## _wp04151

## Source details

**Canonical URL:** [_wp04151](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04151.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04151.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04151.pdf.json)

---

### Introduction and scope
- Focus: sources and sustainability of recent economic growth in the CIS-7 countries (Armenia, Azerbaijan, Georgia, the Kyrgyz Republic, Moldova, Tajikistan, Uzbekistan).
- Time frame emphasized: up to and including 2002; key summary statistic: in 1998–2002, average annual real GDP growth in the CIS-7 countries was 5.4 percent.
- Exclusion: Turkmenistan excluded due to "large gaps in data and poor quality of the available data".
- Analytical approaches used:
  - Demand-side decomposition of GDP.
  - Growth accounting and analysis of total factor productivity (TFP).
  - Sectoral breakdown of GDP and employment.
  - Econometric analysis to identify factors driving economic growth.

### Recent growth developments in the CIS-7 countries
- Growth performance (Real GDP growth, percent):
  - Armenia: 1998 = 7.3; 1999 = 3.3; 2000 = 6.0; 2001 = 9.6; 2002 = 12.9; 2003 = 7.0; 1998–2003 average = 7.7
  - Azerbaijan: 1998 = 10.0; 1999 = 7.4; 2000 = 11.1; 2001 = 9.9; 2002 = 10.6; 2003 = 9.2; 1998–2003 average = 9.7
  - Georgia: 1998 = 2.9; 1999 = 3.0; 2000 = 1.9; 2001 = 4.7; 2002 = 5.5; 2003 = 11.1; 1998–2003 average = 4.9
  - Kyrgyz Republic: 1998 = 2.1; 1999 = 3.7; 2000 = 5.3; 2001 = 5.4; 2002 = -0.5; 2003 = 5.6; 1998–2003 average = 3.6
  - Moldova: 1998 = -6.5; 1999 = -3.4; 2000 = 2.1; 2001 = 6.1; 2002 = 7.8; 2003 = 6.3; 1998–2003 average = 2.1
  - Tajikistan: 1998 = 5.3; 1999 = 3.7; 2000 = 8.3; 2001 = 10.2; 2002 = 9.1; 2003 = 6.0; 1998–2003 average = 7.1
  - Uzbekistan: 1998 = 2.1; 1999 = 3.4; 2000 = 3.3; 2001 = 4.1; 2002 = 3.2; 2003 = 0.3; 1998–2003 average = 2.7
  - CIS-7 country average: 1998 = 3.3; 1999 = 3.0; 2000 = 5.4; 2001 = 7.1; 2002 = 6.9; 2003 = 6.5; 1998–2003 average = 5.4
- Comparative patterns:
  - CIS-7 growth during 1998–2003 exceeded CEEB average (1998–2003 average = 4.0) and was broadly equal to other CIS countries (Other CIS average 1998–2003 = 5.1).
- Contraction depth (1989 = 100; Year of first positive real GDP growth; cumulative drop from 1989, percent):
  - Armenia: Year = 1994; Cumulative drop = 53.3
  - Azerbaijan: Year = 1996; Cumulative drop = 42.6
  - Georgia: Year = 1996; Cumulative drop = 78.8
  - Kyrgyz Republic: Year = 1996; Cumulative drop = 48.7
  - Moldova: Year = 1997; Cumulative drop = 63.2
  - Tajikistan: Year = 1997; Cumulative drop = 68.6
  - Uzbekistan: Year = 1996; Cumulative drop = 13.9
- Exchange rate and external shocks:
  - During 1998–2002, real depreciation of CIS-7 currencies was on average 20 percent against the Russian ruble and 14 percent against all currencies combined.
  - The Russian 1998 crisis initially slowed CIS-7 but subsequent Russian recovery supported CIS-7 export demand.

### Demand-side decomposition and export/investment dynamics
- Shift in growth drivers:
  - Early 1990s: recovery and growth mostly driven by consumption, financed by external borrowing and income from the underground economy.
  - Since the late 1990s: growth in most CIS-7 countries resulted from expansion of exports and an increase in investment (data limitations for Tajikistan and Uzbekistan limit full expenditure decomposition for those two).
- Selected contributions to real GDP growth (1996–2002 averages and shifts):
  - Armenia: GDP growth increased to Avg. (01–02) = 11.3; net export contribution (01–02) = 4.0; investment contribution (01–02) = 1.7; final consumption contribution (01–02) = 5.5.
  - Azerbaijan: GDP growth (01–02) = 10.3; final consumption (01–02) = 4.9; investment (01–02) = 3.2; net export (01–02) = 2.2.
  - Georgia: net export contribution rose (98–00) = 5.0 and (01–02) = 4.3, while final consumption went negative in (98–00) = -3.2.
  - Moldova: recovery (01–02) GDP growth = 7.0; final consumption (01–02) = 7.9.
- Investment ratios, 1999–2002 average (percent of GDP):
  - CIS-7 countries = 17.1
  - CIS-7 w/o Azerbaijan = 15.4
  - Other CIS countries = 20.7
  - CEEB = 22.7
  - OECD = 21.9
  - Exception: Azerbaijan — investment in oil-producing industry accounted for 27.4 percent of GDP in 1999–2002.
- Export shares (exports of goods and services, percent of GDP, 1997–2002 snapshots):
  - Armenia: 1997 = 20; 1998 = 21; 1999 = 23; 2000 = 25; 2001 = 25; 2002 = 30
  - Azerbaijan: 1997 = 29; 1998 = 23; 1999 = 28; 2000 = 39; 2001 = 41; 2002 = 44
  - Georgia: 1997 = 19; 1998 = 20; 1999 = 26; 2000 = 34; 2001 = 32; 2002 = 29
  - Kyrgyz Republic: 1997 = 38; 1998 = 36; 1999 = 42; 2000 = 42; 2001 = 37; 2002 = 37
  - Moldova: 1997 = 48; 1998 = 41; 1999 = 47; 2000 = 45; 2001 = 46; 2002 = 46
  - Tajikistan: 1997 = 72; 1998 = 49; 1999 = 64; 2000 = 86; 2001 = 69; 2002 = 64
  - Uzbekistan: 1997 = 27; 1998 = 23; 1999 = 18; 2000 = 25; 2001 = 28; 2002 = 31
- Competitiveness factor:
  - Average monthly wages stayed below 40 dollars during 1997–2002, supporting competitiveness.

### Growth accounting and measurement of TFP
- Production function and decomposition:
  - Assumes Cobb-Douglas: Yt = At Kt^α Lt^(1−α).
  - TFP change (log) computed from: gA = gY − α gK − (1−α) gL (presented in source as ,)1( LKYA gggg αα −−−= ).
  - Capital accumulation: Kt+1 = (1−δ) Kt + It with δ = 3 percent (assumed constant).
  - Elasticities of output w.r.t capital and labor assumed constant and summing to 1; specific α values in source summarized as "0.3 and 0.7, respectively, which is consistent with previous studies on growth accounting."
- Data and measurement caveats:
  - Measured TFP is a residual capturing efficiency of input use and various measurement and institutional factors (working hours, hidden employment, labor hoarding, capacity utilization, legalization of informal sector activity).
  - No capacity utilization series for labor and capital available for CIS-7.
  - Alternative extreme depreciation scenarios (very high obsolescence early on) yield qualitatively similar TFP and capital stock patterns to constant δ = 3 percent.
- Aggregate TFP and productivity summary (selected averages, percent):
  - CIS-7 countries Avg. 1971–97: Output Growth Rate = -0.4; Capital Growth Rate = 3.7; Labor Growth Rate = 1.3; TFP Growth Rate = -2.4; Labor Productivity Growth Rate = -1.8.
  - CIS-7 countries Avg. 1991–97: Output Growth Rate = -10.5; Capital Growth Rate = -0.1; Labor Growth Rate = -1.3; TFP Growth Rate = -9.5; Labor Productivity Growth Rate = -7.9.
  - CIS-7 countries Avg. 1971–2000: Output Growth Rate = 0.5; Capital Growth Rate = 3.1; Labor Growth Rate = 1.2; TFP Growth Rate = -1.6; Labor Productivity Growth Rate = -1.1.
  - CIS-7 countries Avg. 1998–2000: Output Growth Rate = 4.3; Capital Growth Rate = -0.3; Labor Growth Rate = 0.6; TFP Growth Rate = 4.6; Labor Productivity Growth Rate = 4.1.
- Annual TFP growth (selected country-year examples, percent):
  - Armenia: 1998 = 7.2; 1999 = 9.6; 2000 = 6.1; 2001 = 7.7.
  - Azerbaijan: 1999 = 5.9; 2000 = 9.7.
  - Georgia: 1998 = 18.4; 1999 = 5.4; 2002 = 6.3.
  - Kyrgyz Republic: 1997 = 9.6; 2000 = 4.3; 2002 = -3.6.
  - Moldova: 2000 = 4.7; 2001 = 8.3; 2002 = 0.3.
  - Tajikistan: 1998 = 5.4; 1999 = 6.3; 2000 = 8.2; 2001 = 7.0.
  - Uzbekistan: 1997 = 3.1; 1998 = 3.4; 1999 = 4.1.
- Regression correlates of TFP changes (fixed-effect estimation, 1991–2002, CIS-7; coefficients from source):
  - Unemployment rate: -0.627 (col 1), -0.81 (col 2), -0.944 (col 3).
  - Average EBRD transition index: 12.427 (col 2).
  - Large-scale privatization: 3.439 (col 1), 3.874 (col 3).
  - Infrastructure reform: 13.549 (col 1), 12.636 (col 3).
  - Country fixed-effect examples (col 2 and col 3): Armenia = -16.95 and -20.16; Azerbaijan = -27.79 and -26.59; Georgia = -19.52 and -24.36; Kyrgyz Republic = -28.98 and -23.78; Moldova = -26.18 and -25.21; Tajikistan = -22.77 and -21.61; Uzbekistan = -23.24 and -22.35.
  - R-squared: 0.29 (col 1), 0.29 (col 2), 0.33 (col 3). Adjusted R-squared: 0.24, 0.19, 0.21. Number of observations: 62 in each specification.
- Interpretation:
  - TFP increases correlate positively with infrastructure reform and large-scale privatization and negatively with unemployment; strong negative country fixed effects reflect adverse initial conditions and measurement differences.

### Sectoral decomposition and sectoral TFP
- Sectoral coverage: industry, agriculture, construction, transport (including communications), and trade (including procurement); services largely excluded due to data limitations.
- Employment and output structure (selected facts):
  - Agriculture provides >40 percent of labor force and >25 percent of GDP in the CIS-7 on average.
  - Example sector shares (percent of GDP, selected years):
    - Armenia 1995: Agriculture 41; Industry 24; Other 25.
    - Azerbaijan 2001: Industry 36; Agriculture 16; Other 31.
  - Example employment shares (percent of employment):
    - Armenia 2000: Agriculture 44; Industry 14.
- Sectoral investment shares (percent of total investment, selected country examples):
  - Azerbaijan industry investment shares: 1995 = 46.4; 1996 = 72; 1997 = 70.7; 1998 = 74.9; 1999 = 61.7; 2000 = 68.1; 2001 = 70.5; 2002 = 77.
  - Kyrgyz Republic industry investment shares: 1995 = 76.9; 1996 = 80; 1997 = 69.3; 1998 = 33.52; 1999 = 34.7; 2000 = 40.5; 2001 = 32.34; 2002 = 18.35.
- Sectoral TFP trends (selected points):
  - Agriculture: average TFP growth positive but unstable; Moldova had highest agricultural TFP growth between 1998 and 2002 (breakup of collective farms partly explains).
  - Industry: TFP growth turned positive after 1998 in all CIS-7 except Moldova.
  - Trade: highest recent TFP growth among sectors — varied from 5 percent in the Kyrgyz Republic to 33 percent in Armenia.
  - Transport: TFP continued to decline on average in all countries but the Kyrgyz Republic in 1998–2002.
  - Construction: average TFP growth resumed in recent years.
- Export composition risk (share of low-degree-processed commodities in exports, percent of total export):
  - Armenia: 1995 = 61.0; Latest = 59.0
  - Azerbaijan: 1995 = 64.5; Latest = 90.7
  - Georgia: 1995 = 30.1; Latest = 44.3
  - Kyrgyz Republic: 1995 = 36.2; Latest = 73.5
  - Moldova: 1995 = 62.8; Latest = 76.6
  - Tajikistan: 1995 = 59.0; Latest = 84.0
  - Uzbekistan: 1995 = 64.0; Latest = 64.5

### Econometric analysis of growth determinants (24 transition countries and CIS-7 subsample)
- Regression specification includes: country fixed effects; exchange rate regime dummy (ER); fiscal policy (general government budget deficit in percent of GDP); initial shock variable (IS); WarCIS7 dummy; and EBRD transition indices (small-scale privatization, large-scale privatization, governance and enterprise reforms, price liberalization, trade liberalization, competition policy, banking reform and interest rate liberalization, securities markets reform, infrastructure reform). EBRD indices range 1–4.
- Key regression findings (summary):
  - Initial shock variable appears insignificant in all specifications (fixed effects likely capture initial conditions).
  - De facto exchange rate stability and prudent fiscal policy exert a large positive effect on real GDP growth in the 24-country sample; exchange rate dummy effect weakens when reform variables are included.
  - War dummy for CIS-7 is negative but insignificant.
  - Net effect of privatization on growth is positive overall; small-scale privatization is positive and significant; large-scale privatization shows negative or weakly significant effects in some specifications.
  - Price liberalization variable has a negative sign (timing effect: price liberalization often occurred during early transition years of output decline).
  - Reform of nonbank financial institutions is positive and significant at the 10 percent level.
- Selected coefficient examples from Table 14 (24-country fixed effects):
  - Exchange rate dummy: 0.05 (col 1), 0.046 (col 2), 0.045 (col 3), 0.048 (col 4).
  - Fiscal: 0.47 (col 1), 0.533 (col 2), 0.511 (col 3), 0.493 (col 4).
  - Small-scale privatization: 0.03 (col 1), 0.038 (col 2), 0.037 (col 3), 0.046 (col 4).
  - Large-scale privatization: -0.01 (col 1), examples of negative coefficients -0.054, -0.047 in other columns.
  - R-squared (24-country sample): 0.58 (cols 1–3), 0.57 (col 4). Adjusted R-squared: 0.53 (cols 1–4).
  - Number of observations: 284 for 24-country regressions; 77 for CIS-7 regressions.

### Foreign Direct Investment (FDI) patterns
- FDI in CIS-7, percent of GDP (selected years 1996–2002):
  - Armenia: 1996 = 1.1; 1997 = 3.2; 1998 = 11.6; 1999 = 6.6; 2000 = 5.5; 2001 = 3.3; 2002 = 4.6.
  - Azerbaijan: 1996 = 20.8; 1997 = 27.4; 1998 = 23.0; 1999 = 11.1; 2000 = 2.8; 2001 = 5.2; 2002 = 17.1.
  - Georgia: 1996 = 1.8; 1997 = 6.6; 1998 = 6.1; 1999 = 2.2; 2000 = 5.0; 2001 = 2.5; 2002 = 3.8.
  - Kyrgyz Republic: 1996 = 2.6; 1997 = 4.7; 1998 = 5.3; 1999 = 3.1; 2000 = -0.5; 2001 = -0.1; 2002 = 1.0.
  - Moldova: 1996 = 1.4; 1997 = 4.1; 1998 = 4.4; 1999 = 3.3; 2000 = 11.1; 2001 = 10.5; 2002 = 6.6.
  - Tajikistan: 1996 = 1.7; 1997 = 1.6; 1998 = 1.9; 1999 = 1.9; 2000 = 2.4; 2001 = 0.9; 2002 = 3.0.
  - Uzbekistan: 1996 = 0.7; 1997 = 1.1; 1998 = 0.9; 1999 = 0.7; 2000 = 0.5; 2001 = 0.7; 2002 = 0.7.
  - CIS-7 average: 1996 = 4.3; 1997 = 6.9; 1998 = 7.6; 1999 = 4.1; 2000 = 3.8; 2001 = 3.3; 2002 = 5.3.
  - CIS-7 excluding Azerbaijan: 1996 = 1.5; 1997 = 3.6; 1998 = 5.0; 1999 = 3.0; 2000 = 4.0; 2001 = 3.0; 2002 = 3.3.

### Main conclusions
- Recovery drivers:
  - Early recovery driven by consumption.
  - Since 1998, growth in Armenia, Azerbaijan, Georgia, and Kyrgyz Republic driven by export expansion and investments in export-oriented sectors.
  - Consumption remained a primary driver in Moldova, and possibly in Tajikistan and Uzbekistan.
- Exports and vulnerabilities:
  - Exports resilient to global slowdown due to comparative advantages: high energy endowments in selected countries, artificially low energy prices, and low real wages.
  - Increasing dependence on exports of raw materials, agricultural products, and low-degree-processed goods heightens vulnerability to world commodity prices and weather.
- Productivity and reform:
  - Since 1998, overall TFP appears to be increasing in all CIS-7 countries — consistent with reform effects — but interpretation must account for capacity utilization, improved statistical coverage of private sector, and hidden employment changes.
  - Leading growth sectors: agriculture, industry, and trade; industrial growth concentrated in export-oriented raw-material sectors.
- Investment and sustainability:
  - Investment share in GDP is rising but remains low compared with other CIS and CEEB countries (exception: petroleum sector in Azerbaijan).
  - Low investment levels are insufficient to guarantee long-term sustainable growth.
- Agriculture:
  - Agricultural output recovery may be transitory; trends include increasing labor intensity, declining share of investment, and movement toward subsistence farming in poorest CIS countries.
- Trade potential and constraints:
  - Realizing trade potential requires reducing restrictive trade policies, repairing physical infrastructure, and addressing corruption and governance problems in customs and transport services.
  - Insufficient foreign investment constrains diversification and improvement of trade environment.

### Policy recommendations
- Continue trade liberalization, emphasizing:
  - Reducing corruption and improving governance in customs and transport services.
  - Removing man-made barriers to trade.
  - Facilitating development of physical infrastructure conducive to trade, including regional transportation networks to improve access to world markets.
- Improve general investment climate to attract new investment needed for diversification and sustainable growth.
- Continue structural reforms to support diversification away from export concentration in raw materials and low-degree-processed goods.

### Data description and measurement issues (Appendix I)
- Key variable definitions and sources:
  - GDP growth: Percentage change of GDP in constant prices. Source: WEO (2003), data from IMF economists working on the countries.
  - Exchange rate dummy: 1 = stable rate (de facto peg or bands 2–5 percent per Reinhart and Rogoff (2004)); 0 = floating rate. Source: Authors’ estimates.
  - Fiscal: General government budget balance in percent of GDP. Source: WEO.
  - Avg Index: Average of all EBRD liberalization indices. Source: Authors’ estimates. EBRD indices source: EBRD, 2002.
- Measurement and conceptual issues emphasized:
  - Official data subject to conceptual and measurement problems; large statistical discrepancies for some countries, especially early transition years.
  - For some countries, statistical discrepancies are captured in different expenditure items (example: Uzbekistan—“Investments” contains large sub-item “Inventory” used to balance discrepancies).
  - Tajikistan: GDP by expenditure data available only from 1998 onwards.
  - In absence of reliable official data, staff often derive expenditure components with consumption as a residual.
  - Likely biases: underreporting of the shadow economy more likely to affect final consumption than investment; incentives suggest possible overstatement of investment and understatement of consumption.
  - Net exports contribution to real GDP growth computed as residual because net exports in constant prices are not available.

*Source: _wp04151 (IMF working paper content provided).*

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

### _wp04151 - References

### Introduction and scope
- Focus: sources and sustainability of recent economic growth in the CIS-7 countries (Armenia, Azerbaijan, Georgia, the Kyrgyz Republic, Moldova, Tajikistan, Uzbekistan).
- Time frame emphasized: up to and including 2002; key summary statistic: in 1998–2002, average annual real GDP growth in the CIS-7 countries was 5.4 percent.
- Exclusion: Turkmenistan is excluded due to "large gaps in data and poor quality of the available data".
- Analytical approaches used in the source:
  - Demand-side decomposition of GDP.
  - Growth accounting and analysis of total factor productivity (TFP).
  - Sectoral breakdown of GDP and employment.
  - Econometric analysis to identify factors driving economic growth.

### Findings: demand-side decomposition and export/investment dynamics
- Shift in sources of growth during the 1990s:
  - Early 1990s: recovery and growth mostly driven by consumption, financed by external borrowing and income from the underground economy.
  - Since the late 1990s: growth in most CIS-7 countries resulted from expansion of exports and an increase in investment.
- Factors contributing to resilience of exports to the global slowdown:
  - High energy endowments in some countries.
  - Artificially low energy prices.
  - Low wages.

### Findings: growth accounting and productivity
- Growth accounting results:
  - Starting from 1998, overall TFP growth resumed for the first time since the breakup of the Soviet Union.
  - TFP growth averaged over 5 percent in three of the CIS-7 countries.
- Caveats and data limitations affecting interpretation:
  - Data weaknesses and measurement difficulties related to:
    - Changes in informal sector activity.
    - Labor hoarding.
    - Capacity utilization.
    - Depreciation rates.

### Findings: sectoral composition and vulnerabilities
- Sectoral concentration of recent GDP growth:
  - Concentrated in agriculture, extraction industries, production of goods with low degree of processing, trade, and services.
- Structural concerns:
  - Stagnant share of manufacturing.
  - Increasing dependence on exports of raw materials and agricultural products.
  - Resulting vulnerability to changes in external conditions.

### Econometric results: drivers of growth
- Econometric findings (based on data through 2002) are consistent with prior studies and indicate that:
  - Lower fiscal deficits are associated with higher growth.
  - Exchange rate stability is associated with higher growth.
  - Implementing structural reforms leads to higher growth.

### Organization of the full paper (as described)
- Section II: growth performance in the CIS-7 countries compared with the rest of the CIS and Central and Eastern Europe and the Baltics (CEEB).
- Section III: sources of growth — demand-side decomposition of GDP.
- Section IV: growth accounting framework and analysis of TFP growth rates.
- Section V: sectoral breakdown of GDP and employment.
- Appendices and tables cover data description, GDP by expenditure, annual TFP and labor productivity growth rates, components of GDP by expenditures, and foreign direct investment in CIS-7.

*Source: _wp04151 - References (excerpt).*

### Section VI presents the results of the econometric analysis of the main factors that determine

### Section VI presents the results of the econometric analysis of the main factors that determine

### Recent growth developments in the CIS-7 countries
- The CIS-7 countries exhibited strong, although uneven, growth in the five-year period 1998–2003.
- During 1998–2003, growth in the CIS-7 countries was higher than in the CEEB and other developing countries, and was broadly equal to that in the other CIS countries.
- Table 1 — Real GDP Growth in the CIS-7 Countries, 1998–2003 (In percent):
  - Armenia: 1998 = 7.3; 1999 = 3.3; 2000 = 6.0; 2001 = 9.6; 2002 = 12.9; 2003 = 7.0; 1/ = 7.7; 1998–2003 average = 7.7
  - Azerbaijan: 1998 = 10.0; 1999 = 7.4; 2000 = 11.1; 2001 = 9.9; 2002 = 10.6; 2003 = 9.2; 1/ = 9.7
  - Georgia: 1998 = 2.9; 1999 = 3.0; 2000 = 1.9; 2001 = 4.7; 2002 = 5.5; 2003 = 11.1; 1/ = 4.9
  - Kyrgyz Republic: 1998 = 2.1; 1999 = 3.7; 2000 = 5.3; 2001 = 5.4; 2002 = -0.5; 2003 = 5.6; 1/ = 3.6
  - Moldova: 1998 = -6.5; 1999 = -3.4; 2000 = 2.1; 2001 = 6.1; 2002 = 7.8; 2003 = 6.3; 1/ = 2.1
  - Tajikistan: 1998 = 5.3; 1999 = 3.7; 2000 = 8.3; 2001 = 10.2; 2002 = 9.1; 2003 = 6.0; 1/ = 7.1
  - Uzbekistan: 1998 = 2.1; 1999 = 3.4; 2000 = 3.3; 2001 = 4.1; 2002 = 3.2; 2003 = 0.3; 1/ = 2.7
  - CIS-7 country average: 1998 = 3.3; 1999 = 3.0; 2000 = 5.4; 2001 = 7.1; 2002 = 6.9; 2003 = 6.5; 1/ = 5.4
  - Other CIS 2/ countries: 1998 = -0.2; 1999 = 3.1; 2000 = 7.9; 2001 = 8.1; 2002 = 5.8; 2003 = 6.1; 1/ = 5.1
  - CEEB: 1998 = 3.9; 1999 = 2.2; 2000 = 4.6; 2001 = 3.8; 2002 = 4.1; 2003 = 5.4; 1/ = 4.0
  - Memorandum items:
    - Developing countries: 1998 = 3.5; 1999 = 3.9; 2000 = 5.7; 2001 = 4.1; 2002 = 4.6; 2003 = 5.0; 1/ = 4.5
    - Industrial countries: 1998 = 2.7; 1999 = 3.4; 2000 = 3.9; 2001 = 1.0; 2002 = 1.8; 2003 = 1.8; 1/ = 2.4
  - Source: World Economic Outlook (IMF, 2003). 1/ Preliminary. 2/ Hereafter “Other CIS” group excludes Turkmenistan.
- Stabilization-time comparison:
  - Stabilization programs in the CIS-7 and other CIS countries started between September 1993 and February 1995; stabilization programs in the CEEB countries began during January 1990–October 1993.
  - Using stabilization time, the relative magnitude of the recovery in the CIS-7 countries surpassed that in the other CIS and the CEEB (Figure 1: Real GDP Recovery, Stabilization time; T(-1) = 100).
- Contraction depth of real GDP (Table 2; 1989 = 100; Year of the First Positive Real GDP Growth; Cumulative Drop from 1989, In percent):
  - Armenia: Year = 1994; Cumulative drop = 53.3
  - Azerbaijan: Year = 1996; Cumulative drop = 42.6
  - Georgia: Year = 1996; Cumulative drop = 78.8
  - Kyrgyz Republic: Year = 1996; Cumulative drop = 48.7
  - Moldova: Year = 1997; Cumulative drop = 63.2
  - Tajikistan: Year = 1997; Cumulative drop = 68.6
  - Uzbekistan 1/: Year = 1996; Cumulative drop = 13.9
  - Kazakhstan: Year = 1997; Cumulative drop = 39.4
  - Russia: Year = 1997; Cumulative drop = 41.3
  - Source: World Economic Outlook, IMF, 2003.
  - Note: 1/ Official data show a much lower initial output decline in Uzbekistan; potential explanations include less accurate GDP data, low initial industrialization, readily sold cotton production, and self-sufficiency in energy.
- External shocks and exchange rates:
  - The Russian crisis in 1998 initially led to a slowdown in most CIS-7 countries; as Russia recovered, stronger import-demand from Russia provided an impetus for recovery.
  - Real depreciation of CIS-7 currencies against the ruble and other currencies could have increased CIS-7 market shares.
  - During 1998–2002, real depreciation of CIS-7 currencies was on average 20 percent against Russian ruble and 14 percent against all currencies combined.
- Overall interpretation:
  - The recent growth spurt in the CIS-7 countries could be partially explained by low initial real GDP levels, increased demand from Russia and other countries, and improved productivity.
  - However, EBRD transition indicators show reforms in the CIS-7 countries did not progress as much as in other transition countries.

### Demand decomposition: has growth been driven by consumption?
- Data limitations:
  - Insufficient data for Tajikistan and Uzbekistan prevent calculation of contributions of expenditure components to GDP growth for those countries.
- Key findings on expenditure components (Table 3 — Contribution of Expenditure Components to Real GDP Growth, 1998–2002):
  - For five CIS-7 countries (Armenia, Azerbaijan, Georgia, Kyrgyz Republic, Moldova), recovery after 1998 was driven more by net export and investment, rather than by consumption as in earlier years.
  - High consumption prior to 1998 likely stimulated capacity utilization and possibly investment.
  - Country-specific highlights:
    - Armenia:
      - GDP growth (Avg. 96–97) = 4.6; (Avg. 98–00) = 5.5; (Avg. 01–02) = 11.3
      - Final consumption (96–97) = 5.4; (98–00) = 5.3; (01–02) = 5.5
      - Investment (96–97) = 1.3; (98–00) = 0.9; (01–02) = 1.7
      - Net export (96–97) = -2.1; (98–00) = -0.7; (01–02) = 4.0
      - Share of components (In percent of total): Final consumption (96–97) = 117.4; (98–00) = 96.2; (01–02) = 48.9; Investment (96–97) = 28.3; (98–00) = 15.9; (01–02) = 15.2; Net export (96–97) = -45.7; (98–00) = -12.1; (01–02) = 35.9
    - Azerbaijan:
      - GDP growth (96–97) = 3.5; (98–00) = 9.5; (01–02) = 10.3
      - Final consumption (96–97) = 4.7; (98–00) = 10.3; (01–02) = 4.9
      - Investment (96–97) = 2.9; (98–00) = 4.1; (01–02) = 3.2
      - Net export (96–97) = -4.1; (98–00) = -4.9; (01–02) = 2.2
      - Share of components (In percent of total): Final consumption (96–97) = 134.3; (98–00) = 108.4; (01–02) = 47.4; Investment (96–97) = 82.9; (98–00) = 43.3; (01–02) = 31.2; Net export (96–97) = -117.1; (98–00) = -51.7; (01–02) = 21.4
    - Georgia:
      - GDP growth (96–97) = 6.9; (98–00) = 5.5; (01–02) = 3.8
      - Final consumption (96–97) = 7.7; (98–00) = -3.2; (01–02) = -0.8
      - Investment (96–97) = 1.9; (98–00) = 3.7; (01–02) = 0.3
      - Net export (96–97) = 1.7; (98–00) = 5.0; (01–02) = 4.3
      - Share of components (In percent of total): Final consumption (96–97) = 111.6; (98–00) = -57.9; (01–02) = -22.7; Investment (96–97) = 27.5; (98–00) = 66.4; (01–02) = 7.3; Net export (96–97) = 24.6; (98–00) = 91.5; (01–02) = 115.4
    - Kyrgyz Republic:
      - GDP growth (96–97) = 8.5; (98–00) = 3.7; (01–02) = 2.5
      - Final consumption (96–97) = 4.9; (98–00) = 2.5; (01–02) = 2.8
      - Investment (96–97) = -1.0; (98–00) = 1.9; (01–02) = -0.5
      - Net export (96–97) = 4.6; (98–00) = -0.8; (01–02) = 0.1
      - Share of components (In percent of total): Final consumption (96–97) = 57.6; (98–00) = 68.6; (01–02) = 116.3; Investment (96–97) = -11.8; (98–00) = 52.4; (01–02) = -18.4; Net export (96–97) = 54.1; (98–00) = -21.0; (01–02) = 2.1
    - Moldova:
      - GDP growth (96–97) = -2.2; (98–00) = -2.6; (01–02) = 7.0
      - Final consumption (96–97) = 10.9; (98–00) = 0.6; (01–02) = 7.9
      - Investment (96–97) = -1.7; (98–00) = -1.0; (01–02) = 0.5
      - Net export (96–97) = -11.3; (98–00) = -2.1; (01–02) = -1.4
      - Share of components (In percent of total): Final consumption (96–97) = 495.5; (98–00) = 21.6; (01–02) = 119.4; Investment (96–97) = -77.3; (98–00) = -39.5; (01–02) = 7.4; Net export (96–97) = -513.6; (98–00) = -82.0; (01–02) = -26.8
  - Notes:
    - Final consumption was financed by retail trade, foreign remittances, and possibly underground economy income; government consumption by accumulated foreign borrowing, grants, and inflation financing (seignorage).
    - In the Kyrgyz Republic, data do not fully reflect export-driven growth because of a supply shock (landslide in the gold mine Kumtor) in 2002.
- Investment trends:
  - Investment share remains low in CIS-7 compared to other groups (Table 4 — Investment Ratios, 1999–2002, Average; in percent of GDP):
    - CIS-7 countries = 17.1
    - CIS-7 w/o Azerbaijan = 15.4
    - Other CIS countries = 20.7
    - CEEB = 22.7
    - OECD = 21.9
  - Exception: Azerbaijan — investment mostly in oil-producing industry accounted for 27.4 percent of GDP in 1999–2002.
- Exports:
  - Increased contribution of net exports to growth reflects high or rising share of exports in GDP (Table 5 — Export of Goods and Services, 1997–2002, In percent of GDP):
    - Armenia: 1997 = 20; 1998 = 21; 1999 = 23; 2000 = 25; 2001 = 25; 2002 = 30
    - Azerbaijan: 1997 = 29; 1998 = 23; 1999 = 28; 2000 = 39; 2001 = 41; 2002 = 44
    - Georgia: 1997 = 19; 1998 = 20; 1999 = 26; 2000 = 34; 2001 = 32; 2002 = 29
    - Kyrgyz Republic: 1997 = 38; 1998 = 36; 1999 = 42; 2000 = 42; 2001 = 37; 2002 = 37
    - Moldova: 1997 = 48; 1998 = 41; 1999 = 47; 2000 = 45; 2001 = 46; 2002 = 46
    - Tajikistan: 1997 = 72; 1998 = 49; 1999 = 64; 2000 = 86; 2001 = 69; 2002 = 64
    - Uzbekistan: 1997 = 27; 1998 = 23; 1999 = 18; 2000 = 25; 2001 = 28; 2002 = 31
  - Competitiveness:
    - Average monthly wages stayed below 40 dollars during 1997–2002, helping maintain high competitiveness.
  - Exports growth may be partially attributed to the Russian crisis in 1998 and the recent global economic slowdown, as other countries searched for cheaper suppliers.

### Growth accounting
- Goal: Explore the contribution of improvements in total factor productivity (TFP) to real GDP growth in 1998–2002.
- Methodology:
  - Uses a growth accounting framework following De Broeck and Koen (2000) with their dataset extended until 2002.
  - Assumes a neoclassical Cobb-Douglas production function: Yt = At Kt^α Lt^(1−α), where Yt = total output, At = total factor productivity index, Kt = capital stock, Lt = labor stock, and 0 < α < 1 is the elasticity of output with respect to capital.
  - Output growth decomposed into capital growth, labor growth, and the residual (TFP growth).
  - In the absence of factor price information, elasticities of output with respect to capital and labor are assumed constant over time and across countries and sum to 1.
- Data caveat:
  - Period of extension differs across countries due to employment data availability.
  - Because of data uncertainties, any point estimate of TFP is likely to be misleading.

*Source: _wp04151 - Section VI presents the results of the econometric analysis of the main factors that determine (PDF).*

### 0.3 and 0.7, respectively, which is consistent with previous studies on growth accounting

### _wp04151 - 0.3 and 0.7, respectively, which is consistent with previous studies on growth accounting

### Growth accounting framework and measurement of TFP
- TFP change (log) is computed from equation (2): ,)1( LKYA gggg αα −−−=  where gi is the growth rate of variable i.
- Capital accumulation uses the standard formula (3): ,)1( 1ttt IKK+−= − δ where It is investment at time t and δ is depreciation.
- The paper assumes a constant depreciation rate δ = 3 percent for the total capital stock (footnote: assumption consistent with De Broeck and Koen (2000)).
- Caveats on interpretation:
  - Measured TFP is a residual reflecting efficiency of input use and can capture changes in working hours, hidden employment, labor hoarding, capacity utilization, resource allocation, measurement accuracy, legalization of informal sector activity, and other factors — not purely exogenous technological progress.
  - Capacity utilization is not explicitly controlled for because no capacity utilization series for labor and capital are available for the CIS-7 countries.
  - Alternative depreciation scenarios (very high up to 70 percent in 1992 with 1 percent afterwards) yield qualitatively similar TFP and capital stock patterns to the constant 3 percent scenario.

### Assumptions and robustness checks
- Depreciation: constant 3 percent assumption; recognized as possibly low for some sectors but consistent with prior studies.
- Data limitations: 13-year span (including catastrophic systemic collapse and limited business cycles) constrains TFP inference; nonetheless, trends provide useful insights.
- Alternative calculations using extreme obsolescence/depreciation patterns do not materially change qualitative results for TFP and capital stock.

### Aggregate empirical findings (TFP, inputs, productivity)
- TFP changes were largely negative in the early 1990s and turned positive in the CIS-7 countries after 1993 as stabilization policies took hold.
- “As growth of factor inputs cannot account for the growth performance in the CIS-7 countries during the five-year time span 1993–2002, total factor productivity seems to be important.”
- By 2002:
  - Average capital productivity almost surpassed its 1993 level.
  - Average labor productivity as of 1999 remained far below the level in 1993.
- Average (CIS-7) summary figures (as presented in the source):
  - CIS-7 countries Avg. 1971-97: Output Growth Rate -0.4; Capital Growth Rate 3.7; Labor Growth Rate 1.3; TFP Growth Rate -2.4; Labor Productivity Growth Rate -1.8.
  - CIS-7 countries Avg. 1991-97: Output Growth Rate -10.5; Capital Growth Rate -0.1; Labor Growth Rate -1.3; TFP Growth Rate -9.5; Labor Productivity Growth Rate -7.9.
  - CIS-7 countries Avg. 1971-2000: Output Growth Rate 0.5; Capital Growth Rate 3.1; Labor Growth Rate 1.2; TFP Growth Rate -1.6; Labor Productivity Growth Rate -1.1.
  - CIS-7 countries Avg. 1991-2000: Output Growth Rate -4.3; Capital Growth Rate -0.1; Labor Growth Rate -0.6; TFP Growth Rate -3.5; Labor Productivity Growth Rate -3.8.
  - CIS-7 countries Avg. 1993-2000: Output Growth Rate -1.2; Capital Growth Rate -0.5; Labor Growth Rate -0.3; TFP Growth Rate -0.6; Labor Productivity Growth Rate -0.9.
  - CIS-7 countries Avg. 1998-2000: Output Growth Rate 4.3; Capital Growth Rate -0.3; Labor Growth Rate 0.6; TFP Growth Rate 4.6; Labor Productivity Growth Rate 4.1.
- Regression with changes in TFP (fixed-effect estimation, 1991–2002, CIS-7):
  - Unemployment rate coefficients: -0.627 (col 1), -0.81 (col 2), -0.944 (col 3). (t-statistics shown in parentheses in source; negative correlation.)
  - Average EBRD transition index: coefficient 12.427 (col 2).
  - Large-scale privatization: coefficients 3.439 (col 1) and 3.874 (col 3).
  - Infrastructure reform: coefficients 13.549 (col 1) and 12.636 (col 3).
  - Country fixed-effect coefficients (examples, col 2 and col 3): Armenia -16.95 and -20.16; Azerbaijan -27.79 and -26.59; Georgia -19.52 and -24.36; Kyrgyz Republic -28.98 and -23.78; Moldova -26.18 and -25.21; Tajikistan -22.77 and -21.61; Uzbekistan -23.24 and -22.35.
  - R-squared: 0.29 (col 1), 0.29 (col 2), 0.33 (col 3). Adjusted R-squared: 0.24, 0.19, 0.21 respectively.
  - Number of observations: 62 in each specification.
- Interpretation of regression: positive correlation between TFP changes and infrastructural reform and large-scale privatization; negative correlation with unemployment; negative and highly significant country fixed effects reflect adverse initial conditions.

### Sectoral decomposition and sectoral TFP
- Sectoral focus: industry, agriculture, construction, transport (including communications), and trade (including procurement). Services largely excluded due to data limitations.
- Sectoral shares and employment (selected highlights from Tables 8–10):
  - Agriculture provides >40 percent of labor force and >25 percent of GDP in the CIS-7 on average (textual statement).
  - Table 8, Armenia 1995: Agriculture 41; Industry 24; Other 25 (percent of GDP).
  - Table 8, Azerbaijan 2001: Industry 36; Agriculture 16; Other 31.
  - Table 10, Armenia 2000: Agriculture 44; Industry 14 (percent of employment).
- Investment by sector (selected highlights from Table 9):
  - Assumption note: sectoral investment shares reported as percent of total investment.
  - Azerbaijan industry investment shares: 1995 46.4; 1996 72; 1997 70.7; 1998 74.9; 1999 61.7; 2000 68.1; 2001 70.5; 2002 77 (percent).
  - Kyrgyz Republic industry investment shares: 1995 76.9; 1996 80; 1997 69.3; 1998 33.52; 1999 34.7; 2000 40.5; 2001 32.34; 2002 18.35 (percent).
- Sectoral TFP trends (Table 11 and Table 13 summary points):
  - Agriculture: average TFP growth positive but unstable; Moldova had highest agricultural TFP growth between 1998 and 2002 (breakup of collective farms partly explains).
  - Industry: TFP growth turned positive after 1998 in all CIS-7 except Moldova.
  - Trade: highest recent TFP growth among sectors — varied from 5 percent in the Kyrgyz Republic to 33 percent in Armenia (textual statement).
  - Transport: TFP continued to decline on average in all countries but the Kyrgyz Republic in 1998–2002.
  - Construction: average TFP growth resumed in recent years.
- Table 12: Share of low-degree-processed commodities in exports (1995 and latest available year, percent of total export):
  - Armenia: 1995 61.0; Latest 59.0.
  - Azerbaijan: 1995 64.5; Latest 90.7.
  - Georgia: 1995 30.1; Latest 44.3.
  - Kyrgyz Republic: 1995 36.2; Latest 73.5.
  - Moldova: 1995 62.8; Latest 76.6.
  - Tajikistan: 1995 59.0; Latest 84.0.
  - Uzbekistan: 1995 64.0; Latest 64.5.

### Econometric analysis of growth determinants (24 transition countries and CIS-7 subsample)
- Regression specification includes: country fixed effect, exchange rate regime dummy (ER), fiscal policy (general government budget deficit in percent of GDP), initial shock variable (IS defined in equation (4)), WarCIS7 dummy, and EBRD transition indices (small-scale privatization, large-scale privatization, governance and enterprise reforms, price liberalization, trade liberalization, competition policy, banking reform and interest rate liberalization, securities markets reform, infrastructure reform). EBRD indices range 1–4.
- Key aggregate regression findings (summary bullets from the source):
  - Initial shock variable appears insignificant in all specifications (fixed effects likely capture initial conditions).
  - De facto stability of the exchange rate regime and prudent fiscal policy exert a large positive effect on real GDP growth in the 24-country sample; the exchange rate dummy effect weakens when reform progress variables are included.
  - War dummy for CIS-7 is negative but insignificant.
  - Net effect of privatization on growth is positive overall; small-scale privatization is positive and significant; large-scale privatization shows negative or weakly significant effects in some specifications.
  - Price liberalization variable has a negative sign (reflecting timing: price liberalization often occurred during early transition years of output decline).
  - Reform of nonbank financial institutions is positive and significant at the 10 percent level.
- Selected coefficient examples from Table 14 (Sample of 24 transition countries, fixed effects):
  - Exchange rate dummy: 0.05 (col 1), 0.046 (col 2), 0.045 (col 3), 0.048 (col 4).
  - Fiscal: 0.47 (col 1), 0.533 (col 2), 0.511 (col 3), 0.493 (col 4).
  - Small-scale privatization: 0.03 (col 1), 0.038 (col 2), 0.037 (col 3), 0.046 (col 4).
  - Large-scale privatization: -0.01 (col 1) and negative in other specifications (e.g., -0.054, -0.047 in other columns).
  - R-squared (24-country sample): 0.58 (cols 1–3), 0.57 (col 4). Adjusted R-squared: 0.53 (cols 1–4).
  - Number of observations: 284 for 24-country regressions; 77 for CIS-7 regressions.

### Main conclusions
- Recovery drivers:
  - Early recovery driven by consumption.
  - Since 1998, growth in Armenia, Azerbaijan, Georgia, and Kyrgyz Republic driven by export expansion and investments in export-oriented sectors.
  - Consumption remained primary driver in Moldova, and possibly Tajikistan and Uzbekistan.
- Exports and vulnerabilities:
  - Exports resilient to global slowdown due to comparative advantages: high energy endowments in selected countries, artificially low energy prices, and low real wages.
  - Reliance on agricultural products, raw materials, and low-degree-processed goods increases vulnerability to world commodity price changes and weather.
- Productivity and reform:
  - Since 1998, overall TFP appears to be increasing in all CIS-7 countries — consistent with reform effects on production — but interpretation must account for capacity utilization, improved statistical coverage of private sector, and hidden employment changes.
  - Leading growth sectors: agriculture, industry, and trade; industrial growth concentrated in export-oriented raw-material sectors.
- Investment and sustainability:
  - Share of investment in GDP rising but remains very low compared to rest of CIS and CEEB (exception: petroleum sector in Azerbaijan).
  - Low investment is insufficient for long-term sustainable growth.
- Agriculture:
  - Output recovery in agriculture may be transitory; agriculture shows increasing labor intensity, declining share of investment, and possible movement toward subsistence farming in poorest CIS countries.
- Trade potential and constraints:
  - Growth prospects could be improved by realizing trade potential: reducing restrictive trade policies, repairing physical infrastructure, and addressing corruption and governance problems in customs and transport services.
  - Insufficient foreign investment limits diversification and improvement of trade environment.

### Policy recommendations (as stated in the source)
- Continue trade liberalization, with emphasis on:
  - Reducing corruption and improving governance in customs and transport services.
  - Removing man-made barriers to trade.
  - Facilitating development of physical infrastructure conducive to trade, including regional transportation networks to improve access to world markets.
- Improve general investment climate to attract new investment needed for diversification and sustainable growth.
- Continue structural reforms to support diversification away from export concentration in raw materials and low-degree-processed goods.

*Source: authors’ estimates and tables as presented in the provided IMF working paper content.*

### APPENDIX                                                                        I

### _wp04151 - APPENDIX I

### Data Description and Data Issues
- Variables and definitions:
  - GDP growth: Percentage change of GDP in constant prices. Source: WEO (2003), data from IMF economists working on the countries.
  - Exchange rate dummy: Dummy for “stable” exchange rate regime, 1 = stable rate, 0 = floating rate. The dummy for the “stable” exchange rate arrangement is defined as for fixed exchange rate in classification of Reinhart and Rogoff (2004). That is, the exchange rate is defined as “stable,” if there is de facto peg or bands (2–5 percent). Source: Authors’ estimates.
  - Fiscal: General government budget balance in percent of GDP. Source: WEO, data from IMF economists working on the countries.
  - Avg Index: Average of all EBRD liberalization indices. Source: Authors’ estimates.
  - EBRD indices: Bank Reform, Comp Policy, Ent Reform, Infr Reform, Price Liberalization, Trade Liberalization, Large Privatization, Small Privatization. Source: EBRD, 2002.
- Source note: "Sources: IMF staff and national authorities estimations."
- Measurement and conceptual issues emphasized:
  - Official data are subject to conceptual and measurement problems; statistical discrepancies were large for some countries, especially early transition years.
  - For some countries, statistical discrepancies are captured in different expenditure items (example: Uzbekistan—“Investments” contains a large sub-item “Inventory” used to balance statistical discrepancies).
  - Tajikistan: data on GDP by expenditure available only from 1998 onwards; nominal decomposition available, real decomposition work is recent.
  - In absence of reliable official data, staff often derive expenditure components with consumption as a residual.
  - Likely biases: inaccurate measurement of the shadow economy more likely to affect final consumption than investment; incentives suggest possible overstatement of investment and understatement of consumption.
  - Contributions of expenditure components to real GDP growth: final consumption and gross capital formation contributions calculated from constant prices data from the WEO; net exports contribution calculated as residual difference because data on net exports in constant prices are not available.

### GDP by Expenditure — Key Averages and Patterns (Table A4 and excerpts from Table A1)
- Table A4: Components of GDP by Expenditures, averages for 1994–98 and 1999–02 (percent)
  - Armenia:
    - Consumption (average): 100.8
    - Private (94–98): 11.4
    - Public (94–98): 86.0
    - Fixed investment private (99–02): 10.3
    - Fixed investment public (99–02): 15.9
    - Net exports (94–98): -35.9
    - Net exports (99–02): -22.6
  - Azerbaijan:
    - Consumption (average): 84.1
    - Private (94–98): 13.4
    - Public (94–98): 65.5
    - Fixed investment private (99–02): 13.6
    - Fixed investment public (99–02): 23.0
    - Net exports (94–98): -21.6
    - Net exports (99–02): -4.3
  - Georgia:
    - Consumption (average): 95.7 (data series starts at 1996; first average for 1996–98)
    - Private (94–98): 10.6
    - Public (94–98): 85.8
    - Fixed investment private (99–02): 10.7
    - Fixed investment public (99–02): 10.4
    - Net exports (94–98): -22.1
    - Net exports (99–02): -15.9
  - Kyrgyz Republic:
    - Consumption (average): 78.5
    - Private (94–98): 18.4
    - Public (94–98): 68.9
    - Fixed investment private (99–02): 18.7
    - Fixed investment public (99–02): 16.1
    - Net exports (94–98): -14.9
    - Net exports (99–02): -6.1
  - Moldova:
    - Consumption (average): 63.7
    - Private (94–98): 26.5
    - Public (94–98): 83.1
    - Fixed investment private (99–02): 16.1
    - Fixed investment public (99–02): 19.4
    - Net exports (94–98): -15.7
    - Net exports (99–02): -22.4
  - Tajikistan (data series for 1998–2001; averages adjusted accordingly):
    - Consumption (average): 67.4
    - Private (94–98): 9.3
    - Fixed investment private (99–02): 72.3
    - Fixed investment public (99–02): 9.1
    - Fixed investment (94–98): 13.4
    - Net exports (94–98): -12.2
    - Net exports (99–02): -8.0
  - Uzbekistan:
    - Consumption (average): 56.7
    - Private (94–98): 29.0
    - Public (94–98): 66.1
    - Fixed investment private (99–02): 22.8
    - Fixed investment public (99–02): 16.5
    - Fixed investment (99–02): 11.0
    - Net exports (94–98): -2.2
    - Net exports (99–02): 1.5
- Table A1 (selected country-year patterns; percent of GDP):
  - Azerbaijan: Final consumption ranged from 95.7 (1993) to 74.4 (2002); Gross fixed capital formation ranged from 14.9 (1995) to 36.0 (1997); Net exports ranged from -31.9 (1998) to 3.6 (2001).
  - Armenia: Final consumption peaked at 176.8 (1993) and was 96.7 (2002); Gross fixed capital formation ranged from 7.7 (1993) to 20.2 (1994); Net exports improved from -43.1 (1993) to -15.1 (2002).
  - Georgia: Series starts at 1996; Final consumption was 112.4 (1996) and 97.4 (2002); Gross fixed capital formation was 9.7 (1996) and 16.7 (2002); Net exports ranged from -26.2 (1997) to -11.6 (2002).
  - Kyrgyz Republic: Final consumption 96.0 (1993) to 85.6 (2002); Gross fixed capital formation 13.3 (1993) to 17.2 (2002); Net exports from -25.8 (1996) to -0.3 (2001).
  - Moldova: Final consumption 75.4 (1994) to 102.7 (2002); Gross fixed capital formation 19.3 (1994) to 16.6 (2002); Net exports from -26.8 (1998) to -7.8 (1995).
  - Tajikistan: GDP by expenditure available from 1998; Final consumption 76.7 (1998) to 81.5 (2001); Gross fixed capital formation 13.4 (1998) to 9.2 (2001); Net exports -7.8 (1998) to 10.2 (2000).
  - Uzbekistan: Final consumption 105.2 (1993) to 86.2 (2002); Gross fixed capital formation 2.9 (1993) to 11.0 (1999–2002); Net exports -9.4 (1996) to 3.1 (2000).

### Total Factor Productivity (TFP) and Labor Productivity (Tables A2 and A3)
- Table A2. Annual TFP Growth Rates, 1991–2002 (selected country-year values; percent)
  - Armenia: 1992 = -10.2; 1993 = -59.0; 1994 = -7.4; 1995 = 12.9; 1996 = 5.7; 1997 = 5.1; 1998 = 7.2; 1999 = 9.6; 2000 = 6.1; 2001 = 7.7; 2002 = n.a.
  - Azerbaijan: 1991 = -7.9; 1992 = -21.0; 1993 = -18.3; 1994 = -17.5; 1995 = -12.2; 1996 = -1.5; 1997 = 0.6; 1998 = -9.1; 1999 = 5.9; 2000 = 9.7; 2001 = n.a.; 2002 = n.a.
  - Georgia: 1991 = -18.3; 1992 = -37.0; 1993 = -18.2; 1994 = -9.4; 1995 = -5.6; 1996 = 10.8; 1997 = 6.4; 1998 = 18.4; 1999 = 5.4; 2000 = -0.8; 2001 = 3.6; 2002 = 6.3.
  - Kyrgyz Republic: 1991 = -6.9; 1992 = -24.8; 1993 = -10.7; 1994 = -26.3; 1995 = -9.6; 1996 = 3.0; 1997 = 9.6; 1998 = 1.2; 1999 = 0.7; 2000 = 4.3; 2001 = 3.7; 2002 = -3.6.
  - Moldova: 1991 = -21.4; 1992 = -34.4; 1993 = 8.9; 1994 = -36.7; 1995 = -6.6; 1996 = -5.4; 1997 = 1.3; 1998 = -6.0; 1999 = 3.8; 2000 = 4.7; 2001 = 8.3; 2002 = 0.3.
  - Tajikistan: 1991 = -7.6; 1992 = -25.5; 1993 = -18.9; 1994 = -12.5; 1995 = -11.0; 1996 = -13.2; 1997 = -1.2; 1998 = 5.4; 1999 = 6.3; 2000 = 8.2; 2001 = 7.0; 2002 = n.a.
  - Uzbekistan: 1991 = -4.2; 1992 = -9.5; 1993 = -2.2; 1994 = -5.3; 1995 = -2.4; 1996 = -1.1; 1997 = 3.1; 1998 = 3.4; 1999 = 4.1; 2000 = n.a.; 2001 = n.a.; 2002 = n.a.
  - Sources noted: De Broeck and Koen (2000); and authors’ estimates.
- Table A3. Annual Labor Productivity Growth, 1991–2002 (selected country-year values; percent)
  - Armenia: 1991 = -13.9; 1992 = -38.4; 1993 = -6.7; 1994 = 9.3; 1995 = 7.7; 1996 = 8.9; 1997 = 7.9; 1998 = 10.1; 1999 = 6.4; 2000 = 7.7; 2001 = n.a.; 2002 = n.a.
  - Azerbaijan: 1991 = -5.2; 1992 = -19.3; 1993 = -23.0; 1994 = -17.9; 1995 = -11.4; 1996 = -0.7; 1997 = 5.5; 1998 = -13.8; 1999 = 7.4; 2000 = 11.0; 2001 = n.a.; 2002 = n.a.
  - Georgia: 1991 = -13.4; 1992 = -30.1; 1993 = -21.8; 1994 = -8.2; 1995 = -15.8; 1996 = 10.8; 1997 = 6.5; 1998 = 27.2; 1999 = 5.9; 2000 = -2.7; 2001 = 2.4; 2002 = 6.1.
  - Kyrgyz Republic: 1991 = -6.9; 1992 = -18.8; 1993 = -7.7; 1994 = -18.4; 1995 = -5.2; 1996 = 6.5; 1997 = 7.9; 1998 = 1.2; 1999 = 0.2; 2000 = 5.2; 2001 = 4.2; 2002 = -3.9.
  - Moldova: 1991 = -17.4; 1992 = -28.3; 1993 = 20.0; 1994 = -30.6; 1995 = -1.4; 1996 = -7.1; 1997 = 2.2; 1998 = -6.3; 1999 = 6.1; 2000 = 0.8; 2001 = 7.2; 2002 = 6.8.
  - Tajikistan: 1991 = -9.9; 1992 = -27.7; 1993 = -14.9; 1994 = -11.8; 1995 = -12.8; 1996 = -10.9; 1997 = -1.7; 1998 = 5.0; 1999 = 7.2; 2000 = 7.8; 2001 = 5.1; 2002 = n.a.
  - Uzbekistan: 1991 = -5.1; 1992 = -10.5; 1993 = -2.2; 1994 = -6.6; 1995 = -2.1; 1996 = 0.4; 1997 = 3.7; 1998 = 3.0; 1999 = 4.0; 2000 = n.a.; 2001 = n.a.; 2002 = n.a.
  - Source: Authors’ estimates.

### Foreign Direct Investment — CIS-7 (Table A5)
- Foreign Direct Investment in CIS-7, 1996–2002 (In percent of GDP)
  - Armenia: 1996 = 1.1; 1997 = 3.2; 1998 = 11.6; 1999 = 6.6; 2000 = 5.5; 2001 = 3.3; 2002 = 4.6.
  - Azerbaijan: 1996 = 20.8; 1997 = 27.4; 1998 = 23.0; 1999 = 11.1; 2000 = 2.8; 2001 = 5.2; 2002 = 17.1.
  - Georgia: 1996 = 1.8; 1997 = 6.6; 1998 = 6.1; 1999 = 2.2; 2000 = 5.0; 2001 = 2.5; 2002 = 3.8.
  - Kyrgyz Republic: 1996 = 2.6; 1997 = 4.7; 1998 = 5.3; 1999 = 3.1; 2000 = -0.5; 2001 = -0.1; 2002 = 1.0.
  - Moldova: 1996 = 1.4; 1997 = 4.1; 1998 = 4.4; 1999 = 3.3; 2000 = 11.1; 2001 = 10.5; 2002 = 6.6.
  - Tajikistan: 1996 = 1.7; 1997 = 1.6; 1998 = 1.9; 1999 = 1.9; 2000 = 2.4; 2001 = 0.9; 2002 = 3.0.
  - Uzbekistan: 1996 = 0.7; 1997 = 1.1; 1998 = 0.9; 1999 = 0.7; 2000 = 0.5; 2001 = 0.7; 2002 = 0.7.
  - CIS-7, average: 1996 = 4.3; 1997 = 6.9; 1998 = 7.6; 1999 = 4.1; 2000 = 3.8; 2001 = 3.3; 2002 = 5.3.
  - CIS-7, excluding Azerbaijan: 1996 = 1.5; 1997 = 3.6; 1998 = 5.0; 1999 = 3.0; 2000 = 4.0; 2001 = 3.0; 2002 = 3.3.
  - Other CIS, average: 1996 = 2.4; 1997 = 2.8; 1998 = 2.4; 1999 = 3.7; 2000 = 2.4; 2001 = 4.1; 2002 = 3.4.
  - Source: IMF: FSU database.

### Statistical Issues in Analysis of GDP by Expenditure Components (Box 1)
- Key issues highlighted:
  - Decomposition based on official data subject to conceptual and measurement problems; large statistical discrepancies in early transition years.
  - Different countries capture discrepancies in different expenditure items; example: Uzbekistan’s large “Inventory” sub-item within “Investments” used for balancing.
  - Tajikistan: real expenditures available only from 1998; nominal decomposition used earlier.
  - When official expenditure data unreliable, staff derive components with consumption as residual.
  - Biases likely: underreporting of the shadow economy likely reduces measured final consumption; incentives exist to overstate investment (higher depreciation allowances), implying potential understatement of consumption.
  - Real contribution computation method: contributions of final consumption and gross capital formation calculated from constant-price series in the WEO; net exports contribution computed as residual because net exports in constant prices are not available (current-price exports/imports and deflators produce inconsistent real growth results).

*Source: _wp04151 - APPENDIX I (PDF content provided).*

### Chapter 4 in Economic Survey of Europe, No. 2/3.

### _wp04151 - Chapter 4 in Economic Survey of Europe, No. 2/3.

### References cited in Chapter 4
- Levine, R., and D. Renelt, 1992, “A Sensitivity Analysis of Cross-Country Growth Regressions,” American Economic Review, Vol. 82 (September), pp. 942–963.
- Loukoianova, E. and A.Unigovskaya, 2004, “Growth in the CIS-7: Recent Developments and Prospects,” a background paper to the IMF Board paper “Recent Policies and Performance of the Low-Income CIS Countries—An Update of the CIS-7 Initiative” SM/04/150 (April 24) (Washington: International Monetary Fund), also available via the Internet: www.cis7.org.
- Raiser, M., M. Schaffer, and J. Schuchhardt, 2003, “Benchmarking Structural Change in Transition,” CERT Discussion Paper 2003/01.
- Reinhart, C., and K. Rogoff, 2002, “The Modern History of Exchange Rate Arrangements: A Reinterpretation,” NBER Working Paper No. 8963 (Cambridge, Mass.: National Bureau of Economic Research).
- Sala-I-Martin, 1997, “I Just Ran Four Million Regressions,” American Economic Review, Papers and Proceedings, Vol. 87, pp. 178–183.
- Sarel, M., 1997, “Growth and Productivity in ASEAN Countries,” IMF Working Paper 97/97 (Washington: International Monetary Fund).
- Senhadji, A., 1999, “Sources of Economic Growth: An Extensive Growth Accounting Exercise,” IMF Working Paper 99/77 (Washington: International Monetary Fund).
- Staehr, K., 2003, “Reforms and Economic Growth in Transition: Complementarity, Sequencing and Speed,” BOFIT Discussion Paper 2003/01 (Helsinki: Bank of Finland Institute for Economies in Transition).
- Zettelmeyer, J., 1998, “The Uzbek Growth Puzzle,” IMF Working Paper 98/133 (Washington: International Monetary Fund).

*Chapter 4 references, Economic Survey of Europe, No. 2/3.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04151.pdf_