Between 1980 and 1995, labor productivity in the business sector grew at an average annual rate in Canada, which was slightly faster than productivity growth in Germany, but significantly slower than labor productivity growth in France, Italy, Japan, and the United States. To better understand developments in labor productivity, it is useful to decompose its growth rate into changes in the capital/labor ratio and in total factor productivity. The contribution of information technology to labor productivity growth has been more modest in Canada than in the United States.

Abstract

Between 1980 and 1995, labor productivity in the business sector grew at an average annual rate in Canada, which was slightly faster than productivity growth in Germany, but significantly slower than labor productivity growth in France, Italy, Japan, and the United States. To better understand developments in labor productivity, it is useful to decompose its growth rate into changes in the capital/labor ratio and in total factor productivity. The contribution of information technology to labor productivity growth has been more modest in Canada than in the United States.

IV. Price Stability and the Choice of Inflation Target for Monetary Policy1

1. Since its introduction in 1991, Canada’s inflation targeting framework has successfully curbed and then maintained inflation at low levels. At the last extension of the inflation target in 1998, the Government and the Bank of Canada announced that before the end of 2001 an appropriate long-run target consistent with price stability would be determined. In deciding whether to change the inflation target, the authorities will have to consider what are the possible benefits and costs of lowering inflation further. Moreover, the decision to examine the issue of a target consistent with price stability has raised a question among some observers as to whether a regime change from inflation targeting to price level targeting may be warranted. In coming years, other countries that have adopted inflation targeting will be faced with similar questions as they consider adjustments to their targets (Table 1).

Table 1.

Selected Countries: Advanced Economies with Explicit Inflation Targeting Frameworks and Key Features

article image
Sources: Masson, Savastanoand Sharma (1997); and Debelle (1997).

In May 1997, the Bank of England was given operational independence to set interest rates in order to achieve the inflation target (set by the U.K. Treasury). Inflation outside the target range would require the Governor to write an open letter to the Chancellor to explain the reasons for the deviation.

2. The empirical literature does not offer much insight on the relative magnitude of the benefits to be gained in moving from low to lower inflation, perhaps in part because of the limited recent experience with very low inflation. Stable, lower inflation provides a firmer anchor for expectations and diminishes volatility in the aggregate price level. Thus, lower inflation can foster more effective longer-term planning which would entail efficiency gains for the economy. Costs associated with such a change in the inflation target to lower inflation center on nominal rigidities in the economy, with consideration generally focused on nominal wage rigidities. Empirical evidence suggests that wage rigidity may not be as severe as previously expected (at least when downward wage adjustments are relatively small) and that such rigidities could dissipate over time as individuals adjust to stable, low inflation. However, some nominal rigidities (such as those associated with the costs of contract renegotiations, information asymmetries, and the costs of monitoring inflation) are likely to continue to persist, imposing significant costs on the economy in moving to lower inflation. In considering a regime change, the economic literature has largely focused on the possible benefits of inflation versus price level targeting; it has not investigated all of the possible costs of such a regime change, especially the costs of the transition to a price level target. Moreover, with theoretical benefits of such a regime being highly dependent on the assumptions made in the economic models used, there is substantial uncertainty regarding what might be gained from switching from an inflation to a price level target.

A. Moving from Low to Lower Inflation and Price Stability

3. In contrast to the case of lowering inflation from high rates,2 the benefits of moving from a low inflation rate closer to price stability are difficult to evaluate. Moving toward price stability, which is specially important for economic decisions with long gestation periods, would entail efficiency gains to the economy. The costs associated with such a change are also highly uncertain. The potential costs of further inflation reduction are mainly associated with the extent and duration of downward nominal rigidities that impede adjustment in the economy and increase employment and output volatility. These nominal rigidities can result from a variety of sources, with the most commonly recognized one being nominal wage rigidity. Some of these costs (like those arising from wage rigidities) could be expected to be transitory, but the length of this transition period and the magnitude of the associated costs are difficult to estimate, especially when inflation is being reduced by only a small amount. A further reduction in inflation also would entail other adjustment costs since it would involve an effective transfer of real resources from debtors to creditors.

4. The empirical literature provides little guidance to draw firm conclusions on the desirability of moving from low to lower inflation. Statistically significant evidence is hard to find because of the lack of recent experience with very low inflation rates. Nevertheless, part of the literature has placed the burden of proof on the cost side, with the recurrent question of whether some level of inflation is needed to “grease the wheels” of the economy and eliminate the potential negative effects of nominal rigidities, particularly wage rigidity. The argument is that workers may strongly resist nominal wage rollbacks, but inflation might be able to produce a decline in real earnings. However, the behavior of workers should not be taken as exogenous to the regime they face. Thus, nominal wages could be expected to become more downwardly flexible in an environment of relative price stability. This process, nevertheless, could take time.

5. Several papers have explored the extent of downward wage inflexibility in Canada in recent years. The bulk of the evidence suggests that nominal wage rigidity may not be a pervasive phenomenon, at least over the range of small nominal wage reductions observed. Evidence of wage rigidity tends to be more prevalent in some specific sectors (e.g., large unionized firms). The impact of wage rigidity on aggregate employment and output appears small because it does not affect large segments of the economy (non-unionized and smaller firms) and because of the flexibility embedded in variable compensation schemes. Empirical studies suggest that nominal wage flexibility in Canada exhibits patterns closer to those in the United States and the United Kingdom than to those in other European economies; in the latter group of countries, the length, extent of indexation, and synchronization of wage contracts are much higher.

6. Wage freezes and lack of rollbacks have frequently been considered a first test of downward nominal wage rigidity.3 Crawford and Harrison (1997) analyze the distribution of base salary settlements in large unionized contracts during the 1990s. They find that a large number of cases of unchanged wages could be partly explained by the presence of downward wage rigidity. However, broader wage measures, including bonuses, that provide additional flexibility in overall labor compensations, show a higher incidence of rollbacks. This indicates that total wage compensation exhibits a higher degree of nominal downward flexibility than suggested by base salaries. Moreover, wage settlement data cover only a relatively small share of total employment in Canada. Thus, from the analysis of various data sources, the paper concludes that wage settlement data overstates the extent of downward rigidity in the Canadian economy.

7. With respect to employment growth in Canada, Faruqui (2000) finds that downward nominal wage rigidities have had no discernible effect Likewise, Crawford and Hogan (1999) show that wage inflexibility has had a relatively small effect on the outcome of wage negotiations and employment in the low-inflation period of the 1990s. The paper estimates that menu costs can explain at least half of the wage freezes in the private sector data over the period 1992-98, and concludes that the number of wage freezes significantly overstates the importance of wage rigidity. In addition, it finds that variable pay schemes provide an additional margin of flexibility to wage costs that lessens the impact on employment and real activity.

8. In a different exercise, Farès and Lemieux (2000) also find a lack of significant impact from downward wage rigidity during the 1990s. Controlling for composition effects,4 they estimate a Phillips-curve relationship between real wages and unemployment. Nominal wage rigidity, by preventing real wages from adjusting to negative employment shocks in periods of low inflation, should flatten the Phillips curve compared to periods of high inflation. The paper finds that, during the low-inflation period of the 1990s, the Phillips-curve estimates do not suggest that the slope of the curve became flatter. The authors interpret this fact as evidence that downward rigidity did not have a significant effect on wages and employment. In part, this is because new entrants (young workers and workers on new jobs) seemed to bear a large share of the real wage adjustments over the business cycle, more than offsetting the effect of rigidities that may have been binding for other groups of workers.

9. It is possible, however, that nominal wages are flexible over a relatively narrow range (e.g., when downward adjustments in nominal wages of, say, less than 5 percent are required), but they may turn out to be much more rigid when larger wage cuts are needed. Larger wage cuts might prompt more worker resistance because of their impact on incomes and uncertainties about the need for such cuts given information asymmetries between management and workers regarding the financial viability of firms.

10. Other sources of nominal rigidities (such as menu costs) may impose significant costs on an economy as it moves from low to lower inflation. Also, in contrast to the case of nominal wages, the incidence of some of these rigidities could be expected to continue after the transition period to a regime of price stability, because such rigidities can be independent of the monetary regime. They may be rooted in the costs of contract renegotiation, information asymmetries, and costs of monitoring inflation, among other reasons.

11. Moreover, in moving from low to lower inflation (particularly moving to an inflation target consistent with “price stability”), the basic question of bias in the measurement of inflation has to be considered. In the case of Canada, the bias in the measurement of the CPI is estimated to add ½ of a percentage point a year to the measured inflation rate. This is generally less than the measurement bias in other major industrial countries (Table 2).

Table 2.

Canada: Bias in the Consumer Price Index in Major Countries

article image
Sources: Shiratsuka (2000); and Congressional Budget Office (1999).“<” indicates estimated bias is lower than the figure in the table.

B. Changing the Regime from Inflation to Price-Level Targeting

12. With inflation in recent years at low and stable rates, the question has arisen whether there might be significant benefits to be gained with little costs by shifting from targeting inflation to some form of targeting the price level. Pure price-level targeting consists of minimizing the deviation of the price level from a target level, thus precluding long-run price-level drift, while inflation targeting aims at minimizing inflation rate deviations from a target inflation rate, which allows for the possibility of price-level drift. Some hybrid alternatives could, in principle, limit the scope of price level drift by incorporating some form of error-correction mechanism which, for example, might provide for some correction of deviations in inflation from the mid-point of the inflation target band.5

Conventional views

13. Price-level targeting provides two potential advantages with respect to inflation targeting. First, it favors long-term planning and nominal contracting by precluding price-level drift.6 Second, it provides a firmer anchor for expectations. A more effective information environment allows people to more efficiently differentiate changes in relative prices.7 However, there are drawbacks to price-level targeting. In particular, it induces a higher probability of deflationary episodes, and thereby, it increases the possibility of financial instability. Under price-level targeting, unexpected shocks to the price level are not treated as bygones, so they must be offset. This requires a more contractionary monetary policy to reverse the overshooting of the price level, producing a higher probability of deflationary episodes.8 In economies where long-duration debt contracts entail fixed nominal interest payments, price deflation increases the liability burden of debtors in real terms9. Not only does this balance-sheet effect reduce firms’ financial capacity, but it also affects borrowers’ repayment incentives, as net worth and collateral values fall. This, in turn, exacerbates adverse selection and moral hazard problems common to financial contracts and increases the probability of default and financial instability. Moreover, the gains from lower long-term inflation variability resulting from price-level targeting may be limited in developed countries where financial markets have developed an array of hedging instruments that can efficiently reduce the risk of price fluctuations.10

14. A price-level target that entails a very low or zero inflation rate would increase the possibility of episodes where short-term interest rates fall to zero. By having a zero-bound constraint on short-term nominal interest rates, monetary authorities are limited in their ability to lower interest rates to affect consumption and investment behavior.11

15. The recent Japanese experience presents an example of monetary policy being constrained by zero nominal interest rates and having a reduced ability to influence monetary conditions. From February 1999 to August 2000, the Bank of Japan held its prime interest rate (Uncollateralized Call Rate) at zero (except for 1-2 basis points in transaction fees). However, the zero interest rate constraint initially led to an increase in real interest rates, owing to a worsening of deflationary expectations, which was only reversed after the Bank announced its commitment to maintain the zero interest rate policy until deflationary concerns were dispelled. Once the Bank of Japan deemed that deflationary pressures had abated it reversed course and raised its prime interest rate above the zero floor.12 There are, however, other means by which a central bank can conduct open market operations and inject liquidity into the system. For example, it can extend the scope of its money market operations along the yield curve, widen the range of private securities eligible for open market operations, and intervene in the foreign exchange market. In the case of Japan, however, these options were not pursued, mainly out of concern over their effect on the Bank of Japan’s balance sheet and the potential impact of a weakening in the yen on Japan’s trading partners.

16. Finally, price-level targeting could generate higher output volatility in the short run if there exists nominal price rigidities in the economy. Because the monetary authorities need to reverse price-level deviations from their target, monetary policy could generate more output fluctuations in the short run if nominal price adjustments are not rapid.13

Recent literature

17. New arguments in favor of price-level targeting claim that it can generate lower short-run inflation variability without affecting output volatility, while eliminating any potential inflation bias that might exist under inflation-targeting regimes.14 This result is obtained in a neoclassical framework where deviations of actual and expected prices are associated with a gap between actual and potential output (the output gap).15 The result holds if the output gap is persistent over time. As output-gap persistence dampens the effect of price changes in the short term, the monetary authorities would gain from lowering long-term price variability through price-level targeting. Output-gap persistence would reduce the tradeoff between low long-term price-level variability and high short-term inflation variability.

18. The reduced inflation-variability/output-variability result also holds in the context of a New-Keynesian framework where deviations of current and expected inflation are linked to the output gap due to partial price adjustments.16 In this case, the result depends on two crucial assumptions. First, monetary policy takes private sector expectations as given, thus restricting policymakers from manipulating expectations as to avoid time-consistency problems.17 Second, forward-looking inflation expectations rely on the existence of partial price adjustment processes stemming from market imperfections. Because people incorporate the price adjustments into the future, the monetary authorities can avoid short-term inflation variability by pinning down the price level in the long term.

19. These models, however, do not consider the costs associated with adjusting inflation expectations in a low and stable inflationary environment. Some recent literature has argued that people may prefer not to adjust their inflation expectations when the costs of inflation changes are minimal. In low inflationary environments, gains from incorporating inflation information may be so low that agents will choose to ignore it.18 In this context, the benefits of price stability under price-level targeting would also apply to inflation targeting when the inflation target is set at a sufficiently low rate.

20. The benefits of moving from inflation targeting to price-level targeting remain an empirical question on which there is little available evidence. While many developed countries have explicit or implicit inflation-targeting policy regimes, no country has used price-level targeting since the 1930s.19 Although economists have sought to use theoretical models to illustrate the potential magnitude of possible benefits from price-level targeting, these models are constrained by their specific assumptions that may bias the results. Moreover, these models do not address the cost implications of the transition from one policy regime to another.

List of References

  • Akerlof, George, William Dickens and George Perry, 2000, “Near-Rational Wage and Price Setting and the Optimal Rates of Inflation and Unemployment,Brookings Papers on Economic Activity, March.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Barnett, Richard and Merwan Engineer, 2000, “When is Price-Level Targeting a Good Idea?Bank of Canada Conference on Price Stability and the Long-Run Target for Monetary Policy, June.

    • Search Google Scholar
    • Export Citation
  • Beaudry, P. and M. Doyle, 2000, “What Happened to the Phillips Curve in the 1990s in Canada?Working Paper, presented at the Bank of Canada Conference, June 8, 2000.

    • Search Google Scholar
    • Export Citation
  • Crawford, Allan and Seamus Hogan,Downward Wage Rigidity,Bank of Canada Review, Winter 1998-99.

  • Crawford, Allan and Alan Harrison, 1997, “Testing for Downward Rigidity in Nominal Wage Rates,Bank of Canada’s conference on Price Stability, Inflation Targets and Monetary Policy, May.

    • Search Google Scholar
    • Export Citation
  • Debelle, G., 1997, “Inflation Targeting in Practice,IMF Working Paper No. WP/97/35, March.

  • Dittmar R. and W. Gavin, 2000, “What do New-Keynesian Phillips Curves Imply for Price Level Targeting?Federal Reserve Bank of St. Louis, March.

    • Search Google Scholar
    • Export Citation
  • Farès, Jean and Thomas Lemieux, 2000, “Downward Nominal Wage Rigidity: A Critical Assessment and Some New Evidence for Canada,Working Paper, presented at the Bank of Canada Conference, June 8, 2000.

    • Search Google Scholar
    • Export Citation
  • Faruqui, Umar. A., 2000, “Employment Effects of Nominal Wage Rigidity: An Examination Using Wage-Settlement Data,Bank of Canada Working Paper, July.

    • Search Google Scholar
    • Export Citation
  • Feldstein, M., 1997, “Costs and Benefits of Going from Low Inflation to Price Stability,” in Romer, C. and D. Romer, eds., Reducing Inflation: 123156, Chicago University Press.

    • Search Google Scholar
    • Export Citation
  • Fischer, S., 1994, “Modern Central Banking,” in Capie, F. et al, The Future of Central Banking, Cambridge University Press, Cambridge, United Kingdom.

    • Search Google Scholar
    • Export Citation
  • Fujiki, H., K. Okina and S. Shiratsuka, 2000, “Monetary Policy Under Zero Interest Rate,IMES Discussion Paper Series 2000-E-11, May.

    • Search Google Scholar
    • Export Citation
  • Jonung, L. and C. Berg, 1999, “Pioneering Price Level Targeting: The Swedish Experience 1931-1937Journal of Monetary Economics, 43, 525551.

    • Search Google Scholar
    • Export Citation
  • Lucas, R. E. (1972) “Expectations and the Neutrality of Money,Journal of Economic Theory 4 (April 1972): 103–124.

  • Maclean, Dinah and Hope Pioro, 2000, “Price Lever Targeting—The Role of CredibilityWorking Paper presented at the Bank of Canada Conference June 8, 2000.

    • Search Google Scholar
    • Export Citation
  • Masson P., M. Savastano and S. Sharma, 1997, “The Scope of Inflation Targeting in Developing Countries,IMF Working Paper No. WP/97/130, October.

    • Search Google Scholar
    • Export Citation
  • Mishkin, F., 1997, “The Causes and Propagation of Financial Instability: Lessons for Policymakers,Maintaining Financial Stability in a Global Economy, Federal Reserve Bank of Kansas City, Kansas City, MO: 55–96.

    • Search Google Scholar
    • Export Citation
  • Mishkin, Frederic S., 2000, “Issues in Inflation Targeting,Bank of Canada Conference on Price Stability and the Long-Run Target for Monetary Policy, June.

    • Search Google Scholar
    • Export Citation
  • Sarel, Michael, 1995, “NonLinear Effects of Inflation on Economic Growth,IMF Working Paper No. WP/95/56, May.

  • Shiratsuka, S., 2000, “Is There a Desirable Rate of Inflation? A Theoretical and Empirical Survey,IMES Discussion Paper No. 2000-E-32, October.

    • Search Google Scholar
    • Export Citation
  • Summers, L. H. (1991) “How Should Long-Term Monetary Policy Be Determined?Journal of Money, Credit and Banking, 23: 625–631.

  • Svensson, L.E.O., 1999, “Price-Level Targeting vs. Inflation Targeting: a Free Lunch,Journal of Money, Credit and Banking (31): 277–295.

  • Thiessen, Gordon G., 1998, “The Canadian Experience with Targets for Inflation Control,The Gibson Lecture, Queen’s University, Ontario, October.

    • Search Google Scholar
    • Export Citation
  • Thiessen, Gordon G.,Does Canada Need More Inflation to Grease the Wheels of the Economy?Bank of Canada Review, Winter 1996-97.

  • Walsh, Carl E., 2000, “Should Central Banks Stabilize Prices?Federal Bank of San Francisco, Economic Letter, No. 24, Research Department, August 11.

    • Search Google Scholar
    • Export Citation
  • Woodford, M., 1999, “Price-Level Determination Under Interest-Rate Rules,Princeton University, unpublished.

1

Prepared by Martin Kaufman and Rodolfo Luzio.

2

Sarel (1995) shows using panel data that there is a nonlinear effect of inflation on economic growth. When inflation is below a certain threshold, its effect on growth is null or slightly positive, while higher inflation has a negative, significant, and robust effect on growth. This is sometimes referred to in the economic literature as the “grease” versus “sand” effects of inflation.

3

Wage freezes alone are an incomplete test for downward rigidity because freezes can result from other sources, such as menu costs associated with changing wages and a lower variance of wage changes induced by a reduction in inflation uncertainty.

4

The composition effect refers to the changes in the composition of the workforce. This effect can bias up aggregate wage changes since rigidity only affects workers that stay with the same employer.

5

In practice, however, these hybrid alternatives may be very difficult to implement due, in part, to problems in trying to clearly explain the regime and to potentially less predictability in how the correction of deviations in inflation would be done.

6

See Feldstein (1997) for a detailed discussion on the role of expectations and price signals.

8

Mishkin (1997) discusses the relation between deflation and financial instability.

9

Under price level targeting, the commitment to revert deviations of the the price level away from the target would determine that any unexpected increase in real debt burdens would be temporary. However, debtors do suffer a loss, and the target-reversion process can be protracted exhacerbating liquidity problems.

10

Financial contracts such as indexed bonds, price-level contingent debt contracts, and option contracts efficiently reduce the cost associated with price uncertainty in the long run. See Fischer (1994).

11

See Summers (1991). This argument assumes that changes in short-term nominal interest rates affect real interest rates at least in the short run.

14

See Svensson (1999) in the context of a neoclassical framework, and Dittmar and Gavin (2000) in the context of a New-Keynesian model.

15

See Lucas’s island model (1972) or Fischer’s wage-contracting model (1977).

16

See Dittmar and Gavin (2000). The basic intuition is that given adjustment costs, forward-looking agents incorporate the effect of expected future price changes into current decisions. For instance, when prices are costly to change, firms raise current prices in response to an anticipated future aggregate demand expansion, thus affecting the current output level.

17

Intuitively, the time-consistency problem arises when the monetary authority can use current policy announcements to influence public expectations, but then deviate from their announced policy path in order to reap the benefits from such deviation.

18

See Akerlof et al. (2000). Federal Reserve Chairman Alan Greenspan has defined price stability as the level of inflation at which price changes no longer play a significant role in economic decision making, suggesting that price stability essentially may be associated with some low, positive inflation rate.

19

Sweden implemented price-level targeting from 1931 to 1937 (see Berg and Jonung (1999)). The Swedish price stabilization program resulted from the suspension of the international gold standard in 1931 and was viewed as a temporary solution before a return to the gold standard. Its initial goal was to arrest the ongoing deflation and “maintain the domestic purchasing power of the Swedish krona” (Jonung and Berg (1999), p. 535). In 1937, employment stabilization became the primary objective of Swedish policy authorities; the implementation of active countercyclical fiscal policies marked the end of the price level stabilization program.