Russian Federation
Selected Issues

Output has rebounded strongly and is now above the pre-crisis level in Russia. Inflation pressure stemming from the ruble depreciation has been reigned in and inflation has been brought down. The paper contains a detailed description of developments in the enlarged government budget, including the main determinants of revenue performance. It also describes the main elements of the recent tax reform and provides an overview of outstanding structural issues in the fiscal sector. The new government has proposed a set of far-reaching tax policy reforms.

Abstract

Output has rebounded strongly and is now above the pre-crisis level in Russia. Inflation pressure stemming from the ruble depreciation has been reigned in and inflation has been brought down. The paper contains a detailed description of developments in the enlarged government budget, including the main determinants of revenue performance. It also describes the main elements of the recent tax reform and provides an overview of outstanding structural issues in the fiscal sector. The new government has proposed a set of far-reaching tax policy reforms.

Russian Federation: Basic Data

article image
Sources: Russian authorities; and Fund staff estimates.

Data for 1998 or latest available.

Population declined to 145.2 million as of June 1, 2000.

Share of gross value added generated by sectors in factor prices to GDP in market prices.

Agriculture, including companies servicing agriculture and forestry.

1. This report provides background information for the discussion contained in SM/00/196, 8/23/00 of the main policy issues currently facing the authorities. The primary focus is on two key features of recent macroeconomic developments—the sustained recovery in output and the substantial improvement in the fiscal position—notably on the relative importance of the sharp ruble depreciation and the increase in international energy prices in accounting for these developments. A detailed analysis of the factors underlying the 1998 crisis and of macroeconomic developments immediately after the crisis was provided in SM/99/178,7/14/99.

2. Chapter I and Annex I describe the recovery in output and the main components of demand. A main conclusion is that the recovery in output after the 1998 crisis was originally driven primarily by import substitution, but that the recovery has now become much more broad based with robust increases in non-energy exports and the principal components of domestic demand. An important conclusion for the policy discussions is the finding that the ruble depreciation has been relatively more important than the increase in oil prices in explaining the strong output performance, although the impact of the latter is by no means negligible.

3. Chapter II and Annex II are concerned with the fiscal adjustment. While there has been a significant improvement in the overall balance at all levels of government, the adjustment at the federal level has been almost entirely accounted for by higher revenues, whereas adjustment at other levels has been mostly due to expenditure compression. The relative importance of discretionary policy changes, changes in the macroeconomic environment, and a residual item (which would include improved tax compliance) in explaining the strong increase in revenues at the federal level are analyzed. In this regard, the most important factor appears to be discretionaiy policy changes. Chapter n also includes a description of recent tax reforms as well as a discussion of the major outstanding structural fiscal problems, reflecting the priority assigned by the new government to fiscal reform.

4. Annexes m and IV provide a chronology of recent changes in the exchange and trade systems, respectively.

I. The Recovery in Output

A. Introduction

5. Output has rebounded strongly and is now above the pre-crisis level.1 Following a decline of almost 5 percent in real GDP during 1998, with a particularly sharp decline in the third quarter, output grew by over 3 percent in 1999. Estimates for the first quarter of 2000 suggest that the recovery has since gained further momentum. The output recovery was initially driven by import substitution in response to the large real depreciation. Subsequently, the recovery has become more broadly based as exports are growing and domestic demand, including both investment and, more recently, private consumption, is becoming more buoyant.

6. Inflationary pressure stemming from the ruble depreciation was quickly reigned in and inflation has been brought down to relatively low levels. Following an initial burst of inflation stemming from the depreciation, inflation was reduced by late 1999 to about percent per month on a seasonally adjusted basis.

7. Both the ruble depreciation and the increase in world energy prices have played important roles in the recovery. Breaking out the contribution of each of these factors is difficult but, the data suggest that the increase in world energy prices has, to date, played a secondary role in the recovery. In this regard, the bulk of the improvement in the external current account during 1999 came from a reduction in imports rather than an increase in energy exports.

8. The strong growth in output has resulted in a decline in unemployment. Despite the existence of important rigidities in the labor market, the unemployment rate fell to 11 percent at end-June 2000.

B. Developments Since the 1998 Crisis

9. Output began recovering in the last quarter of 1998 and has since gained further momentum. In the first quarter of 2000, real GDP stood 8.4 percent above its level one year earlier, almost 3 percent above its previous, end-1997 peak. Further, in a sharp break with pre-crisis experience, the current expansion involves most sectors and almost all regions— 84 out of 89 regions experienced growth in industrial output in 1999.

10. Any analysis of the dynamics and causes of this recovery is complicated by severe data limitations. Seasonal adjustment is extremely difficult, both because the data series are very short and unstable, and because there have been significant changes in the structure of the economy (see Box 1).

Issues In Seasonal Adjustment

Seasonal fluctuations in Russian real activity are unusual in both magnitude and timing, but the amplitude seems to be diminishing, Box Figure 1 (1 and 2). The seasonal pattern, with a very sharp fall in January, followed by strong growth in subsequent months and a peak in December, differs somewhat from the seasonal pattern normally found in countries with a similar climate and industrial structure. The Russian seasonal pattern partly reflects the orthodox holiday period that falls in January as well as old central planning behavior, in which the desire to fulfil the annual plan led to increased activity towards the end of the planning period. Central planning-based accounting and data-reporting practices artificially exacerbated the pattern; in particular, the output of smaller enterprises, which only reported once every year, was incorporated into the series on a cumulative year-to-date basis during the last period. Over the period 1995-99, the estimated seasonal January decrease in industrial output shrank from 7.8 to 6.7 percent, and the estimated seasonal first quarter decrease in real GDP shrank from 12.7 to 12 percent. This reflects both changes over time in the structure of the economy, and in particular the weakening influence of the old central planning mentality and improvements in statistical practices.

Seasonal adjustment in such a situation must allow the seasonal pattern to change over time and to take into account the impact of shocks, notably the August 1998 financial crisis. Allowing for changes in the seasonal pattern is only feasible if it can be assumed that the changes are sufficiently small and/or smooth. With short time series, a single isolated shock, such as the August 1998 financial crisis, can have a large and misleading impact on estimates of seasonal factors. Box Figure 1 (3) contrasts two seasonally adjusted (SA) time series for real GDP growth. In the first (“SA incl. crisis”), seasonal adjustment is achieved by passing the entire, unadjusted time series through the X-12 filter. In the second (“SA excl. crisis”), the crisis and post-crisis period (1998 Q3 onwards) is ignored when estimating the seasonal factors. The substantive difference is that, in the first time series, the output dynamics stemming from the August 1998 crisis and the post-crisis recovery influence the estimates of the normal seasonal movement. As a result, the procedure will tend to estimate higher SA growth rates for Q3 and lower SA growth rates for Ql, Q2, and Q4.

Controlling for the crisis, Box Figure 1 (4) shows SA growth for real GDP, an “output of bask sectors” index (covering about 60 percent of GDP), and industrial output. The figure suggests that, in 1999 Q1-Q2, growth was very fast, particularly for industrial output. In 1999 Q3, there was a sharp slow-down, although the real GDP and the output of basic sectors series give contrasting indications about the precise magnitude. In 1999 Q4, the growth rate rose again. Use of monthly data, see Box Figure 1 (5), confirms this pattern.

Using instead an ARIMA approach has both disadvantages and advantages. Estimating a simple ARIMA regression with seasonal dummies has two disadvantages: the estimates can be heavily influenced by outliers, and changes in the seasonal pattern cannot be accommodated. The approach does, however, permit the calculation of standard eirors of the estimated seasonal factors. Box Figure 1 (6) plots the upper and lower bounds of a 90 percent confidence interval for the estimated SA growth rate, using both the full sample and the non-crisis period alone to estimate the seasonal factors and their standard errors. There are two key conclusions, which reflect the feet that quarter-on-quarter changes in Russia are large and variable. First, the magnitude of the depicted confidence interval is extremely large; indeed, so large that in 1999 the null hypothesis of zero growth cannot (or can only barely) be rejected for any individual quarter. Second, the magnitude of the depicted confidence interval greatly overwhelms the differences implied by using different estimation methods, for example, including or excluding the crisis period, or using X-12 filters versus an econometric approach. Note though that the standard errors diminish dramatically over periods greater than an individual quarter. Indeed, by definition, the standard error is zero for the year as a whole. On the other hand, the confidence intervals would be even wider if account were taken of the (unknown) measurement error associated with the collection of the underlying, unadjusted data.

Box Figure 1.
Box Figure 1.

Russian Federation: Seasonal Factors in GDP and Industrial Production, 1996-2000

Citation: IMF Staff Country Reports 2000, 150; 10.5089/9781451833089.002.A001

Source: Goskomstat and Fund staff estimates.

In addition, key macroeconomic time series are unavailable, internally inconsistent, or of only limited reliability, especially at a quarterly frequency.2

11. The post-crisis recovery was initially led by import substitution, in response to the sharp real ruble depreciation. This first phase began in late 1998 and lasted until the middle of 1999. It was characterized by rapid growth of the external trade-oriented industrial sector, whose output increased on average by over 5 percent per quarter (Text Table 1, Figure 1). Import volumes, seasonally adjusted, fell by 50 percent in the second half of 1998 as a result of the depreciation-related substitution effect as well as an income effect originating from a sharp decline in real wages and consumption demand in the aftermath of the crisis. As for exports, energy export volumes remained broadly flat after the crisis, since the scope for expansion is limited by extraction and transportation constraints (see below). In contrast, non-energy exports (seasonally adjusted) declined by about 15 percent in dollar terms between the crisis and the second quarter of 1999, although customs data suggest that export volumes increased significantly, especially to non-CIS countries.3,4 The resultant dramatic increase in net foreign demand more than offset a decline in domestic demand, in particular in consumption, which fell sharply as real wages declined to three-fifths of their pre-crisis level (Figure 2).

Text Table 1.

Key Real Sector Growth Rates

(Seasonally adjusted, one-period growth)

article image
Sources: Goskomstat; and Fund staff estimates.
Figure 1.
Figure 1.

Russian Federation: Output and Income, 1995-2000

Citation: IMF Staff Country Reports 2000, 150; 10.5089/9781451833089.002.A001

Source: Goskomstat.
Figure 2.
Figure 2.

Russian Federation: Wages and Unit Labor Costs, 1992-2000

Citation: IMF Staff Country Reports 2000, 150; 10.5089/9781451833089.002.A001

Sources: Goskomstat; and IMF staff calculations.

12. The recovery appeared to be petering out in the second half of 1999. There was a significant slowdown in growth, with imports stabilizing while domestic consumption remained depressed. Investment rebounded strongly as the depreciation and rising oil prices improved the financial condition of enterprises, albeit from a very low base.

13. Output growth gained new and more broad-based momentum from late 1999. Output growth rates have returned to the levels observed in early 1999 and the expansion is now more broadly based, with a significant increase in domestic demand. In particular, private consumption has been growing sharply reflecting both a 20 percent increase in real wages since August 1999 and the near-elimination of wage arrears (Table 18), in turn linked to the strong financial situation of enterprises. Capital formation remains strong as rising international oil prices have further improved profits, and overall public confidence in economic prospects has strengthened in the wake of the recent elections. Three firms alone (Lukoil and Sifneft in the fuel sector, and Norilsk Nickel in the metal sector, which over the last six months has also experienced a major increase in export prices) have announced capital expenditure plans for 2000 that imply a (combined) increase in fixed capital formation of almost 1 percent of GDP. While such increases occur from a very low base, they do nonetheless represent a major break with the long-running trend decline.5

Table 1.

Russian Federation: Selected Indicators of Economic Activity, 1991-99

(Annual percentage change)

article image
Source: Goskomstat.

Plant growing.

Preliminary data.

Turnover of transport companies (including pipelines).

Table 2.

Russian Federation: GDP by Expenditure, 1991-99

article image
Source: Goskomstat and Fund staff estimates.

In last year’s comparable prices.

Table 3.

Russian Federation: GDP by Sector, 1991-99 1/

article image
Source: Goskomstat and Fund staff estimates.

Unit weight of gross added values generated by economic sectors in basis prices to GDP in basis prices unadjusted by indirectly measured financial intermediary services.

Agriculture, including companies servicing agriculture and forestry.

Transport, communications, road infrastructure.

Table 4.

Russian Federation: Gross Industrial Output by Sector, 1991-99

article image
Source: Goskomstat
Table 5.

Russian Federation: Employment, Labor Productivity and Real Wages in Industry by Sector, 1991-99

article image
Source: Goskomstat and Fund waff calculation.

Preliminary data.

The table contains the average payroll fund data.

Measured as the ratio of production to workforce.

Deflated by industrial PPI.

Measured as die ratio of real producer wages to average labor productivity.

Table 6.

Russian Federation: Labor Force Turnover, 1993-99 1/

article image
Sources: Goskomstat.

Data for large and medium enterprises.

Table 7.

Russian Federation: Employment by Sector, 1991 -99 1/

article image
Source: Goskomstat

Average for the year; does not include students.

Preliminary data.

Table 8.

Russian Federation: Indicators of Hidden Unemployment, 1993-99 1/

article image
Source: Goskomstat.

In industry, construction, transportation, communication, services, science, and scientific support.

For 1993, 1995-99 data include number of people on shortened workday at the end of each quarter; for 1994 data show those on shortened workdays over the course of the period.

Without pay or with partial pay.

Table 9.

Russian Federation: Selected Labor Market Indicators, 1992-99

article image
Source: Goskomstat.

Annual percentage change.

Table 10.

Russian Federation: Unemployment Rate by Regions (ILO methodology), 1993-99

(In percent of labor force)

article image
article image
Source: Goskomstat