## _wp07164

## Source details

**Canonical URL:** [_wp07164](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07164.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07164.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07164.pdf.json)

---

### Study purpose, scope, and definitions
- Purpose: Analyze the sources of the recent rapid growth in the Commonwealth of Independent States (CIS) and the prospects for its continuation.
- CIS definition for this paper: Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Mongolia, Russia, Tajikistan, Ukraine, and Uzbekistan (Turkmenistan excluded; Mongolia included).
- Key contextual observations:
  - Unweighted average growth of the CIS and the three Baltics countries has been above that of most other regions in the past six years.
  - As of end-2006 real GDPs of Moldova, Georgia, Tajikistan, Ukraine, and Kyrgyzstan were still well below their 1990 levels.
  - By the time output had bottomed, it had fallen by more than 50 percent in Armenia, Azerbaijan, Georgia, Moldova, and Tajikistan.
- Recovery projection examples (conditional on steady growth of about eight percent a year):
  - Ukraine: regain 1990 real GDP level by 2010.
  - Georgia and Tajikistan: regain 1990 real GDP level by 2012.
  - Moldova: regain 1990 real GDP level by 2015.
- Investment and employment context:
  - Sustaining growth of at least six percent a year typically requires investment more than 25 percent of GDP (examples: China, South Korea, Malaysia, Thailand, and Vietnam).
  - Investment outlays for the CIS, excluding Azerbaijan and Mongolia, averaged about 20 percent of GDP in 1996–2006.
  - Total employment for the region as a whole at end-2006 was about 10 percent below its 1990 level.

### Overview of output development (2001–06)
- Regional growth and demand:
  - Real GDP growth for the CIS region: unweighted average of eight percent a year during 2001–06.
  - Real private consumption: grew by about nine percent per year for the region as a whole.
  - Contribution of net exports: negative in some CIS countries and marginally positive in others due to large import growth.
- Investment patterns and gaps:
  - Real investment: double-digits growth in several CIS countries, but investment-to-GDP ratio remained relatively low.
  - Investment-to-GDP comparisons:
    - CIS average about five percentage points below the simple average for the three Baltics.
    - CIS average about three percentage points below the five Central European economies.
    - CIS average about 10 percentage points below fast-growing East Asian countries.
  - Investment concentrated in oil, gas, and metallurgy; FDI outside commodity sectors remained low.
- Country-specific drivers:
  - Russia: eight years (1999–2006) of strong growth; private consumption driven by higher oil and gas incomes; real fixed investment grew by an annual average rate of nine percent.
  - Ukraine: strong growth in 2000–06 from structural reforms and positive terms-of-trade shocks; FDI surged to an estimated US$10 billion in 2005–06.
  - Azerbaijan and Kazakhstan: FDI-led reconstruction of energy sector central; Baku-Tbilisi-Ceyhan pipeline capacity of one million barrels a day; Azerbaijan’s annual growth rate in 2005–06 rose to 27 percent; Kazakhstan averaged about 10 percent a year in 2001–06.
  - Armenia, Georgia, Moldova, Tajikistan: growth driven by private consumption and small-scale private investment; construction and services dominated.
  - Kyrgyzstan: lower growth due to 2002 Kumar gold mine accident and 2005 political upheaval.
  - Belarus: about seven percent a year average growth over 10 years.
  - Turkmenistan: official figures show real GDP growth averaged more than 15 percent in 2000–06 (likely upward biased).
  - Uzbekistan: average growth of five percent in 2001–06 driven by gold, cotton, and natural gas exports; extensive state intervention remains a constraint.

### Methodology and measurement issues
- Growth accounting framework:
  - Production function: Y = e^θ K^α L^β where θ is TFP growth, α is capital share, β is labor share.
  - Decomposition: g = θ + α (growth rate of capital) + β (growth rate of labor).
- Capital measurement:
  - Capital stock via perpetual inventory method (PIM); initial capital often estimated using Harberger (1978) approach: K_{t-1} = I_t / (g + δ).
  - One-time adjustment: capital stock for CIS countries reduced by the same rate as output between 1990 and 1995 so capital-output ratio is not allowed to rise during sharp contraction.
- Factor shares and estimation strategy:
  - Capital share estimated econometrically using fixed effects, 2SLS, and cointegration techniques; estimation regional rather than country-specific.
  - Instruments for 2SLS: lagged capital and labor growth rates.
- Data constraints and measurement caveats:
  - New database compiled from IMF WEO, ILO, UNECE and other sources; data quality issues due to underreporting and legacy statistics focused on public enterprises.
  - Tajikistan, Turkmenistan, and Uzbekistan: incomplete and not always good-quality data.
  - Possible underestimation of capital accumulation if a significant portion of communist capital stock was permanently scrapped.
  - Labor and capital quality adjustments largely unavailable—potential overestimation of TFP.
  - Capacity utilization data limited (available for Russia only): U-shaped capital utilization falling until 1996 and rising from 1999; labor utilization rose from around 70 percent during 1994–98 to around 85 percent during 2000–04. Russia’s utilization used as proxy for other CIS countries.
  - Education and hours-worked data limited; relatively high secondary school attainment across transition economies reduces expected cross-country variation.

### Estimated results (1996–2006) — regional TFP and factor shares
- Fixed effects and 2SLS estimates (1996−2006):
  - CIS-12:
    - TFP = 2.3 percent
    - Capital elasticity (α) = 0.63
    - Labor elasticity (β) = 0.51
    - R^2 = 0.62
    - Number of observations = 132
  - Baltics-3:
    - TFP = 3.0 percent
    - Capital elasticity (α) = 0.49
    - Labor elasticity (β) = 0.52
    - R^2 = 0.66
    - Number of observations = 33
  - CE-5:
    - TFP = 1.6 percent
    - Capital elasticity (α) = 0.40
    - Labor elasticity (β) = 0.62
    - R^2 = 0.66
    - Number of observations = 51
  - SEE-6:
    - TFP = 0.8 percent
    - Capital elasticity (α) = 0.78
    - Labor elasticity (β) = 0.29
    - R^2 = 0.63
    - Number of observations = 60
- Alternative fixed-effects (cross-section weights) highlights:
  - CIS-12: TFP = 2.6 percent; Capital elasticity = 0.66; Labor elasticity = 0.53; R^2 = 0.59; Observations = 131.
  - Baltics-3: TFP = 3.0 percent; Capital elasticity = 0.49; Labor elasticity = 0.52; R^2 = 0.66; Observations = 33.
  - CE-5: TFP = 1.6 percent; Capital elasticity = 0.40; Labor elasticity = 0.62; R^2 = 0.66; Observations = 55.
  - SEE-6: TFP = 0.7 percent; Capital elasticity = 0.71; Labor elasticity = 0.34; R^2 = 0.62; Observations = 60.
- Interpretation:
  - TFP growth 1996−2006: Baltics 3.0 percent > CIS 2.3 percent > CE 1.6 percent.
  - For CIS sample: capital elasticity ≈ 0.63 and labor elasticity ≈ 0.51 (fixed effects).
  - Sum of capital and labor elasticities close to unity for Baltics and CE; slightly greater than one for CIS.
  - 2SLS estimates nearly identical to fixed effects, addressing endogeneity concerns.
  - Literature typically finds capital elasticity of 0.3 to 0.5 in industrial countries; shares of physical capital likely higher in developing countries.

### Box 1 — Labor Market: trends, measurement, and links to growth
- Employment trends:
  - Employment declined by about 20 percent from 1990 to 1997 for the region as a whole.
  - As of 2006, the region still lost slightly more than 10 percent of its employment relative to 1990.
  - Employment increases after 1998 were modest in Russia, Ukraine, and Central Asia.
- Public vs. private sector:
  - Private sector by 2006 generally accounted for between 60−80 percent of total employment.
  - Reduction in public sector employment was not offset by private sector increases except in Azerbaijan, Kyrgyzstan, Mongolia, Tajikistan, and Uzbekistan.
- Demographics and human capital:
  - Population and labor force growth (1995–2006): 1-2 percent a year increase in Uzbekistan, Tajikistan, Kyrgyzstan, and Azerbaijan; declines of 1 to 1½ percent a year in Armenia, Georgia, Moldova, Russia, and Ukraine.
  - Secondary school enrollment rates remain much higher than in most low- and middle-income developing countries.
- Informal sector and unemployment measurement:
  - ILO-style registered unemployment rates low, but labor force survey unemployment:
    - 9 percent in Kazakhstan, Russia, and Ukraine
    - 20 percent in Armenia, Georgia, and Moldova
  - Informal sector employment about one-half of all employment in the CIS (compared with about one-quarter in CE economies).
- Links to growth accounting:
  - For 1996–2006 the most important source of growth has been capital accumulation for most countries, except Armenia, Tajikistan, and Ukraine where TFP contribution was higher.
  - Labor contribution low given sharp employment decline in the 1990s; improvements in labor quality reflected in TFP.
  - Cross-country variation in TFP (1996–2006):
    - Armenia and Azerbaijan: above 4 percentage points
    - Georgia and Belarus: between 2−3 percentage points
    - Moldova, Russia, and Ukraine: from 1 to 2 percentage points
    - Uzbekistan: TFP growth less than one percentage point
    - Mongolia: TFP growth negative
  - Drivers of high TFP growth include managerial improvements, labor shedding, inter-industry reallocation, low initial technology endowment, and increases in capacity utilization.

### Key quantitative findings on TFP and investment (CIS-12, 2001–2006 — Table 6 scenarios)
- Aggregate changes and Scenario results:
  - Average ratio of investment to GDP for the CIS increased by about four percentage points from 1996–2000 to 2001–06.
  - With capacity utilization adjustment (Scenario B), average TFP growth in 2001–06 fell from 3.8 to 2.9 percentage points.
- CIS-12, 2001–2006 (Table 6, Scenario A / Scenario B):
  - Real GDP growth: 8.2
  - Investment share of GDP: 24.8 (In percent)
  - Growth of GDP: 6.7 (In percent, annual average)
  - Capital growth: 0.9 (In percent)
  - Labor growth: 7.5 (In percent)
  - Productivity: 4.0 (In percent)
  - Contribution of Capital: 0.4 (percentage points)
  - TFP (Scenario A): 3.8 (percentage points)
  - TFP (Scenario B): 2.9 (percentage points)

### Historical perspective, dynamics, and interpretation
- Comparative benchmarks:
  - Postwar Western Europe and Japan: average contribution of TFP to output growth 2.7 percentage points for seven major industrial countries—close to estimated CIS TFP contribution.
  - U.S. per capita income growth 1950–1973: 2.4 percent a year; Germany: 5 percent; Japan: slightly more than 8 percent.
  - East Asia (“four tigers”): physical capital accumulation major contributor; Young (1995) and Shigeru, Khan, and Murao (2003) estimates highlight a strong role for capital in some periods and substantial TFP contributions in others.
  - China: TFP growth averaged about 4 percentage points over two decades, contributing to 41 percent of output growth.
- Technical efficiency decomposition:
  - TFPi(t) = Ft + ûi(t): TFP change = technological progress (best-practice frontier) + change in technical efficiency.
  - Rapid positive technical efficiency change (movement toward best practice) can generate high measured TFP growth, consistent with CIS experience.
  - The paper assumes technological change constant across countries; variance in TFP change then derives from variance in technical efficiency change.
- Measurement caveats:
  - Factor input quality improvements (capital embodied technology, labor education/on-the-job training) may be measured as TFP rather than adjusted as capital or labor quality.
  - Mis-measurement of TFP could be significant following move from central planning to market economies.

### Dynamics, limits, and medium-term prospects
- Transitory vs. sustained sources:
  - Much of recent CIS productivity growth reflects reallocation, better use of investment, capacity utilization increases, elimination of inefficiency, and higher work intensity—factors that are largely transitory.
  - Such sources can have substantial impact over one or two decades but unlikely to produce indefinite high productivity growth.
- Catch-up potential and limits:
  - Some CIS economies (Moldova, Georgia, Tajikistan, Ukraine) still have substantial catch-up potential given low real GDP bases relative to 1990.
  - As countries approach the world technology frontier, TFP gains from eliminating central-planning inefficiencies will diminish; alternative channels will be needed.

### Central conclusions and policy implications
- Central conclusions:
  - Estimated TFP growth for the former Soviet Union republics was higher than for other fast-growing economies in the sample.
  - Capital accumulation in most CIS countries made modest contributions to growth—on average much smaller than in the three Baltics and the five Central European economies.
  - Investment outlays remained relatively low except in Azerbaijan and Mongolia.
  - Inefficiencies inherited from central planning left scope for managerial improvements, labor shedding, and inter-industry reallocation that supported rapid growth with low investment rates.
- Policy implications to sustain growth:
  - Further improvement in market reforms and institutions to support productivity as catch-up opportunities narrow.
  - Greater labor use and continued faster capital accumulation expected to play a more important role in medium-term growth if TFP catch-up opportunities are exhausted.
  - Prepare for the transitory nature of some TFP gains (reallocation, capacity utilization, intensity) by promoting sustained technological progress and efficiency-enhancing structural reforms.

*Source: _wp07164 — References and Section V summary (excerpts).*

### References .............................................................................................................

### References

### Figures and Tables (inventory)
- Figures listed include:
  - Real GDP Index in 2006
  - Real Private Consumption, 1996–2006
  - Investment, 1996–2006
  - Sources of Growth, Transition Countries, 1996–2006
  - Production Frontier
  - Technical Efficiency, 1996–2006
  - Output Profile, 1991–2006
  - Employment, 1991–2006
- Tables listed include:
  - Output Decline and Recovery
  - Regional Estimates of the Shares of Inputs
  - Growth Accounting Results for Transition Economies
  - Sensitivity of TFP Growth Estimates, 1996–2006
  - Sources of Growth, Historical Perspective, 1950–2006
  - Growth-Accounting Results for Transition Economies, 1991–2006
  - Real GDP Growth, 1991–2006
  - Gross Fixed Capital Formation, 1991–2006
  - Employment, 1990–2006

### Acronyms and country/region definitions
- Country codes provided (selected examples as listed): ALB (Albania), ARM (Armenia), AZE (Azerbaijan), BEL (Belarus), BGR (Bulgaria), BIH (Bosnia and Herzegovina), CZE (Czech Republic), EST (Estonia), GEO (Georgia), HRV (Croatia), KAZ (Kazakhstan), KGZ (Kyrgyzstan), MDA (Moldova), MAC (FYR Macedonia), MON (Mongolia), LTU (Lithuania), LVA (Latvia), POL (Poland), ROM (Romania), RUS (Russia), SLK (Slovakia), SLN (Slovenia), TAJ (Tajikistan), TUR (Turkey), HUN (Hungary), UKR (Ukraine), UZB (Uzbekistan).
- Regional groupings defined:
  - Baltics: Estonia, Latvia, and Lithuania
  - CE: Central Europe (Czech Republic, Poland, Hungary, Slovakia, and Slovenia)
  - CIS: Commonwealth of Independent States (Armenia, Azerbaijan, Belarus, Georgia, Moldova, Mongolia, Kyrgyzstan, Kazakhstan, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan)
  - FSU: Former Soviet Union Republics
  - SEE: Southeast Europe (Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Macedonia, and Romania)
- Other acronyms: CEA, GDP, EBRD, FDI, GKS, ILO, OECD, REB, TFP, UNCTAD, UNECE, WEO

### Study purpose, scope, and definitions (I. INTRODUCTION, PURPOSE, AND SCOPE OF STUDY)
- Purpose: Analyze the sources of the recent rapid growth in the Commonwealth of Independent States (CIS) and the prospects for its continuation.
- CIS definition for this paper: Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Mongolia, Russia, Tajikistan, Ukraine, and Uzbekistan. Turkmenistan excluded due to poor quality of national accounts statistics. Mongolia included though not a CIS member.
- Historical context and key observations:
  - Unweighted average growth of the CIS and the three Baltics countries has been above that of most other regions in the past six years.
  - As of end-2006 real GDPs of Moldova, Georgia, Tajikistan, Ukraine, and Kyrgyzstan were still well below their 1990 levels.
  - By the time output had bottomed, it had fallen by more than 50 percent in Armenia, Azerbaijan, Georgia, Moldova, and Tajikistan.
  - Example comparisons:
    - In the United States during the Great Depression, output per capita fell by 31 percent and recovered to its pre-crisis level in 10 years.
    - Output fall from pre-World War II (1938) to postwar trough was 51 percent in West Germany and 45 percent in Japan; both regained their 1938 level of output by 1953—eight years after the end of the war.
- Specific recovery projections cited:
  - It would take until 2010 for Ukraine, 2012 for Georgia and Tajikistan, and 2015 for Moldova to regain the 1990 real GDP level assuming these countries manage to grow steadily at about eight percent a year (close to their annual average for the past five years).
- Investment and employment context:
  - Sustaining economic growth of at least six percent a year for a long period typically requires high investment (more than 25 percent of GDP) (examples: China, South Korea, Malaysia, Thailand, and Vietnam).
  - Investment outlays for the CIS, excluding Azerbaijan and Mongolia, averaged about 20 percent of GDP in 1996–2006.
  - Total employment for the region as a whole at end-2006 was about 10 percent below its 1990 level.

### Research questions and methodology
- Research questions:
  - What explains the strong economic recovery in the CIS?
  - Is either investment or TFP growth responsible for the major shifts in economic growth?
  - What can be learned from the experiences of other countries that sustained rapid growth for a long time?
- Methodological approach:
  - Uses the growth-accounting approach to analyze sources of recent growth in transition economies and compare with other fast growing economies.
  - Data for transition economies cover 1991−2006.
  - Capital share in income is estimated econometrically (fixed effects, 2SLS, and cointegration techniques) rather than assumed arbitrarily.
  - Estimation is regional rather than country-specific due to the short time span of reliable and comparable national accounts statistics.
  - Sensitivity analysis performed under different assumptions of capital shares and initial capital output ratios.

### Main contributions and key findings
- Contributions of the paper:
  - Examines sources of recent growth for all transition economies and compares them with fast growing economies in East Asia and post-war Western Europe and Japan.
  - Adopts up-to-date econometric techniques to estimate capital shares in income.
  - Provides sensitivity analysis under different capital-share and initial capital-output ratio assumptions.
- Main growth-accounting finding:
  - The CIS’s recent strong growth has been driven largely by growth in TFP.
  - On average, capital accumulation made a modest contribution.
  - Employment rates continued to drop until recently in some countries in the region.
- Policy-relevant implication:
  - Critical question: whether the rapid TFP growth can be sustained.
  - Assuming that TFP growth slows down, other sources of growth (notably faster capital accumulation and improvement in employment) will be essential to sustain a rapid catch-up.
  - Recent trend of faster capital accumulation and improvements in employment in some CIS countries are expected to play a more important role in future growth.

*Source: _wp07164 - References (content excerpt)._

### Section V summarizes the findings and draws conclusions.

### _wp07164 - Section V summarizes the findings and draws conclusions.

### Overview of output development (2001–06)
- Real GDP growth for the CIS region picked up strongly to an unweighted average of eight percent a year during 2001–06.
- The rapid growth was driven by domestic demand:
  - Real private consumption grew by about nine percent per year for the region as a whole, underpinned by large hikes in real wages.
  - In some CIS countries, substantial increases in remittances also supported private consumption.
- External sector developments:
  - Benign global environment and higher commodity prices encouraged a rapid increase in exports.
  - The sharp increase in consumption led to an even higher growth in imports; net exports contributed negatively in some CIS countries and marginally positively in others.
- Investment patterns:
  - Real investment grew by double-digits in several CIS countries, but the investment-to-GDP ratio remained relatively low.
  - The average investment-to-GDP ratio for the CIS was about five percentage points below the simple average for the three Baltics, three percentage points below the five Central European economies, and some 10 percentage points below the fast-growing countries in East Asia.
  - Investment remained low outside oil, gas, and metallurgy; FDI inflows outside commodity sectors remained low.

- Country-specific drivers:
  - Russia: eight years (1999–2006) of strong growth; post-2000 expansion driven mainly by private consumption fuelled by rising incomes from higher oil and gas prices; real fixed investment grew by an annual average rate of nine percent, but investment-to-GDP remained relatively low, particularly in non-energy sectors.
  - Ukraine: strong growth in 2000–06 due to structural reforms (2000–01) and positive terms-of-trade shocks; export-led upturn in steel and chemicals; FDI surged since 2004 to an estimated US$10 billion in 2005–06 (more than cumulative total of all preceding years), largely driven by privatization of Kryvorizhstal and a handful of bank deals; other sectors remained largely untouched by foreign capital due to bureaucratic interference and oligarch influence.
  - Azerbaijan and Kazakhstan: FDI-led reconstruction of the energy sector was important; completion of the Baku-Tbilisi-Ceyhan oil pipeline (capacity of one million barrels a day) and extraction from the Azeri-Chriag-Guneshi oilfield raised Azerbaijan’s annual growth rate in 2005–06 to 27 percent; Kazakhstan averaged about 10 percent a year in 2001–06.
  - Armenia, Georgia, Moldova, Tajikistan: growth driven by private consumption and small-scale private investment; construction and services accounted for most growth; high remittances and real wage growth fueled consumption; Georgia benefited as a transit corridor and from pipeline-related construction demand.
  - Kyrgyzstan: much lower growth due to the 2002 Kumar gold mine accident and political upheaval of 2005 (Tulip Revolution).
  - Belarus: 10 years of uninterrupted strong growth averaging about seven percent a year; growth from a relatively high initial base; benefited from importing energy from Russia at below-market prices and rising prices for energy-intensive exports; gains used to raise wages, subsidize state enterprises, and finance large-scale investment under strong government interference in the economy.
  - Turkmenistan and Uzbekistan: strong growth underpinned by higher commodity prices; Turkmenistan official figures show real GDP growth averaged more than 15 percent in 2000–06 but these estimates likely have significant upward bias; Uzbekistan’s average growth of five percent in 2001–06 driven by world prices for gold and cotton and increasing exports of natural gas and cotton; binding constraint in nonresource sectors remains extensive state intervention and relatively closed economies.

### Methodology and data issues
- Growth accounting framework:
  - Uses Cobb-Douglas aggregate production function: Y = e^θ K^α L^β where θ is TFP growth, α is capital share, β is labor share.
  - Output growth decomposition: g = θ + α (growth rate of capital) + β (growth rate of labor).
- Measuring capital stock:
  - Capital stock measured via perpetual inventory method (PIM); initial capital often estimated using Harberger (1978) approach: K_{t-1} = I_t / (g + δ).
  - The effect of initial capital stock on capital series decreases rapidly with sample size of investment figures.
  - This paper estimates TFP growth using different levels of initial capital stock to GDP ratios and applies a one-time adjustment: capital stock for CIS countries reduced by the same rate as output between 1990 and 1995 so that the capital-output ratio is not allowed to rise during the sharp contraction in output.
- Measuring capital and labor shares:
  - Several approaches discussed:
    - Perfect competition assumption (factor earnings proportional to productivities) often infeasible for CIS due to data limitations.
    - A priori capital share in range 0.3–0.4 commonly used, though some studies find higher shares for developing countries.
    - Regression approach estimating production function coefficients (intercept measures TFP); endogeneity and correlation between factor growth and TFP are concerns.
    - Nonparametric kernel derivative estimation (Shigeru, Khan, and Murao, 2003) finds much lower capital elasticity (~0.20) for some East Asian countries.
    - Cointegration techniques for long-run share estimation are robust to endogeneity, simultaneity, and measurement error.
- Data constraints for CIS:
  - New database compiled from IMF WEO, ILO, UNECE and other sources; data quality issues due to underreporting by private enterprises and initial statistical systems focused on public enterprises.
  - Data for Tajikistan, Turkmenistan, and Uzbekistan are incomplete and not always of good quality; interpret cautiously.
  - Capital stock measurement concern: significant portion of communist capital stock may have been permanently scrapped—if so, contribution of capital accumulation could be underestimated during recovery.
  - Labor and capital quality adjustments largely unavailable; this causes potential overestimation of TFP.
  - Capacity utilization data limited: available for Russia only (industry surveys), showing U-shaped capital utilization pattern falling until 1996 and rising from 1999 onward; labor utilization increased from around 70 percent during 1994–98 to around 85 percent during 2000–04. Russia’s utilization results used as proxy for other CIS countries.
  - Education and hours-worked data limited; transition economies show relatively high secondary school attainment and little variation across countries, reducing expected correlation between education level and growth in this sample.

### Estimated results and regional TFP and factor shares (1996–2006)
- Estimated TFP growth (1996−2006), capital and labor elasticities (fixed effects and 2SLS reported):
  - CIS-12: TFP = 2.3 percent; Capital elasticity (α) = 0.63; Labor elasticity (β) = 0.51; R^2 = 0.62; Number of observations = 132.
  - Baltics-3: TFP = 3.0 percent; Capital elasticity (α) = 0.49; Labor elasticity (β) = 0.52; R^2 = 0.66; Number of observations = 33.
  - CE-5: TFP = 1.6 percent; Capital elasticity (α) = 0.40; Labor elasticity (β) = 0.62; R^2 = 0.66; Number of observations = 51.
  - SEE-6: TFP = 0.8 percent; Capital elasticity (α) = 0.78; Labor elasticity (β) = 0.29; R^2 = 0.63; Number of observations = 60.
- Alternative estimates (Method: Fixed effects using cross-section weights) presented with slightly different elasticities:
  - CIS-12: TFP = 2.6 percent; Capital elasticity = 0.66; Labor elasticity = 0.53; R^2 = 0.59; Number of observations = 131.
  - Baltics-3: TFP = 3.0 percent; Capital elasticity = 0.49; Labor elasticity = 0.52; R^2 = 0.66; Number of observations = 33.
  - CE-5: TFP = 1.6 percent; Capital elasticity = 0.40; Labor elasticity = 0.62; R^2 = 0.66; Number of observations = 55.
  - SEE-6: TFP = 0.7 percent; Capital elasticity = 0.71; Labor elasticity = 0.34; R^2 = 0.62; Number of observations = 60.
- Notes on estimation:
  - All coefficients are significant at least at 1 percent confidence level.
  - Instruments for 2SLS are the lagged capital and labor growth rates.
  - Baltics-3 includes Estonia, Latvia, and Lithuania.
  - CE-5 includes Czech Republic, Hungary, Poland, Slovaka, and Slovenia.
  - SEE-6 includes Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Macedonia, and Romania.
- Interpretation:
  - TFP growth during 1996−2006 was highest for the Baltics (3.0 percent), followed by the CIS (2.3 percent), and CE (1.6 percent).
  - For the CIS sample, fixed effects estimate elasticity of capital = 0.63 and elasticity of labor = 0.51.
  - Sum of capital and labor elasticities is close to unity for the Baltics and CE, but slightly higher than one for the CIS.
  - Endogeneity partially addressed via 2SLS/instrumental variables; 2SLS estimates nearly identical to fixed effects estimates.
  - The majority of existing literature shows capital elasticity of 0.3 to 0.5 in industrial countries; the share of physical capital for industrial countries is likely to be lower than for developing countries where marginal product of capital is higher.

*Source: _wp07164 - Section V summarizes the findings and draws conclusions.*

### Box 1. Labor Market

### Box 1. Labor Market

### Employment trends during transition
- For the region as a whole, employment declined by about 20 percent from 1990 to 1997.
- As of 2006 and relative to the pre-transition level of 1990, the region still lost slightly more than 10 percent of its employment.
- The economic expansion after the Russian crisis in 1998 led only to modest employment increases in Russia, Ukraine, and Central Asia.
- During the sharp contraction of 1991–95, TFP fell dramatically and accounted for about half of the contraction in output. With the exception of Mongolia and Uzbekistan, total labor employment fell in all CIS and Baltic countries.

### Public vs. private sector employment
- Most transition economies experienced large shifts in labor between the public and private sectors.
- The reduction in employment in the public sector was not offset by the increase in the private sector, except for Azerbaijan, Kyrgyzstan, Mongolia, Tajikistan, and Uzbekistan.
- By 2006, the private sector generally accounted for between 60−80 percent of total employment.
- The share of employment in the public sector remains higher in slower-reforming economies (such as Turkmenistan and Uzbekistan).

### Demographics and human capital
- In 1995–2006, population and labor force growth rates increased in the range of 1-2 percent a year in Uzbekistan, Tajikistan, Kyrgyzstan, and Azerbaijan.
- In contrast, population and labor force rates declined by 1 to 1½ percent a year in Armenia, Georgia, Moldova, Russia, and Ukraine due to reduced fertility, increasing mortality and emigration.
- The secondary school enrollment rate in the CIS countries is still much higher than in most low- and middle-income developing countries. The human capital stock inherited from the Soviet era was very high.

### Informal sector, unemployment, and labor market measurement
- Official statistics may understate employment improvements in small and medium-size enterprises and in service sectors that are less well-monitored.
- A difficult business environment in some CIS countries has limited the ability of small and medium-sized enterprises to play a key role as employment generators.
- Self employment has increased substantially and is mainly concentrated in farming, wholesale and retail trade, and construction.
- Registered unemployment rates, according to an ILO definition, are relatively low, but labor force survey data indicate higher unemployment rates:
  - 9 percent in Kazakhstan, Russia, and Ukraine
  - 20 percent in Armenia, Georgia, and Moldova
- Engagement in the informal sector by those officially listed as unemployed or economically inactive is widespread in the CIS countries.
- According to the ILO, the size of the informal sector employment accounts for about one-half of all employment in the CIS as compared to about a quarter in Central European (CE) economies.
- Informal sector jobs are defined as value-adding activities outside the tax net and regulation; these may be unregistered and untaxed by nature (household subsistence economy) or emerge because of purposeful evasion and noncompliance.

### Links to growth accounting and TFP
- For 1996–2006 the most important source of growth by far has been capital accumulation, with the exception of Armenia, Tajikistan, and Ukraine, where the contribution of TFP growth was higher.
- The contribution of labor has been remarkably low, given the sharp decline in employment in the 1990s (improvements in quality of labor are reflected in TFP growth).
- During 1996-2006, average annual TFP growth in the CIS was higher than in CE and SEE, but lower than in the Baltics.
- Variation in estimated TFP growth among CIS countries (1996–2006):
  - Armenia and Azerbaijan above 4 percentage points
  - Georgia and Belarus between 2−3 percentage points
  - Moldova, Russia, and Ukraine from 1 to 2 percentage points
  - TFP growth less than one percentage point in Uzbekistan
  - TFP growth negative in Mongolia
- Factors explaining high TFP growth:
  - Managerial improvements, labor shedding, and gains from inter-industry resource reallocation following inefficiencies inherited from central planning.
  - Relatively low initial endowment of technology in some CIS economies implies larger measured TFP gains for a given technological innovation.
  - When capital is scarce, its marginal productivity is considerable, so the contribution of capital deepening is larger in economies with less capital.
  - Increases in capacity utilization from low mid-1990s levels could raise measured TFP growth.
  - Strong correlation between successful macroeconomic stabilization and progress in market reforms in the late 1990s and TFP growth.
  - A significant portion of high productivity growth is attributable to the rebound effect after the sharp fall in output in the first half of the 1990s.

*Source: Box 1. Labor Market, _wp07164 - Box 1. Labor Market*

### 2001. The average ratio of investment to GDP for the CIS increased by about four

### _wp07164 - 2001. The average ratio of investment to GDP for the CIS increased by about four

### Key quantitative findings on TFP and investment
- The average ratio of investment to GDP for the CIS increased by about four percentage points from 1996–2000 to 2001–06.
- With adjustment for capacity utilization (Scenario B in Table 6), the average TFP growth in 2001–06 went from 3.8 to 2.9 percentage points.
- For CIS-12, 2001–2006 (Table 6, Scenario A / Scenario B):  
  - Real GDP growth: 8.2  
  - Investment share of GDP: 24.8 (In percent)  
  - Growth of GDP: 6.7 (In percent, annual average)  
  - Capital growth: 0.9 (In percent)  
  - Labor growth: 7.5 (In percent)  
  - Productivity: 4.0 (In percent)  
  - Contribution of Capital: 0.4 (percentage points)  
  - TFP (Scenario A): 3.8 (percentage points)  
  - TFP (Scenario B): 2.9 (percentage points)

### Historical perspective and cross-region comparisons
- Postwar “Golden Age” Western Europe and Japan:
  - The average contribution of TFP to output growth was 2.7 percentage points for the seven major industrial countries—close to the estimated TFP growth for the CIS and accounting for about half of the growth in output.
  - U.S. per capita income growth averaged 2.4 percent a year between 1950 and 1973; over the same period, per capita income grew on average by 5 percent a year in Germany; and by slightly more than 8 percent in Japan.
- East Asia and the “four tigers”:
  - Young (1995) estimates: physical capital accumulation boosted growth in the “four tigers” by 4 percentage points during 1966–90; TFP contributed 1.7 percentage points; labor contributed 3.3 percentage points.
  - Shigeru, Khan, and Murao (2003): during 1960–95, estimated TFP growth explains 44 to 47 percent of output growth; capital growth contributes only 25 to 28 percent of output growth in East Asian countries.
- Country-specific long-run patterns:
  - Estimates in this paper: factor inputs (capital and labor) in Korea over 1975–2006 accounted for two-thirds of growth.
  - China: TFP growth averaged about 4 percentage points over the past two decades, contributing to 41 percent of output growth.
- Examples of sustained rapid growth with relatively low investment shares:
  - Chile: factor accumulation accounted for two-thirds of the growth in 1986-1995, and about 90 percent in 1996−2006.
  - Ireland: large increase in labor utilization—employment rate increase since the early 1990s averaged about 4 percent per annum; unemployment declined to 4 percent in 2005.
  - India: growth achieved with relatively little emphasis on capital accumulation and more substantial gains in TFP; since 2000 the volume of exports has grown three times faster than in the latter half of the 1990s.

### Technical efficiency, measurement, and interpretation
- Decomposition approach:
  - TFPi(t) = Ft + ûi(t) (equation (7)): TFP change is the sum of technological progress (movement of the best-practice frontier) and change in technical efficiency observed at the individual level.
  - Observed output G(t) = Gf(t) eu(t) = F(Z(t); t) eu(t) (equation (5)), with 0 < eu(t) = G(t)/Gf(t) < 1.
  - Time derivative form (equation (6)): Ġ(t)/G(t) = Fz Ż(t)/Z(t) + Ft + ů(t).
- Interpretation for transition economies:
  - A rapid shift from average practice to best practice—positive technical efficiency change—can generate high measured TFP growth, as observed in the CIS in recent years.
  - The paper assumes technological change (movement of best practice) is constant and does not vary across countries; under this assumption, variance in TFP change derives from variance in technical efficiency change.
- Measurement caveats:
  - Methodology does not adjust factor inputs for quality changes; embodied technological advancement in capital or improvements in labor quality (education, on-the-job training) can be measured as higher TFP rather than as capital or labor quality adjustments.
  - This “mis-measurement” of TFP may be significant for the CIS following the move from central planning to market economies.

### Dynamics, limits, and medium-term prospects
- Many previously fastest-growing economies (Chile, Ireland, Korea, post-war France/Germany/Japan) have not sustained average annual productivity growth rates in excess of 2.5 percent for long periods.
- A large part of recent CIS productivity growth reflects:
  - Improvements in allocation of resources,
  - Better use of investment,
  - Increases in capacity utilization,
  - Elimination of inefficiency,
  - Higher intensity of work.
- These aspects of productivity gains are essentially transitory and may not produce indefinite growth, but can have substantial impact over one or two decades.
- Several CIS economies still have catch-up potential:
  - Moldova, Georgia, Tajikistan, and Ukraine real GDP base is still substantially lower than real GDP level of 1990, suggesting further catch-up potential.
- As CIS countries approach the world technology frontier, opportunities for further TFP growth from eliminating central-planning inefficiencies will diminish; alternative channels will be needed.

### Central conclusions and policy implications
- Central conclusions:
  - Estimated total factor productivity (TFP) growth for the former Soviet Union republics were higher than other fast-growing economies in the sample.
  - Capital accumulation in most CIS countries made modest contributions to growth—on average much smaller than in the three Baltics and the five Central European economies.
  - Investment outlays remained relatively low with the exception of Azerbaijan and Mongolia.
  - Inefficiencies inherited from central planning left much scope for managerial improvements, labor shedding, and gains from inter-industry resource reallocation; these factors contributed to the post-reform recovery and rapid growth with low investment rates.
- Policy implications and areas to sustain growth:
  - Further improvement in market reforms and institutions to support productivity gains as catch-up opportunities narrow.
  - Greater labor use and continued faster capital accumulation expected to play a more important role in medium-term growth as TFP catch-up opportunities are exhausted.
  - Recognize that some recent TFP gains are transitory (reallocation, capacity utilization, intensity) and prepare for policies that promote sustained technological progress and efficiency-enhancing structural reforms.

*Source: Authors' own calculations, as presented in _wp07164 - 2001. The average ratio of investment to GDP for the CIS increased by about four.*

### References

### References

### Listed references

- Aiyer, Shekhar, and Carl-Johan Dalgaard, 2005, “Total Factor Productivity Revisited: A Dual Approach to Development Accounting,” IMF Staff Papers, Vol. 52, No. 1, pp. 82–102.
- Benhabib, Jess, and Mark M. Spiegel, 1994, “The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data.” Journal of Monetary Economics, Vol. 34, pp. 363–390.
- Barro, Robert, 1999, “Notes on Growth Accounting,” Journal of Economic Growth, Vol. 4, June, pp. 119–37.
- Benhabib, Jess, and Mark M. Spiegel, 1994, “The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data,” Journal of Monetary Economics, Vol. 34, No. 2, pp. 143–73.
- Campos, N. F., and F. Coricelli, 2002, “Growth in Transition: What We Know, What We Don’t Know and What We Should,” Journal of Economic Literature, Vol. 40, pp. 793–836.
- Christiansen, Laurits R., Dianne Cummings, and Dale W. Jorgenson, 1980, “Economic Growth, 1947–73: An International Comparison,” In J.W. Kendrick and B. Vaccara (eds.), New Developments in Productivity Measurement and Analysis, Studies in Income and Wealth, Vol. 41 (Chicago: University of Chicago Press).
- Boseworth, Barry P., and Susan M. Collins, 2003, “The Empirics of Growth: An Update,” Brooking Papers on Economic Activity: 2, pp. 113–206.
- De Broeck, M. and V.Koen, 2000, “The Great Contractions in Russia, the Baltics and the Other Countries of the former Soviet Union: A View from the Supply Side,” IMF Working Paper, 2000/32 (Washington DC: International Monetary Fund).
- Denison, E., 1967, Why Growth Rates Differ?, The Brookings Institution, Washington, DC.
- Easterly, William and Ross Levine, 2001, “It’s Not Factor Accumulation: Stylized Facts and Growth Models,” World Bank Economic Review, Vol. 15, No. 2, pp. 177–219.
- Easterly, Williams, 2005, “National Economic Policies and Economic Growth: A Reappraisal,” Handbook of Economic Growth, P. Aghion and S. Durlauf, eds., (Amsterdam, Netherlands: Elsevier BV).
- Elson, Anthony, 2006, “What Happened,” Finance and Development, International Monetary Fund, Vol. 43, No. 2.
- Färe, Rolf, Shawna Grosskopf, Mary Norris, and Zhongyang Zhang, 1994, “Productivity Growth, Technical Progress, and Efficiency Change in Industrialized Countries.” American Economic Review, Vol. 84, pp. 66–83.
- Heston, A. Summers, R. & Aten, B, 2006, Penn World Table Version 6.2, Center for International Comparison at the University of Pennsylvania (CICP).
- International Monetary Fund, 2004, Chile: Selected Issues, IMF Country Report No. 04/292, Washington, D.C.
- _____, 2004, Ireland: Selected Issues, IMF Country Report No. 04/340, Washington, D.C.
- _____, 2007, World Economic Outlook, Washington D.C.: International Monetary Fund, May.
- International Labor Organization, 2005, Statistical Yearbook, Geneva.
- Jorgenson, D. W., 1995, Productivity: International Comparison of Economic Growth, MIT Press, Cambridge, MA.
- Limam, Y. R. and Miller, S. M., 2004, Explaining Economic Growth: Factor Accumulation, Total Factor Productivity, and Production Efficiency Improvement, Department of Economics Working Paper Series 2004–20, University of Connecticut.
- Loukoianova, Elena, and Anna Uigovskaya, 2004, “Analysis of Recent Growth in Low-Income CIS Countries,” IMF Working Paper, 2000/32 (Washington DC: International Monetary Fund).
- Maddison, A., 1996, “Macroeconomic Accounts for European Countries,” in B. van Ark and N.F.R. Crafts (eds.), Quantitative Aspects of Postwar European Economic Growth (Cambridge: Cambridge University Press), pp. 27–83.
- Nishimizu, Mieko, and John M. Page, Jr., 1982, “Total Factor Productivity Growth, Technological Growth, and Technical Efficiency Change: Dimensions of Productivity Change in Yugoslavia, 1965-78,” Economic Journal, Vol. 92, pp. 920–36.
- Oomes, Nienke, and Oksana Dynnikova, 2006, “The Utilization-Adjusted Output Gap: Is the Russian Economy Overheating?” IMF Working Paper, No. 06/68; (Washington, D.C.: International Monetary Fund).
- Owen, D. and D. Robinson, 2003, Russia Rebounds, Washington: International Monetary Fund.
- Pagan, Adrian, and Aman Ullah, 1999, Nonparametric Econometrics (Cambridge: Cambridge University Press).
- Schadler, S., A. Ashole, A. Abiad, and D. Leigh, 2006, Growth in the Central and Eastern European Countries of the European Union, IMF Occasional Paper No. 252, Washington, DC.
- Senhadji, Abdelhak, 2000, “Sources of Economic Growth: An Extensive Growth Accounting Exercise,” IMF Staff Papers, Vol. 47, No. 1, pp. 129–57.
- Shigeru Iwata, Mohsin S. Khan, and Hiroshi Murao, 2003, “Sources of Economic Growth in East Asia: A Nonparametric Assessment,” IMF Staff Papers, Vol. 50, No. 2, pp. 157–77.
- Solow, Robert, 1957, “Technical Change and the Aggregate Production Function,” Review of Economics and Statistics, Vol. 39, No. 3, pp. 312–20.
- World Bank, 1993, The East Asian Miracle, Oxford University Press.
- ______, 2005, Economic Growth in the 1990s: Learning from a Decade of Reform, Washington, DC.
- Young, A., 1994, “Lessons from the NICs: A Contrarian View,” European Economic Review, Vol. 38, pp. 964–73.
- ______, 1995, “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth Experience,” Quarterly Journal of Economics, August, pp. 641–80.

*Source: _wp07164 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07164.pdf_