflation costs between seven inflation targeting countries (including the four European ones) and four different control groups, Almeida and Goodhart (1998) find preliminary evidence that the switch to inflation targeting reduced sacrifice ratios in terms of unemployment and especially foregone output growth. By contrast, Groeneveld, Koedijk and Kool (1998) compare the experiences of New Zealand, Canada and the United Kingdom with those of Australia, the United States and Germany (respectively) and do not find evidence of a structural break in the relationship between inflation and interest rates following the switch to inflation targeting. They conclude that the credibility-enhancing impact of inflation targeting is ambiguous and is in any case not clearly superior to that of other intermediate monetary strategies. Similarly, based on case studies of Germany, New Zealand, Canada and the United Kingdom, Mishkin (1997) concludes that credibility cannot be achieved by adopting an inflation target, but has to be earned the hard way — by consistent policy implementation. This is also the conclusion drawn by Bernanke et al. (1999).Google Scholar
  36. 36.
    Evidence on these two counts is presented in King (1997, p. 436). The control group comprises G7 non-inflation targeting countries.Google Scholar
  37. 37.
    The estimated differential of one-half of a percentage point between inflation as measured by the Retail Prices Index excluding mortgage interest payments (RPIX) and by the Harmonised Index of Consumer Prices (HICP) has been established as the average for the period 1976–98, although it has varied markedly during this period. The differential can be mainly attributed to different weighting schemes (while the RPIX uses arithmetic averages, the HICP uses geometric means) and to different coverage (especially a larger RPIX coverage of housing costs); see UK Treasury (1998).Google Scholar
  38. 38.
    According to the definition determined by the group of EU central bank experts chaired by Raymond (1990).Google Scholar
  39. 39.
    This contrasts sharply with the estimates by Haldane and Salmon (1995). Based on simulations for the United Kingdom (using quarterly data), they find significant inflation uncertainty, suggesting a range as wide as six percentage points may still be missed in about one-third of the time.Google Scholar
  40. 40.
    Based on the data presented in Appendix VI, where the GDP deflator is used as the measure of inflation.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2000

Authors and Affiliations

  • Aerdt C. F. J. Houben
    • 1
  1. 1.De Nederlandsche BankAmsterdamThe Netherlands

Personalised recommendations