## _wp11163

## Source details

**Canonical URL:** [_wp11163](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11163.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11163.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11163.pdf.json)

---

### Key historical growth and reform context
- India: growth rate of 3.5 per cent per annum during 1950/51 to 1979/80; about 5.5 per cent per annum during the 1980s (gradual trade and industrial liberalization and possibly fiscal expansion).
- Post-1990/1991 BOP crisis: deep liberalization of domestic and external policies; growth rate remained in the 5.5-5.8 per cent range during the 1990s.
- Puzzle: limited reforms in the 1980s coincided with a 2 percentage point higher growth rate compared with the relatively major reforms of the 1990s, which had virtually no measurable effect on the trend.

### Analytical framework: J-curve, S-curve, and hybrid hypotheses
- J-curve hypothesis (Virmani): major/large trade (import) liberalization → initial decline in measured productivity and output growth followed by recovery and eventual improvement.
- S-curve hypothesis: gradual/marginal reforms produce conventional S-curve of productivity and growth (initial adjustment then steady higher growth).
- Hybrid (S-J) case: early positive conventional effects followed by slower negative effects or later J-curve recovery; sectoral heterogeneity expected.
- Mechanisms for initial slowdown under major liberalization:
  - deterioration in capacity utilization from obsolescence of product lines and capital,
  - gradual adoption and diffusion of new technology (learning lags),
  - diversion of human resources to learn new technology and markets,
  - labor law rigidities limiting immediate gains from factor reallocation.

### Data, sample, periodization, and variable construction
- Study period: 1980-81 to 2007-08; organized manufacturing analyzed 1981/82 to 2007/08.
- Two broad periods:
  - Period I (pre-1990s reforms): 1981/82 to 1990/91
  - Period II (1991/92 to 2007/08), subdivided:
    - Sub-period 1: 1991/92 to 1997/98
    - Sub-period 2: 1998/99 to 2001/02
    - Sub-period 3: 2002/03 to 2007/08
- Sample: 12 two-digit ASI manufacturing industries comprising 89.1% share in output; remaining 10 two-digit industries combined as ‘Others’ (10.1% share).
- Variable construction and data sources:
  - Deflation: 1993/94=100 price indices; WPI proxies where exact deflators unavailable.
  - Capital stock via Perpetual Inventory method with benchmark for 1973/74; depreciation fixed at 5 per cent per annum (life of capital stock 20 years).
  - Labor input: total persons engaged (ASI); price of labor = total emoluments / total persons engaged; emoluments deflated by price of labor (1993/94=100).
  - Energy, materials, services: fuel and materials consumed (ASI) deflated by weighted indices; services obtained by subtracting materials and energy expenses from total inputs; services deflated using implicit price deflators from NAS and CSO Input-Output matrix 1993/94.
  - Primary data sources: Annual Survey of Industries (ASI) via EPW database (1981/82–2003/04) and CSO for remaining years; WPI series from Ministry of Statistics and Program Implementation; Reserve Bank of India Bulletin (1976) for gross net ratio (GNR) to obtain gross capital stock for 1973/74; NAS for implicit price deflators.

### Structure of costs and input shares (aggregate manufacturing)
- Average shares in total input costs (study period):
  - Materials: 62.9 per cent
  - Capital: 14.5 per cent
  - Services: 9.7 per cent
  - Energy: 6.4 per cent
  - Labor: 6.5 per cent
- Trends (average annual rates in parentheses where provided):
  - Capital share increased at rate of 0.9 per cent per annum.
  - Services share rose from 8.0% in Period I to 10.7% in Period II at an average annual rate of 3.1 per cent.
  - Labor share declined at an average annual rate of -3.2 per cent.
  - Energy share declined at -0.9 per cent per annum.
  - Materials share essentially constant (62.9 per cent; average annual growth -0.1).
- Selected period-specific shares (Period I; Period II; Study Period):
  - Period I: Capital 13.0 (1.8), Labor 8.3 (-2.9), Energy 6.8 (0.4), Materials 63.9 (0.5), Services 8.0 (-3.6)
  - Period II: Capital 15.3 (-0.3), Labor 5.5 (-3.4), Energy 6.2 (-1.9), Materials 62.3 (0.1), Services 10.7 (3.1)
  - Study Period: Capital 14.5 (0.9), Labor 6.5 (-3.2), Energy 6.4 (-0.9), Materials 62.9 (-0.1), Services 9.7 (2.1)

### Trends in output and input growth (constant prices, 1993/94 = 100)
- Output growth (per cent per annum):
  - Period I: 7.7 (1981/82 - 1990/91)
  - Period II: 8.2 (1991/92 - 2007/08)
  - Sub-period 1: 7.4
  - Sub-period 2: 2.7
  - Sub-period 3: 12.9
  - Study Period (1981/82 - 2007/08): 8.1 (note table lists Study Prd. as (8.1 1981/82 - 2007/08))
- Input growth (Study Period averages):
  - Materials: 8.2 per cent
  - Capital: 7.7 per cent
  - Services: 7.5 per cent
  - Energy: 5.1 per cent
  - Labor: 1.6 per cent
- Selected period-specific input growth (per cent per annum):
  - Prd. I: Output 7.7; Capital 9.1; Labour 0.7; Energy 8.0; Materials 8.1; Services 2.7
  - Prd. II: Output 8.2; Capital 7.0; Labour 2.1; Energy 3.6; Materials 8.3; Services 10.1
  - Sub-prd.1: Output 7.4; Capital 8.7; Labour 2.3; Energy 3.8; Materials 6.3; Services 14.7
  - Sub-prd.2: Output 2.7; Capital 5.0; Labour -3.3; Energy -3.2; Materials 4.1; Services 0.0
  - Sub-prd.3: Output 12.9; Capital 6.3; Labour 5.4; Energy 8.0; Materials 13.4; Services 11.4
  - Study Prd.: Output (8.1 1981/82 - 2007/08); Capital 7.7; Labour 1.6; Energy 5.1; Materials 8.2; Services 7.5

### Methodology for TFPG estimation
- Chosen approach: growth accounting using Translog index (discrete approximation of Translog production function) rather than econometric estimation to avoid multicollinearity, autocorrelation, and large-sample problems.
- Translog properties used:
  - Allows variability in elasticity of substitution between factors.
  - Does not assume Hicks-neutral technological progress.
- Inputs in Translog: K (capital), L (labor), E (energy), M (materials), S (services); SK, SL, SE, SM, SS denote income shares summing to unity.
- Rationale: TFP growth captures productivity gains (technological progress, scale, learning-by-doing, R&D, technical efficiency) distinct from output growth driven by larger input application.

### Empirical findings — Partial factor productivity (aggregate and sectoral)
- Aggregate partial factor productivity growth (study period 1981/82 - 2007/08):
  - Capital (Y/K): 0.1
  - Labor (Y/L): 6.6
  - Energy (Y/E): 3.3
  - Materials (Y/M): 0.1
  - Services (Y/S): -1.7
- Period comparisons (aggregate):
  - Period I (1981/82 - 1990/91): Capital -1.1; Labor 7.3; Energy -0.2; Materials -0.8; Services 4.9
  - Period II (1991/92 - 2007/08): Capital 2.0; Labor 7.3; Energy 5.5; Materials 0.3; Services -3.0
  - Sub-prd.1 (1991/92 - 1997/98): Capital -0.4; Labor 6.1; Energy 4.5; Materials 1.5; Services -4.7
  - Sub-prd.2 (1998/99 - 2001/02): Capital 0.5; Labor 6.7; Energy 6.3; Materials -2.7; Services 3.0
  - Sub-prd.3 (2002/03 - 2007/08): Capital 6.2; Labor 6.7; Energy 6.9; Materials -1.3; Services 2.5
- Patterns and interpretation:
  - Labor and energy follow S-curve patterns: steady/accelerating productivity (labor steady then accelerates; energy from negative in 1980s to strong positive in 1990s).
  - Capital and materials show hybrid S-J patterns: acceleration, deceleration, then strong acceleration in third sub-period.
  - Services exhibit a J-curve pattern: productivity decelerated sharply in early post-reform years then recovered from late 1990s onward.
- Sectoral PFP examples (study-period values):
  - Capital PFP: Food products & beverages -1.6; Textiles -0.5; Coke, petroleum products & nuclear fuel -3.7; Chemicals 0.5; Fabricated metal products 1.5; Others 0.0
  - Labor PFP: Food products 5.2; Textiles 7.2; Coke, petroleum 4.7; Chemicals 5.7; Motor vehicles 8.6; Others 5.8
  - Energy PFP: Food products 1.5; Textiles 3.1; Coke, petroleum 0.2; Chemicals 5.0; Non-metallic mineral products 4.1; Others 3.6
  - Materials PFP: Food products 0.8; Textiles 0.1; Coke, petroleum -2.4; Chemicals -0.7; Motor vehicles 2.2; Others 0.1
  - Services PFP: Food products -4.7; Textiles -0.2; Coke, petroleum -3.9; Chemicals -0.9; Motor vehicles -1.8; Others 2.4

### Empirical findings — Total Factor Productivity Growth (TFPG)
- Aggregate TFPG (All Manufactures):
  - Period I (1981/82 - 1990/91): 0.61 per cent per annum
  - Sub-period 1 (1991/92 - 1997/98): 0.25
  - Sub-period 2 (1998/99 - 2001/02): -0.09
  - Sub-period 3 (2002/03 - 2007/08): 1.41
  - 1981/82 - 2007/08 (All/Total): 0.59
- Interpretation:
  - Aggregate manufacturing TFPG shows a clear J-curve: modest growth in the 1980s (0.61), slowdown to 0.25 in the first sub-period, decline to -0.09 in the second sub-period, sharp acceleration to 1.41 in the third sub-period (more than twice the 1980s TFPG).
  - Attributed causes: BOP crisis and 1991 shock; dramatic import liberalization (removal of QRs on capital goods and intermediates; tariff reductions); exchange rate reforms; obsolescence of some capital goods and temporary falls in capacity utilization; delayed diffusion of new technology; labor market rigidities.
- Sectoral TFPG patterns (among 12 major sub-sectors):
  - 2 followed S-curve pattern.
  - 5 followed J-curve pattern.
  - 5 followed hybrid S-J pattern.
- Selected sub-sector TFPG sequences (example patterns):
  - Textiles: 1.2; 0.49; 1.2; 1.8 — J-curve (trough in first sub-period).
  - Chemicals & chemical products: 1.7; -1.25; -0.3; 1.7 — J-curve.
  - Food products & beverages: 1.98; -0.56; -0.24; 0.95 — J-curve with slow recovery.
  - Motor vehicles & trailers: 0.57; 1.09; -2.27; 4.02 — S=>J with large later acceleration.
  - Coke, petroleum & nuclear fuel: -4.70; -3.99; -0.47; 2.39 — S-curve with progressive improvement.
- “Others” (minor industries) aggregate TFPG (1981/82-2007/08): 0.19

### Sources and magnitude of aggregate manufacturing output growth (Table 9 highlights)
- Study period: 1981/82 - 2007/08 — Growth in Output: 8.06
  - K: 1.12 (13.8)
  - L: 0.08 (1.1)
  - E: 0.33 (4.1)
  - M: 5.19 (64.4)
  - S: 0.75 (9.3)
  - TFPG: 0.59 (7.3)
- Prd I: 1981/82 - 1990/91 — Growth in Output: 7.74
  - K: 1.19 (15.3)
  - L: 0.05 (0.6)
  - E: 0.54 (6.9)
  - M: 5.21 (67.4)
  - S: 0.14 (1.8)
  - TFPG: 0.61 (7.9)
- Prd II: 1991/92 - 2007/08 — Growth in Output: 8.23
  - K: 1.08 (13.1)
  - L: 0.10 (1.3)
  - E: 0.22 (2.6)
  - M: 5.18 (63.0)
  - S: 1.07 (13.0)
  - TFPG: 0.58 (7.0)
- Sub-prd 1: 1991/92 - 1997/98 — Growth in Output: 7.41
  - K: 1.37 (18.5)
  - L: 0.15 (2.1)
  - E: 0.25 (3.3)
  - M: 3.98 (53.6)
  - S: 1.42 (19.1)
  - TFPG: 0.25 (3.4)
- Sub-prd 2: 1998/99 - 2001/02 — Growth in Output: 2.67
  - K: 0.77 (28.9)
  - L: -0.18 (-6.8)
  - E: -0.21 (-7.7)
  - M: 2.38 (88.8)
  - S: 0.01 (0.3)
  - TFPG: -0.09 (-3.5)
- Sub-prd 3: 2002/03 - 2007/08 — Growth in Output: 12.88
  - K: 0.94 (7.3)
  - L: 0.24 (1.9)
  - E: 0.46 (3.6)
  - M: 8.46 (65.7)
  - S: 1.36 (10.6)
  - TFPG: 1.41 (11.0)
- Materials accounted for more than 60 per cent of the output growth over the study period; services contributed 12.7 percent; capital contributed 12.3 percent; labor and energy together contributed less than 6 per cent.
- Improvement in capacity utilization indicated by a contribution of 0.9 per cent during the third sub period even as its share in total output growth fell.

### Heterogeneity, sequencing, institutional constraints, and reform composition
- TFPG and output patterns vary across sub-sectors reflecting differing exposure to reforms, timing/sequencing of reforms, and initial technology gaps.
- Sequencing effects: e.g., elimination of QRs on intermediates and capital goods early 1990s while QRs on manufactured consumer goods persisted until late 1990s–early 2000s, producing sectoral differences (some showing hybrid S-J or weak-J outcomes).
- Public sector dominance, de-licensing, privatization and timing of access to dis-embodied technology/FDI influenced sub-sectoral TFPG trajectories (examples: Electrical Machinery, Basic Metals, Coke, Petroleum & Nuclear Fuel).
- Institutional constraints and incomplete complementary reforms:
  - Imbalance in the 1990s: increase in competitive pressure exceeded increases in access to the means to compete (labor, risk capital, debt finance, bankruptcy law, technology reforms not fully implemented).
  - Rigid labour laws and constrained labor mobility reduced incentives for reallocation in firms with 100 or more workers; small-scale, labor-intensive industries faced different incentive structures.
- Aggregate classification of TFPG patterns across twenty two sub-sectors:
  - 3 sub-sectors followed an S-curve pattern (14%).
  - 8 sub-sectors followed a J-curve pattern (36%).
  - 10 sub-sectors followed a hybrid S-J pattern (45%).
  - 1 sub-sector followed a pattern of falling productivity (5%).
- Conclusion: the hypothesis that reforms and opening damaged Indian manufacturing is decisively rejected.

### Mechanisms linking import liberalization and measured productivity (references from appendix)
- Major import liberalization changes relative prices toward world prices; previously highly protected goods may see temporary falls in prices and output.
- Measured productivity can fall if capital immobility and capacity underutilization persist while capital stock measures do not adjust for utilization.
- Laspeyres-type quantity indices may understate growth during reform periods; base-year shifts to post-reform years can raise measured post-reform growth.
- Adoption of unfamiliar/new technologies and required complementary inputs produce diffusion lags (S-curve) and can produce an initial negative effect on measured productivity (J-curve).
- Competition aspects: freedom to compete, pressure to compete, and means to compete; broad import liberalization increases both pressure and access to means (inputs, capital goods, technology, FDI spillovers) but outcomes depend on complementary reforms and institutional capacity.

*Source: _wp11163*

### References                    28

### _wp11163 - References 28

### Key historical growth and reform context
- India moved from a growth rate of 3.5 per cent per annum during 1950/51 to 1979/80 to a growth rate of about 5.5 per cent per annum during the 1980s because of gradual trade and industrial liberalization and possibly fiscal expansion.
- Following the BOP crisis of 1990-1991, India undertook a deep and wide ranging liberalization of domestic and external policies; however, the growth rate barely moved from the 5.5-5.8 per cent range during the 1990s [Virmani (2004, 2005a 2006b)].
- Puzzle identified: limited reforms of the 1980s coincided with a 2 percentage point higher growth rate compared with the relatively major reforms of the 1990s, which had virtually no measurable effect on the growth trend.

### Findings on manufacturing productivity and capacity utilization
- Majority of studies found that productivity growth in post-reforms period of 1990s decelerated from growth rates seen in the 1980s.
- Alternative explanation focusing on capacity utilization:
  - Surge in investment activities and imports in the post-reforms period, unaccompanied by commensurate expansion of demand, led to worsening capacity utilization in manufacturing, adversely affecting productivity growth [Uchikawa (2001); Goldar & Kumari (2003)].
  - Goldar & Kumari (2003) evidence:
    - Ratio of gross fixed capital formation to gross value added (at 1993/94 prices) in organized manufacturing:
      - 44 per cent during 1985-86 to 1989/90
      - 76 per cent during 1995/96 to 1997/98
    - Ratio of gross capital stock to gross value of output (at 1993-94 prices):
      - average of 78.6 per cent between 1992/93 and 1997/98
      - 83.7 per cent between 1998/99 and 2001-02
      - declined sharply to 61.2 per cent during 2002/03 to 2007/08
  - Even after adjusting for capacity utilization, Goldar & Kumari (2003) found productivity growth in 1990s stayed at nearly the same level as during 1980s.
- Explanation of delayed effects:
  - Athukorala & Rajapatirana (2000) and others argue productivity-enhancing effects of economic reforms may be manifested with a time lag; hence an improvement in productivity growth could occur in later years.
  - Virmani (2009) showed macro-level productivity growth pattern from 1990s reforms was in line with the J-curve hypothesis (see appendix).
  - Estimates by Virmani and Hashim (2009) and Hashim et al. (2009) supported J-curve conclusions within data limitations.

### J-curve, S-curve, and alternative scenarios for productivity response
- J-curve hypothesis: initial decline in productivity growth following liberalization, followed by recovery and improvement over time.
- S-curve possibility: productivity and output growth during early sub-periods may be higher (or lower) than later sub-periods if conventional S-curve effects predominate.
- Hybrid case: negative effects slower to take effect than positive conventional ones, yielding an initial S-curve effect followed by a J-curve one.
- Analogous experiences: Eastern European countries and Russia after the fall of the Berlin wall illustrated J-curve at economy level, though those cases involved systemic political-administrative and institutional shocks that complicate separation of institutional-change effects from policy-change effects (Blanchard (1997), Kvintradze (2010)).

### Data, study scope, and periodization
- Present study estimates productivity growth in Indian organized manufacturing industries using data from 1981/82 to 2007/08.
- Two broad periods:
  - Period I (pre-1990s reforms): 1981/82 to 1990/91
  - Period II (reform / post-reform): 1991/92 to 2007/08
- Post-reform period subdivided into three sub-periods (SPs) based on phases of output growth and capacity utilization:
  - Sub-period 1: 1991/92 to 1997/98
  - Sub-period 2: 1998/99 to 2001/02
  - Sub-period 3: 2002/03 to 2007/08
- Expected empirical patterns:
  - Productivity and output growth during the first or first two sub-periods is expected to be higher (lower) than in later sub-periods if conventional S curve (J curve) effects predominate.

*Source: _wp11163 - References 28*

### Section IV explains the methodology adopted for estimation of TFP; Section V presents the

### _wp11163 - Section IV explains the methodology adopted for estimation of TFP; Section V presents the

### Analytical framework: J-curve and S-curve hypotheses
- Virmani (2005a, 2006a) proposed the hypothesis of the J-curve of productivity and growth following from major/large trade (import) liberalization.
- Gradual reforms (small/marginal steps) are expected to produce the conventional S curve of productivity and growth.
- Major liberalization (as in India in the 1990s) is argued to cause structural transformation with an initial slowdown in (measured) productivity and output growth followed by a surge — the J-curve pattern.
- Mechanisms causing the initial slowdown:
  - deterioration in capacity utilization from obsolescence of product lines and capital (still part of measured capital),
  - gradual adoption and spread of new technology,
  - diversion of human resources to learn new technology and markets,
  - rigidity in labor laws limiting immediate resource reallocation gains.
- Hybrid (S-J) outcomes are possible where some sub-sectors experience initial productivity gains (S-curve) before J-curve negative effects materialize due to timing/sequence of reforms.
- Neither S-curve nor J-curve hypotheses rule out declining productivity/growth at disaggregated industry levels.

### Pattern and timing of liberalization (implications for sectors)
- Early 1990s: liberalization of access to dis-embodied technology and capital (FDI, equity and External Commercial Borrowing); elimination of import controls/Quantitative Restrictions on Intermediate and Capital goods in early 1990s.
- QRs on manufactured consumer goods were not eliminated till the end of the 1990s-early 2000s.
- Nominal tariff reductions were broadly proportional but effective tariffs on consumer goods may have increased for much of the 1990s because remaining QRs kept effective protection high.
- Sectoral consequences:
  - Export-oriented industries (duty-free imports of intermediates, lower tariffs on capital goods even before 1990s reforms) were relatively immune from J-curve effects.
  - Highly protected industries had the largest technology gaps and stronger J-curve effects.
  - Public sector dominance (steel, aluminum, refineries) plus controls on capacity and licensing constrained private sector; de-licensing and privatization altered industry productivity dynamics.

### Data, sample selection, and variables
- Study period: 1980-81 to 2007-08; divided into:
  - Period I: 1981/82 to 1990/91
  - Period II: 1991/92 to 2007/08, further subdivided into:
    - Sub-period 1: 1991/92 to 1997/98
    - Sub-period 2: 1998/99 to 2001/02
    - Sub-period 3: 2002/03 to 2007/08
- Sample: 12 two-digit ASI manufacturing industries comprising 89.1% share in output:
  1. Food products & beverages (15.7%)
  2. Chemicals & chemical products (14.6%)
  3. Basic metals (13.1%)
  4. Textiles products (10.2%)
  5. Coke, petroleum products & nuclear fuel (9.6%)
  6. Machinery & equipment n.e.c (4.6%)
  7. Motor vehicles, trailers & semi-trailers (4.5%)
  8. Electrical Machinery and Apparatus nec (3.7%)
  9. Non-metallic Mineral products (3.7%)
  10. Rubber and Plastic products (3.2%)
  11. Other Transport equipments (2.7%)
  12. Fabricated metal products (2.7%)
- Remaining 10 two-digit industries (10.1% share) combined as ‘Others’; component shares listed (e.g., Radio, television & communication equipments (1.8%), Paper & paper products (1.8%), ... Wood & wood products (0.3%)).

### Construction of variables and data sources
- Deflation: series at current prices deflated with relevant price indices (1993/94=100); proxies from WPI where exact deflators unavailable.
- Capital stock: net fixed capital at constant prices constructed via Perpetual Inventory method with:
  - benchmark capital stock for 1973/74,
  - two price series for deflation: weighted price index for machinery & construction averaged backward 15 years from 1958/59 to 1973/74 for benchmark; implicit price deflator for remaining years,
  - annual gross investment = depreciation + change in net fixed capital,
  - depreciation fixed at 5 per cent per annum (life of capital stock 20 years).
- Labor input: total persons engaged (ASI); total emoluments deflated by price of labor (1993/94=100); price of labor = total emoluments / total persons engaged.
- Energy: fuel consumed (ASI); deflated using weighted price index (coal, petroleum, electricity as relevant).
- Materials: materials consumed; deflated by weighted price deflator.
- Services: not directly available; obtained by subtracting materials and energy expenses from total inputs (following Banga and Goldar (2004)). Services deflated using implicit price deflators constructed from service sector GDP series (NAS) and input-output matrix weights (CSO Input-Output matrix 1993/94).
- Data sources: Annual Survey of Industries (ASI) via EPW database for 1981/82–2003/04 and CSO for remaining years; WPI series from Ministry of Statistics and Program Implementation; Reserve Bank of India Bulletin (1976) for gross net ratio (GNR) to obtain gross capital stock for 1973/74; NAS for implicit price deflators.

### Structure of costs and trends (aggregate manufacturing)
- Average shares in total input costs (study period):
  - Materials: 62.9 per cent
  - Capital: 14.5 per cent
  - Services: 9.7 per cent
  - Energy: 6.4 per cent
  - Labor: 6.5 per cent
- Trends:
  - Capital share increased at rate of 0.9 per cent per annum (with sub-period fluctuations).
  - Services share rose from 8.0% in Period I to 10.7% in Period II at an average annual rate of 3.1 per cent.
  - Labor share declined at an average annual rate of -3.2 per cent.
  - Energy share declined at -0.9 per cent per annum.
  - Material share essentially constant (62.9 per cent; average annual growth -0.1).
- Table 1 period-specific shares (selected highlights):
  - Period I (1981/82-1990/91): Capital 13.0 (1.8), Labor 8.3 (-2.9), Energy 6.8 (0.4), Materials 63.9 (0.5), Services 8.0 (-3.6)
  - Period II (1991/92-2007/08): Capital 15.3 (-0.3), Labor 5.5 (-3.4), Energy 6.2 (-1.9), Materials 62.3 (0.1), Services 10.7 (3.1)
  - Study Period (1981/82-2007/08): Capital 14.5 (0.9), Labor 6.5 (-3.2), Energy 6.4 (-0.9), Materials 62.9 (-0.1), Services 9.7 (2.1)
- Sectoral variability: e.g., Materials share highest in Coke, petroleum products & nuclear fuel (82.6%) and Food products & beverages (75.1%); lowest in Non-metallic mineral products (37.4%). Capital highest in Non-metallic mineral products (21.7%) and Chemicals & chemical products (19.4%). Labor highest in Other transport equipments (12.8%), Textiles products (9.9%), Machinery & equipments n.e.c (9.8%).

### Trends in output and inputs (aggregate growth rates, constant prices 1993/94 = 100)
- Output growth:
  - Period I: 7.7 per cent per annum (1981/82 - 1990/91)
  - Period II: 8.2 per cent per annum (1991/92 - 2007/08)
  - Sub-period 1: 7.4
  - Sub-period 2: 2.7
  - Sub-period 3: 12.9
  - Study Period (1981/82 - 2007/08): 8.1 (note table lists Study Prd. as (8.1 1981/82 - 2007/08) and summary line as 7.7 elsewhere)
- Input growth (Study Period averages):
  - Materials: 8.2 per cent
  - Capital: 7.7 per cent
  - Services: 7.5 per cent
  - Energy: 5.1 per cent
  - Labor: 1.6 per cent
- Differences across periods:
  - Capital and energy growth declined in Period II; labor and services growth increased heavily in Period II.
- Table 4 specific entries:
  - Prd. I: Output 7.7; Capital 9.1; Labour 0.7; Energy 8.0; Materials 8.1; Services 2.7
  - Prd. II: Output 8.2; Capital 7.0; Labour 2.1; Energy 3.6; Materials 8.3; Services 10.1
  - Sub-prd.1: Output 7.4; Capital 8.7; Labour 2.3; Energy 3.8; Materials 6.3; Services 14.7
  - Sub-prd.2: Output 2.7; Capital 5.0; Labour -3.3; Energy -3.2; Materials 4.1; Services 0.0
  - Sub-prd.3: Output 12.9; Capital 6.3; Labour 5.4; Energy 8.0; Materials 13.4; Services 11.4
  - Study Prd.: Output (8.1 1981/82 - 2007/08) ; Capital 7.7; Labour 1.6; Energy 5.1; Materials 8.2; Services 7.5

### Methodology for TFPG (Translog index / growth accounting)
- Rationale: TFP growth isolates productivity gains (technological progress, scale, learning-by-doing, R&D, technical efficiency) whereas output growth from larger input application is less relevant to reform effectiveness.
- Partial factor productivity (PFP) calculated as output divided by quantity of a single input; useful but ignores interactions among inputs.
- Chosen approach: growth accounting using Translog index (discrete approximation of Translog production function) rather than econometric estimation, to avoid multicollinearity, autocorrelation, and large sample problems.
- Translog index properties:
  - Allows variability in elasticity of substitution between factors.
  - Does not assume Hicks-neutral technological progress.
- Translog specification used with five inputs: K (capital), L (labor), E (energy), M (materials), S (services); SK, SL, SE, SM, SS denote income shares summing to unity.
- General Translog discrete-change expression provided (symbolic form in source).

### Empirical findings — Partial factor productivity (aggregate and across sectors)
- Aggregate partial factor productivity growth (study period 1981/82 - 2007/08):
  - Capital (Y/K): 0.1
  - Labor (Y/L): 6.6
  - Energy (Y/E): 3.3
  - Materials (Y/M): 0.1
  - Services (Y/S): -1.7
- Period comparisons (aggregate, Table 5):
  - Period I (1981/82 - 1990/91): Capital -1.1; Labor 7.3; Energy -0.2; Materials -0.8; Services 4.9
  - Period II (1991/92 - 2007/08): Capital 2.0; Labor 7.3; Energy 5.5; Materials 0.3; Services -3.0
  - Sub-prd.1 (1991/92 - 1997/98): Capital -0.4; Labor 6.1; Energy 4.5; Materials 1.5; Services -4.7
  - Sub-prd.2 (1998/99 - 2001/02): Capital 0.5; Labor 6.7; Energy 6.3; Materials -2.7; Services 3.0
  - Sub-prd.3 (2002/03 - 2007/08): Capital 6.2; Labor 6.7; Energy 6.9; Materials -1.3; Services 2.5
- Pattern interpretations:
  - Labor and energy show S-curve patterns: Labor productivity steady then accelerates; energy moves from negative in 1980s to strong positive in 1990s.
  - Capital and materials show hybrid S-J patterns: acceleration, then deceleration, then strong acceleration in the third sub-period (capital particularly).
  - Services exhibit a J-curve pattern: productivity decelerated from 4.4% in the 1980s to -8.3% in first sub-period, then recovered but remained below 1980s levels until late 1990s; services became positive from late 1990s onward.
- Sectoral PFP highlights (Table 6):
  - Capital PFP study-period examples: Food products & beverages -1.6; Textiles -0.5; Coke, petroleum -3.7; Chemicals 0.5; Fabricated metal products 1.5; Others 0.0
  - Labor PFP study-period examples: Food products 5.2; Textiles 7.2; Coke, petroleum 4.7; Chemicals 5.7; Motor vehicles 8.6; Others 5.8
  - Energy PFP study-period examples: Food products 1.5; Textiles 3.1; Coke, petroleum 0.2; Chemicals 5.0; Non-metallic mineral products 4.1; Others 3.6
  - Materials PFP study-period examples: Food products 0.8; Textiles 0.1; Coke, petroleum -2.4; Chemicals -0.7; Motor vehicles 2.2; Others 0.1
  - Services PFP study-period examples: Food products -4.7; Textiles -0.2; Coke, petroleum -3.9; Chemicals -0.9; Motor vehicles -1.8; Others 2.4

### Empirical findings — Total Factor Productivity Growth (TFPG)
- Aggregate TFPG (Table 7, column 14; All Manufactures):
  - Period I (1981/82 - 1990/91): 0.61 per cent per annum
  - Sub-period 1 (1991/92 - 1997/98): 0.25
  - Sub-period 2 (1998/99 - 2001/02): -0.09
  - Sub-period 3 (2002/03 - 2007/08): 1.41
  - 1981/82 - 2007/08 (All/Total): 0.59
- Interpretation:
  - Clear J-curve pattern for aggregate manufacturing TFPG: modest growth in 1980s (0.6), slowdown to 0.25 in first sub-period, decline to -0.1 in second sub-period, sharp acceleration to 1.4 in third sub-period (more than twice the 1980s TFPG).
  - Causes attributed to combined effects of:
    - BOP crisis and 1991 shock,
    - dramatic import liberalization (removal of QRs on capital goods and intermediates; tariff reductions),
    - exchange rate reforms (fixed rate to managed float),
    - obsolescence of some capital goods and temporary falls in capacity utilization,
    - delayed diffusion of new technology and capacity adjustments,
    - labor market rigidities limiting immediate static allocation gains while dynamic gains (learning by doing, scale economies, R&D, spillovers from foreign trade and multinationals) materialize later.
- Sub-sector TFPG patterns (Table 7 summary):
  - Among 12 major sub-sectors:
    - 2 followed the S-curve pattern,
    - 5 followed the J-curve pattern,
    - 5 followed a hybrid S-J pattern.
  - Examples:
    - Textiles: J-curve (TFPG sequence 1.2; 0.49; 1.2; 1.8) — complete J-curve with trough in first sub-period.
    - Chemicals & chemical products: J-curve (1.7; -1.25; -0.3; 1.7).
    - Food products & beverages: J-curve but recovery slow (1.98; -0.56; -0.24; 0.95).
    - Motor vehicles & trailers: S=>J pattern with large later acceleration (0.57; 1.09; -2.27; 4.02).
    - Coke, petroleum & nuclear fuel: S-curve pattern with progressive improvement (-4.70; -3.99; -0.47; 2.39).
- “Others” minor industries (Table 8):
  - Mix of patterns: e.g., Wearing Apparel, Fur: J; Leather & related: S=>J; Medical, precision & optical instruments: S; Accounting, Office & computing machinery: declining trend (—).
  - Aggregate “Others” All/Total TFPG (1981/82-2007/08): 0.19

### Conventional factors modifying J-curve effects
- Public sector presence pre-1990 altered trajectories in several sub-sectors:
  - Electrical Machinery: TFPG rose from 1.9% in eighties to 3.1% in first sub-period, then trough -0.05%, then recovery to 2.8% in last sub-period — attributed to private sector expansion after lifting of investment controls and access to technology.
  - Basic Metals: similar pattern with slow recovery (0.34; 3.0; -1.0; -0.74).
  - Coke, Petroleum & Nuclear Fuel: progressive improvement from -4.7% in eighties to 2.4% in last sub-period; private investment and retained protection contributed.
- Delayed liberalization effects (consumer goods QRs removed late 1990s-early 2000s) produced hybrid S-J or weak-J outcomes in sub-sectors like Non-metallic mineral products and Rubber & Plastic products.
- Machinery & equipment sub-sector followed S-curve: sharp TFPG rise in first sub-period then gradual decline as effective protection declined and competition increased.

### J-curve in output growth
- Aggregate manufacturing output followed a hybrid S-J pattern:
  - Accelerated from 7.7 per cent per annum in Period I to 8.2 in Period II (first sub-period),
  - Decelerated to 2.7 in second sub-period,
  - Surged to 12.9 in third sub-period.
- Contribution of productivity to output growth:
  - Share of productivity in output growth moved from 10.7 per cent in first sub-period to -3.5 per cent in second sub-period to nearly 11 per cent in third sub-period.
- Capital contribution examples across sub-periods (from narrative): capital contributed 1.4 percentage points to output growth in first sub-period, fell to nearly 0.8 percentage points in second sub-period, and recovered thereafter.

*Source: Excerpts from the provided IMF working paper content unit.*

### 0.9 per cent during third sub period even as its share in total output growth fell, indicating

### _wp11163 - 0.9 per cent during third sub period even as its share in total output growth fell, indicating

### Sources and magnitude of aggregate manufacturing output growth
- Materials accounted for more than 60 per cent of the output growth over the study period.
- Services contributed 12.7 percent of output growth.
- Capital contributed 12.3 percent of output growth.
- Labor and Energy together contributed less than 6 per cent of the output growth.
- Improvement in capacity utilization is indicated by a contribution of 0.9 per cent during the third sub period even as its share in total output growth fell.
- Contributions from labor, energy, materials and services show a similar pattern of changing contributions across periods.

### Table 9: Sources of output growth in aggregate manufacturing (%) — key figures
- Study period: 1981/82 - 2007/08 — Growth in Output: 8.06
  - K: 1.12 (13.8)
  - L: 0.08 (1.1)
  - E: 0.33 (4.1)
  - M: 5.19 (64.4)
  - S: 0.75 (9.3)
  - TFPG: 0.59 (7.3)
- Prd I: 1981/82 - 1990/91 — Growth in Output: 7.74
  - K: 1.19 (15.3)
  - L: 0.05 (0.6)
  - E: 0.54 (6.9)
  - M: 5.21 (67.4)
  - S: 0.14 (1.8)
  - TFPG: 0.61 (7.9)
- Prd II: 1991/92 - 2007/08 — Growth in Output: 8.23
  - K: 1.08 (13.1)
  - L: 0.10 (1.3)
  - E: 0.22 (2.6)
  - M: 5.18 (63.0)
  - S: 1.07 (13.0)
  - TFPG: 0.58 (7.0)
- Sub-prd 1: 1991/92 - 1997/98 — Growth in Output: 7.41
  - K: 1.37 (18.5)
  - L: 0.15 (2.1)
  - E: 0.25 (3.3)
  - M: 3.98 (53.6)
  - S: 1.42 (19.1)
  - TFPG: 0.25 (3.4)
- Sub-prd 2: 1998/99 - 2001/02 — Growth in Output: 2.67
  - K: 0.77 (28.9)
  - L: -0.18 (-6.8)
  - E: -0.21 (-7.7)
  - M: 2.38 (88.8)
  - S: 0.01 (0.3)
  - TFPG: -0.09 (-3.5)
- Sub-prd 3: 2002/03 - 2007/08 — Growth in Output: 12.88
  - K: 0.94 (7.3)
  - L: 0.24 (1.9)
  - E: 0.46 (3.6)
  - M: 8.46 (65.7)
  - S: 1.36 (10.6)
  - TFPG: 1.41 (11.0)

(Note: Figures in bracket indicate the percentage contribution of the corresponding variable in total output growth.)

### Patterns of productivity and output growth across manufacturing sub-sectors
- Of twelve major sub-sectors considered:
  - 1 showed an S curve pattern.
  - 3 showed a J curve pattern.
  - 8 showed a hybrid S-J pattern.
- In seven sub-sectors the pattern of output growth matched TFPG growth; in five sub-sectors the pattern differed.
- In three sub-sectors (‘rubber and plastic products’, ‘non-metallic mineral products’, and ‘Other transport equipment’, fabricated metal p) changes in output growth and TFPG growth were inversely related across periods/sub-periods.
- Textile sub-sector: output growth accelerated in the first sub period while TFPG growth decelerated.
- Machinery and equipment sector: both output and TFPG accelerated in the first sub-period, but correlation was not maintained with the same sign in subsequent sub-periods.
- At the disaggregated sub-sector level TFPG trends were much more varied than at aggregate level.

### Interpretation: S-curve, J-curve, and the role of reforms
- Economic policy reforms (1980s and 1990s) are expected to accelerate growth through:
  - Short-term gains in Static Efficiency via factor re-allocation.
  - Dynamic efficiency gains.
- Expected sector-level transition: S-shape movement from lower steady state to higher steady state in growth and TFPG.
- Virmani’s “J curve” hypothesis: major/non-marginal liberalization can induce obsolescence of skills, capital and technology in some industries, causing:
  - Initial negative or non-positive effect on productivity and growth (the initial J portion).
  - A subsequent recovery as firms learn new skills and adopt new technologies (rising J and entry into S curve upper portion).
- Empirical findings in this study:
  - Manufacturing TFPG followed a J curve pattern as a consequence of the 1990s reforms.
  - Partial factor productivity of capital, materials and services followed a J-curve pattern.
  - Partial factor productivity of labor and energy increased progressively during the 1990s and 2000s (attributed to capital deepening and access to more energy efficient technologies and capital goods — S curve effects).
  - Manufacturing output growth followed the J-curve; materials input had the largest impact on the pattern of output growth, followed by services and capital.

### Heterogeneity, sequencing, and institutional constraints
- Variation across sub-sectors reflects differing exposure to reforms, timing/sequencing of reforms, and initial technology gaps.
- Sequencing example: removal of QRs/import controls and staged reduction of import duties/tariffs temporarily maintained or raised effective protection in some sectors, allowing initial TFPG acceleration before J-curve effects manifested.
- The J-curve effect is expected to be stronger where the technology gap with global best practice was larger, but:
  - Rigid labour laws and constrained labor mobility reduced incentives in firms subject to strict labor regulations (firms with 100 or more workers).
  - In small-scale, labor-intensive industries (subject to less rigid labor laws) the incentive from a large technology gap could be offset by the disincentive of becoming “stuck” with many workers.

### Aggregate classification of TFPG patterns (twenty two sub-sectors)
- 3 sub-sectors followed an S-curve pattern (14%).
- 8 sub-sectors followed a J curve pattern (36%).
- 10 sub-sectors followed a hybrid S-J pattern (45%).
- 1 sub-sector followed a pattern of falling productivity suggesting potential non-viability in an open economy (5%).
- Conclusion: the hypothesis that reforms and opening damaged Indian manufacturing is decisively rejected.

*Source: _wp11163 - 0.9 per cent during third sub period even as its share in total output growth fell, indicating*

### References

### _wp11163 - References

### Import Liberalization: Mechanism and Measurement Effects
- Import liberalization changes relative prices from distorted domestic ratios to World relative prices, causing highly protected goods to have relatively higher prices and to grow faster in a distorted economy.
- With liberalization, prices and growth rates of previously highly protected goods will fall, temporarily even to negative, while output and growth of previously under-protected goods will rise.
- Any Laspeyres type quantity indices will understate the (true change in) growth rate during reform periods; shifting the base year to a post-reform year will raise post-reform growth rates and reduce pre-reform ones, reflecting reform effects more accurately.
- If reforms are slow and gradual, producers adapt by stopping investment in unprofitable product lines and initiating it in newly profitable ones; the former is limited by the rate of depreciation and the latter by the pace of incremental technical change.
- If reforms are dramatic in a heavily protected economy, a major import liberalization will initially slow measured productivity growth and result in its acceleration only after a lag.

### Major Import Liberalization and the J-curve of Productivity
- Major import liberalization can cause drastic changes in product-specific competitive advantage; capital immobility makes capital in uncompetitive product lines redundant and capacity utilization falls.
- If productivity calculations use capital stock measures based on the accumulation method with no adjustment for capacity utilization, calculated productivity will decline in affected product lines.
- Full utilization of new potential requires adoption of new technology and investment in new capital goods; introduction of completely new (unfamiliar) technology will have short-term negative effects while positive productivity effects take time to emerge due to technology diffusion (S-curve).
- This dynamic produces a J-curve of liberalization-productivity: an initial decline in measured productivity followed by eventual acceleration.
- The J-curve is used to explain observed fall in total factor productivity growth in registered manufacturing during the 1990s and low acceleration of aggregate TFPG and GDP growth rates.

### Reasons and Evidence from Innovation and Technology Adoption Literature
- Incremental change differs from major (“drastic”) inventions which “can trigger an uneven growth trajectory, which starts with a prolonged slowdown followed by a fast acceleration” (Helpman (2004)).
- Adoption of new technologies requires firms to learn how to use them, slowing productivity growth (Hornstein and Krusell (1996); Greenwood and Yorokolgu (1997)).
- Development of complementary inputs takes time and resources, temporarily diverting resources and slowing growth (Helpman and Trajtenberg (1998)).
- On-the-job training raises worker productivity but technology-specific skills may be lost when a new technology replaces an old one, causing temporary declines in labor productivity (Helpman and Rangel (1999)).
- The experience of Indian import liberalization suggests major import liberalization can have similar short-term negative effects followed by longer-term gains.

### Dynamic Competition, FDI, and the Means to Compete
- Competition conceptualized in three aspects: freedom to compete; pressure to compete (competitive pressure); and means and ability to compete.
- India’s historical production, investment and external controls restricted freedom to compete for medium-large firms and limited exploitation of economies of scale.
- Competitive pressure can arise from domestic production (including FDI) or imports; the threat of imports can be as powerful as actual imports.
- Policy instruments: import bans or QRs eliminate both actual imports and the threat of imports; tariffs, even if high, may preserve some competitive pressure by maintaining the threat of imports.
- Broad liberalization of import controls coupled with a rise in average tariffs can paradoxically increase competitive pressure by reducing rents, evasion and corruption and lowering transaction costs of imports.
- The means to compete include access to inputs and capital goods, technology, capital, skills, and flexibility to adjust factors; FDI bundles technology, management, marketing, and capital, rapidly expanding access to these means and generating spillover effects that strengthen domestic entrepreneurs.
- Freedom to import inputs and capital goods expands both competitive pressure and the producer’s means to compete; freedom to import consumer goods increases competitive pressure and information flows that enhance long-run competitive ability.

### Broad and Deep Reforms: Pace, Composition, and Growth Effects
- The 1990s reforms were broad but many were static in nature and not expected to impact growth or TFPG over the short–medium term; such reforms may increase efficiency and/or equity with long and variable lags.
- The 1990s tax reforms (primarily income tax) were expected to lead to a sustainable increase in revenues rather than immediate TFPG gains.
- Imbalance in the 1990s: increase in pressure to compete exceeded increases in access to the means to compete; recommended factor market reforms (labor, risk capital and debt finance, bankruptcy law, technology) were not fully implemented.
- Sharp reduction in protection contributes to a J-curve of productivity and growth (initial negative impact on measured productivity followed by slowly rising factor productivity).
- Elements of the J-curve after major liberalization:
  - Rebalancing of historically distorted prices that change sectoral weights.
  - Immediate reduction in capacity utilization in unprofitable product lines due to capital immobility until depreciation reduces excess capacity.
  - Gestation lags in investment in newly profitable product lines and S-curve technology diffusion.
  - Resources and effort to adopt unfamiliar technology may reduce productivity of existing technology/capital during transition.

### Growth Trend Estimates and Predictions
- Based on the pre-1999-2000 NAS GDP data series Virmani (2005a) predicted: “the underlying trend growth is currently about 6.3% (6.25% to 6.35%). … It is likely to rise to about 6.5% over the next few years as the effect of the 1990s reforms works through the system as hypothesized in our analysis of the J-curve of liberalization.”
- Once the new 1999-2000 based NAS GDP series became available in 2006 the rising growth trend was found to be statistically significant for aggregate growth, and the forecast for underlying growth was revised upward.
- In Virmani (2009) the underlying growth rate of the Indian economy was estimated to be between 8.5% and 9%.

*Source: _wp11163 - References (Appendix: J Curve of Import Liberalization and Productivity)*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11163.pdf_