This Selected Issues paper examines the potential costs of faster disinflation in Russia, drawing on the experience of European transition countries. The paper analyzes this experience, discusses factors contributing to the persistence of moderate inflation, and quantifies the disinflation costs in these countries. It compares the Russian economy with the sample countries. The paper concludes that a relatively rapid reduction of core inflation from above 10 percent in 2002 to less than 5 percent in 2004 would be beneficial for the Russian economy.

Abstract

This Selected Issues paper examines the potential costs of faster disinflation in Russia, drawing on the experience of European transition countries. The paper analyzes this experience, discusses factors contributing to the persistence of moderate inflation, and quantifies the disinflation costs in these countries. It compares the Russian economy with the sample countries. The paper concludes that a relatively rapid reduction of core inflation from above 10 percent in 2002 to less than 5 percent in 2004 would be beneficial for the Russian economy.

I. Speed of Disinflation and Output Costs in Russia: Implications from the Experience of European Transition Countries1

A. Summary

1. This paper examines the potential costs of faster disinflation in Russia, drawing on the experience of European transition countries (ETC).2 We study this experience, discuss the factors contributing to the persistence of moderate inflation, and quantify the disinflation costs in these countries. Then, we compare the Russian economy with the sample countries and draw conclusions about the sources of inflation and possible disinflation costs in Russia.

2. The main results of the paper are:

  • The Russian economy is similar to European transition countries in terms of nominal rigidities—a low degree of nominal wage indexation to past inflation and a qualitatively similar role of downward price rigidities in the inflation process;

  • A relatively rapid reduction of core inflation from above 10 percent in 2002 to less than 5 percent in 2004 would be beneficial for the Russian economy. It appears that the short-term output costs (less than 1 percent based on the experience of ETC) of faster disinflation are likely to be much smaller than the long-term gains from lower inflation.

B. Introduction

3. Over the last decade, inflation in Russia has declined substantially, but has settled at moderate levels of 15-20 percent over the past three years. The start of transition was followed by a surge in inflation to almost 900 percent in 1992. By 1997, inflation had been brought down to around 15 percent. After the 1998 crisis, year-on-year inflation jumped again to over 100 percent in mid-1999, but then dropped to around 20 percent in 2000, and has declined slowly since then, with inflation in 2002 at slightly above 15 percent.

4. Persistent moderate inflation is costly for the economy. Prolonged moderate inflation is associated with distortionary and allocative costs that might amount to several percentage points of GDP. Moderate inflation leads to:

  • Uncertainty about the future price level and relative prices impedes and distorts production decisions—Groshen and Schweitzer (1996) calculate that inflation of 10 percent (compared to 2 percent) for a prolonged period leads to a 2 percentage point reduction in the level of GDP;

  • Distortions to the tax system resulting from infrequent indexation for inflation. Higher taxation which results from inflation causes misallocation of capital, distorting the labor supply and leading to inappropriate corporate financing decisions—Fischer (1994) calculates that the costs from tax-related distortions associated with an inflation rate of 10 percent over a prolonged period amount to 2-3 percent of GDP;

  • Income and wealth redistributions that might result in higher income inequality. It is believed that economic agents that are able to adjust quickly their prices (entrepreneurs and businesses) gain from inflation at the expense of economic agents that are unable to do so (labor). Similarly, the poor find it more difficult to protect their financial savings from inflation. Although difficult to measure, the wealth redistributions resulting from even moderate inflation rates are likely to be economically and politically significant.

As a result, moderate inflation lowers the level of potential output and increases income inequality, which is a strong argument for ending it.

5. At the same time, too fast a disinflation from moderate levels might result in significant short-term output and unemployment costs. For example, Ghosh and Phillips (1998), Blanchard (1998) and Thierry and Griffiths (1998), find that disinflation from moderate inflation levels is costly. In particular, Ghosh and Phillips emphasize the nonlinearity in disinflation costs. They find a threshold of 10 percent for middle income countries below which disinflation is associated with output costs—the results suggest that severe disinflation (halving inflation) is associated with a fall in real GDP growth by 1 percentage point, while reducing inflation by 20-50 percent is associated with a decline in real GDP growth of 0.5 percentage points.

6. However, too slow a disinflation risks entrenching inflation expectations and could result in lasting moderate inflation. This is particularly important in Russia, where the authorities have targeted modest disinflation and followed an accommodative monetary policy. This resulted in persistent overshooting of the inflation targets and slower-than-originally planned disinflation over the last four years. The record of overshooting of the inflation targets, combined with the announced relatively unambitious disinflation path for the next two years, risks entrenching inflation expectations and the associated results discussed above.

7. We have chosen ETC as a comparator sample, because of the following similarities in their economies:

  • Substantial relative price adjustments. These countries, like Russia, faced substantial relative price adjustments during the disinflation process (see Figure 1.1).

  • Massive foreign exchange inflows. ETC experienced massive foreign exchange inflows against the background of underdeveloped monetary instruments, which prevented full sterilization by central banks. Although the nature of foreign exchange inflows in ETC was different from the one in Russia (capital account origin in ETC versus current account source in Russia) the authorities faced a similar dilemma in terms of monetary policy—either to accept higher inflation or some appreciation of the nominal exchange rate.

Figure 1.1.
Figure 1.1.

European Transition Countries: Relative Price Adjustment

(Index)

Citation: IMF Staff Country Reports 2003, 146; 10.5089/9781451833065.002.A001

C. Sources of Moderate Inflation in European Transition Countries

8. There is a vast empirical and theoretical literature on the sources of moderate inflation and the reasons for its persistence. In summary, the main findings in the literature suggest that the sources of moderate inflation are:

  • Excess money growth fueled either by substantial unsterilized foreign exchange inflows or by lax fiscal policy;

  • Relative price adjustments resulting from increases of administered prices combined with downward price rigidities;

  • Fast real wage growth that is in excess of productivity gains, combined with nominal wage rigidities.

9. Money growth was an important determinant of inflation in European transition countries. For instance, Coorey et al. (1997) find that inflation elasticity with respect to money growth is about 0.3 for ETC. Also, they find a relatively rapid inflation response to a monetary shock. Therefore, controlling money growth is important for disinflation.

10. Relative price adjustments can contribute to the persistence of moderate inflation, either because adjustment of natural monopoly prices to cost recovery levels or international prices causes inflationary pressures that are not of primary monetary origin, or because monetary policy accommodates these adjustments. Therefore, in a period of sustained relative price changes, inflation could be higher and sustained at double digits for several years.

11. Empirical evidence in the literature supports the contribution of relative price adjustments to inflation in ETC. In the early years of transition, inflation distributions displayed a high degree of variance, indicating significant price adjustments well beyond the ones suggested by price liberalization alone. In addition, there is supporting evidence that these distributions were positively skewed, in line with downward price rigidity and implying that a small number of large price increases led the inflationary process. For example, Coorey et al. find that the skewness of the relative price distribution has significant explanatory power for inflation in ETC. Similar results are obtained by Pujol and Griffiths for Poland, and Suranyi and Vinzce for Hungary.

12. There is mixed empirical evidence about the contribution of nominal wage indexation to inflation. For example, our estimates of wage indexation (see Table 1.1) show that only Poland had a very high degree of nominal wage indexation to lagged inflation (100 percent); nominal wage indexation in the Baltics and Hungary is below 60 percent, while nominal wages in the Czech Republic appear not to be indexed to lagged inflation.

Table 1.1.

European Transition Countries: Nominal Wage Indexation 1/

(Dependent variable quarterly wage growth)

article image

Estimates are reported in paranthesis

13. In sum, the evidence in the literature shows that monetary policy was the major determinant of inflation in ETC. At the same time, the empirical results in the literature show that relative price rigidity and increases of administered prices contributed to inflation, but they were not the main driving force of inflation. Finally, our empirical results suggest that with the exception of Poland, nominal wages had a minor contribution to inflation in the sample countries.

D. Disinflation and Output Cost: The Experience of European Transition Countries

14. The existing literature on disinflation suggests that, in order to achieve price stability with limited output and unemployment costs, three essential factors should be in place:

  • Credibility of monetary policy. This contributes to successful disinflation with limited output costs through its impact on inflation expectations. However, credibility can only be achieved if the commitment of the central bank to disinflation is supported by sound fiscal policies.

  • Flexibility of nominal wages. The structure of wage negotiations affects disinflation costs, even with a credible monetary policy. Other things being equal, the less frequent and the more spread-out the individual wage settlements, the harder it will be for wage growth to slow down in line with inflation, and consequently the higher output costs will be.

  • Flexibility of real wages. Although, in theory, disinflation requires an equal decrease in the growth of wages and prices, and not a decrease in the growth of real wages, in practice, part of the disinflation comes from a reduction of the real wage growth. In this case, the sensitivity of wages to labor market conditions becomes crucial for disinflation costs. If real wages are relatively sticky, because of backward-looking indexation, unemployment may be high for some time in order to achieve disinflation, thereby contributing to higher costs.

15. We applied three methods to study the output-inflation tradeoff and to quantify the disinflation costs in ETC: (i) single country inflation-output analysis (Figure 1.2), (ii) pooled sample inflation-output analysis (Figure 1.3), and, (iii) single country analysis accounting for trend output (Ball’s method given in Table 1.2). The results of the three methods are discussed below.

Figure 1.2.
Figure 1.2.

European Transition Countries: Inflation-Output Tradeoff

Citation: IMF Staff Country Reports 2003, 146; 10.5089/9781451833065.002.A001

Figure 1.3.
Figure 1.3.

Pooled Data: Inflation-Output Tradeoff

(1994-2001)

Citation: IMF Staff Country Reports 2003, 146; 10.5089/9781451833065.002.A001

Table 1.2.

European Transition Countries: Sacrifice Ratios

article image

16. The output-inflation tradeoff in ETC is non-linear. As shown on Figure 1.2, although the inflation-output pattern for each country is different, two stylized facts are worth mentioning. First, the reduction of inflation from 30-50 percent to about 10-15 percent was generally associated with an increase of growth; and second, further reduction of inflation from 10-15 percent to 3-5 percent was generally associated with some decline of output growth in most of the studied countries, although experience varied considerably.

17. However, the evidence in Figure 1.2 should be treated with caution, since the relationship between inflation and output might be spurious, due to exogenous shocks that affect both variables. For example, the degree of the slowdown in GDP growth in Latvia and real GDP decline in Estonia and Lithuania in 1998-99 could be significantly overstated by the 1998 Russian crisis—which had a negative impact on exports to Russia and the rest of the CIS countries and therefore a depressing effect on economic activity in the Baltics. Regarding the decline of output in the Czech Republic, it is overstated by two factors that coincided with the start of the disinflation at the beginning of 1998—first, structural problems in the enterprise sector; and second, significant cuts in government expenditures implemented to address mounting external current account imbalances.

18. To control for a possible spurious correlation between growth and inflation, we used pooled data. The estimated nonlinear relationship confirms the findings that inflation reduction from high double digits to around 15 percent is associated with higher growth, while further reduction of inflation below 15 percent leads to a small decline in growth. However, this decline in growth is only during the disinflation period (approximately 2 years on average). Thus, bringing inflation down to 5 percent from 15 percent lowers output growth by around 0.5 percentage point in the first year, and a cumulative reduction of less than 1 percent over two years.

19. We further quantify the cost of disinflation in the sample countries by calculating the output loss per percentage point reduction in inflation (sacrifice ratio). To calculate sacrifice ratios we use a modification of the methodology described in Ball (1994) (see Box 1.1).

Computation of Sacrifice Ratio For European Transition Countries

Methodology

The sacrifice raiio is computed as the ratio of the cumulative loss of output resulting from disinflation to the size of disinflation. Therefore, in order to compute the sacrifice ratios, we have lo identify disinflation episodes and measure output loss.

In the first step, we identify disinflationary episodes as those in which trend inflation declines substantially—by more than 5 percentage points. Trend inflation is defined as the centered four-quarter moving average of actual inflation. This definition smoothes actual inflation and allows us to separate policy induced declines in inflation from smaller fluctuations arising from exogenous shocks.

In the next step, we measure output loss as the deviation of actual output from trend output. As a measure of trend output, we use the Hodrick-Prescott filter, which eliminates transitory shocks.

Application to European transition countries

The above methodology was applied to European transition countries using quarterly data from the first quarter of 1993 lo the second quarter of 2002. All steps in computing the sacrifice ratio are provided En Table 1.2.

As Table 1.2 shows, the disinflation in European transition countries continued on average for slightly more than two years, at an annual speed of disinflation of approximately 35 percent.

20. Computed sacrifice ratios vary substantially across the sample countries. While the Czech Republic and Poland experienced costly disinflations (for every percentage point decline in inflation, the output loss was 0.5 percent and 0.2 percent correspondingly), Estonia, Hungary, and Lithuania achieved single digit inflation at zero cost, and Latvia gained from the disinflation approximately 0.5 percent.

21. To summarize, the various methods of computing disinflation costs in European transition countries suggest that disinflation from 10-15 percent to 3-5 percent is associated with relatively small output costs. The single-equation method does not offer conclusive evidence for disinflation costs, while the pooled data methods suggest that reducing inflation from 15 percent to 5 percent will reduce the level of GDP by cumulatively less than 1 percent over two years, and the method of Ball suggests that the same inflation reduction at an annual speed of 35 percent on average had a negligible effect on output (sec Table 1.2).

22. The disinflation costs in European transition countries are generally lower than in other transition economies. As discussed above, three conditions have to be met in order for the disinflation to proceed at low costs and it seems that all of them were in place in ETC during the disinflation process—namely:

  • Credible commitment by the authorities to a steady and fast disinflation. The fast decline of inflation in these countries could be regarded as a sign of credible monetary policy. For example, during 1995-98 inflation in the Baltics declined on average by nearly a half each year and was brought down to low single digits from the mid forties. Hungary and Poland also experienced fast disinflation—during the same period, inflation declined on average by one quarter each year. The fast disinflation created low inflation expectations and further facilitated the disinflation process;

  • Relatively low nominal wage indexation to past inflation, ensuring fast adjustment of nominal wages in line with inflation;

  • Flexible labor markets. It appears that the implemented structural reforms helped create flexible labor markets that facilitated real wage adjustments associated with the disinflation process, without increasing unemployment and therefore output costs.

E. Russian Economy Versus European Transition Economies: How Fast to Disinflate?

23. In this section we analyze the determinants of inflation in Russia and compare the results with our findings for European transition countries.

The impact of exchange rate and nominal rigidities on inflation

24. The policy of continuous nominal exchange rate depreciation has contributed significantly to inflation in Russia. Empirical estimates of the pass-through suggest that a one percentage point change in the nominal exchange rate leads to roughly ½ percentage point change in inflation (Table 1.3).3 This pass-through effect is relatively high compared to Asian developing countries, but is of a similar magnitude to that of Turkey (see Leigh and Rossi 2002).

Table 1.3.

Russia: Pass-Through Effect

article image

25. The distribution of relative prices in Russia is positively skewed, suggesting downward price rigidity. In order to quantify the degree of relative price rigidity in Russia, we studied the statistical properties of the relative price distribution of 30 items from the CPI basket (we used data for eight months of 2001 and 2002). As reported in Figure 1.4, the relative price distribution is positively skewed, suggesting downward price rigidity in the economy. When prices are downward rigid, large increases in the prices of a few goods are accompanied by small price increases (but few price declines) in the rest of the basket resulting in a positive skewness of the distribution.

Figure 1.4.
Figure 1.4.

Russia: Relative Price Distribution

Citation: IMF Staff Country Reports 2003, 146; 10.5089/9781451833065.002.A001

26. Although downward price rigidity contributed to inflation, monetary policy was the main driving force, as in ETC. As reported in Table 1.4, downward price rigidity (captured by the positive coefficient of skewness of the relative price distribution) has a significant role in explaining Russian inflation. However, the nominal exchange rate and the money supply had a much larger impact. For example, the standardized regression coefficients show that an increase in the skewness of the relative price distribution (equivalent to faster increases in regulated prices) by one standard deviation leads to an increase in inflation by 0.06 standard deviations, while changes in the nominal exchange rate and money supply by one standard deviation result in 0.95 and 0.19 standard deviations changes in inflation respectively. This result suggests that persistent moderate inflation over the last couple of years was a result not only of increases in regulated prices, but also of continued nominal exchange rate depreciation and relatively fast growth of money supply.

Table 1.4.

Russia: Contribution of Monetary Policy and Relative Price Variability to Inflation

article image

27. Nominal wages show lower inflation indexation than in most European transition countries. Nominal wages in Russia are only indexed by a quarter to past inflation while in most of European transition countries the wage indexation is much higher, as shown above. Even accounting tor the autoregressive coefficient, the long-run impact of wage indexation in Russia is not higher than the average in the sample countries and is much lower than in Poland.

Table 1.5.

Russia: Contribution

(Dependent variable: quarterly wage growth)

article image

28. The conclusion is that the Russian economy compares favorably to European transition countries in terms of nominal rigidities. As discussed above, nominal wage indexation is somewhat lower in Russia than in most of the sample countries (25 percent in Russia versus more than 55 percent on average in the sample). Also, our analysis shows that downward price rigidity in Russia has a similar qualitative role in the inflation process as in European transition countries—administered price increases contributed to higher overall inflation as a result of downward price rigidity. However, as discussed in other studies and as implied by the lower sacrifice ratios, downward price rigidities appear to have made a minor contribution to disinflation costs in the sample countries.

The risk of entrenching inflation expectations and inflation costs

29. The Russian economy differs substantially from the ETC economies in respect of the commitment of the authorities to faster disinflation.4 With the background of sizable foreign currency inflows, the CBR subordinated its inflation targets to the goal of limiting the pace of real appreciation by accommodating inflationary shocks. This policy resulted in overshooting of the inflation targets over the last four years and, as a result, slow actual disinflation—the disinflation speed since mid-1999 was on average around 15 percent, which is much lower than the average disinflation speed in ETC (35 percent). The history of accommodative monetary policy, combined with somewhat unambitious inflation targets over the next two years, risks entrenching inflation expectations.

Figure 1.4.
Figure 1.4.

Russia: Actual Inflation and Inflation targets

(In precent, 12-month change)

Citation: IMF Staff Country Reports 2003, 146; 10.5089/9781451833065.002.A001

The case for faster disinflation

30. To reduce inflation expectations, the authorities need to aim for a faster disinflation over the next two years. The authorities’ inflation targets for the next two years imply, however, that overall inflation declines at an annual speed of approximately 20 percent, slower than achieved in ETC. At the same time, net inflation (excluding the impact of regulated prices)5 declines at an even slower annual speed—under 20 percent a year. Accounting for the low nominal rigidities in Russia and based on the experience of European transition countries, the authorities could reduce net inflation6 at a speed of 30-35 percent a year, with relatively small output cost.

Table 1.6.

Russia: Regulated Prices and Speed of Disinflation

article image

For projection, authorities’ inflation target wd increase unregulated prues are assumed.

F. Conclusions

31. Our analysis suggests that a relatively rapid reduction of core inflation from above 10 percent in 2002 to less than 5 percent in 2004 would be beneficial, since the short-term output costs are likely to be much smaller than the long-term gains from lower inflation. Given the authorities’ planned increases of regulated prices over the next two years, a faster-than-planned reduction in headline inflation could be achieved at a short-term output cost of less than 1 percent. At the same time, opting for a slower disinflation might entrench inflation expectations and lead to persistent moderate inflation over the medium term, which could inflict substantial output losses.

References

  • Ball, Laurence, 1994, “What Determines the Sacrifice Ratio?in Monetary Policy, ed. by N.G. Mankiw (Chicago: University of Chicago Press).

    • Search Google Scholar
    • Export Citation
  • Blanchard, Oliver, 1998, “Optimal Speed of Disinflation: Hungary,” in Carlo Cottarelli and György Szapáry (eds.), Moderate Inflation: The Experience of Transition Economies, International Monetary Fund and National Bank of Hungary, Washington D.C.

    • Search Google Scholar
    • Export Citation
  • Christoffersen, Peter, and Peter Doyle, 1998, “From Inflation to Growth: Eight Years of Transition,” IMF Working Paper 98/100 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Dornbush, Rudiger, and Stanley Fischer, 1993, “Moderate InflationWorld Bank Economic Review, Vol. 7 (January), pp. 144.

  • Fischer, Stanley, 1994, “Modern Central Banking,” in Forest Capie, Charles Goodhart, Stanley Fischer and Norbert Schnadt, eds. The Future of Central Banking: The Tercentenary Symposium of the Bank of England. Cambridge University Press: Cambridge, England and New York: 262308.

    • Search Google Scholar
    • Export Citation
  • Groshen, Erica L., and Mark E. Schweitzer, 1996, “The Effects of Inflation on Wage Adjustment in Frim-Level Data: Grease or Sand?,” Federal Reserve Bank of New York Staff Reports, January 1996, No. 9.

    • Search Google Scholar
    • Export Citation
  • Leigh, Daniel, and Marco Rossi, 2002, “Exchange Rate Pass-Through in Turkey,” IMF Working Paper 02/204 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Thierry Pujol, and Mark Griffiths, 1998, “Moderate Inflation in Poland: A Real Story,” in Carlo Cottarelli and Gyorgy Szapary (eds.), Moderate Inflation: The Experience of Transition Economies, International Monetary Fund and National Bank of Hungary, Washington D.C.

    • Search Google Scholar
    • Export Citation
1

Prepared by Emii Stavrev (EU2).

2

The following countries are included in the sample: Poland, Hungary, the Czech Republic, and the Baltics (Lithuania, Latvia, and Estonia). We have chosen these countries as a comparator sample instead of considering a larger sample including developing countries as in other studies, because we believe that the experience of European transition countries is more relevant to Russia than the experience of the other countries.

3

See Chapter II for a fuller discussion.

4

For a detailed discussion of the CBR’s monetary policy objectives, see Chapter IV.

5

Several factors limited the accuracy of our calculations of the impact of regulated prices on total inflation both historically and for the projection period: (i) the lack of a precise definition of administered prices in the CPI basket and their weights; (ii) the impact of regulated wholesale (PPI) prices for gas, electricity, and freight transportation on the CPI; (iii) uncertainties with administered prices increases for 2003 and 2004 (administered prices increases for 2003 are included in the government program, but it is uncertain to what extent the authorities will stick to the program; administered prices increases for 2004 are only roughly outlined in the medium term government program accepted in July 2002 and are preliminary—the data are provided by the Economic Expert Group).

6

The implied paths for the overall and net inflation are given in Table 1.6.

Russian Federation: Selected Issues
Author: International Monetary Fund