2006年-世界发展银行全球_Development___A_Study_of_the_Indian_Manufacturing_Industry_18页_145kb

Summary of "Infrastructure, Externalities, and Economic Development: A Study of the Indian Manufacturing Industry"

Core Content

This article investigates the relationship between infrastructure investment and productivity growth in the Indian manufacturing industry, focusing on the presence of spillover externalities—effects that benefit firms beyond the direct use of infrastructure. The study uses data from 1972 to 1992 and applies a production function model to estimate the impact of infrastructure on total factor productivity (TFP) and total productivity (TP).

Main Views and Key Findings

• Infrastructure as a Factor of Production: Infrastructure is integrated directly into the production function as an unpaid factor, alongside labor, capital, and intermediate inputs.

• Two Channels of Impact: Infrastructure affects manufacturing productivity through:

  1. Market-mediated effects: Lowering the cost of intermediate inputs such as transportation and electricity.
  2. Non-market-mediated effects: Improving the overall efficiency of production, captured by the Hicksian efficiency term $A(B,t)$.

• Spillover Effects in India: The study finds that infrastructure spillovers accounted for nearly half of the growth in the Solow productivity residual for India's registered manufacturing industry, suggesting a significant indirect effect on productivity.

• Comparison with U.S. Manufacturing: Unlike the U.S., where Hulten and Schwab found no evidence of infrastructure externalities, the Indian case shows that infrastructure improvements had a substantial role in productivity growth.

• Empirical Methodology: The Solow residual method is used to estimate TFP, and the study applies a translog index procedure to calculate the relative productivity levels across states. This allows for isolating the infrastructure spillover effect ($\gamma$) from other factors like time ($\lambda$).

• Regional Productivity Dynamics: There was significant variation in productivity growth across Indian states. Some states, like Himachal Pradesh, had lower initial productivity levels but experienced faster growth, indicating convergence in productivity over the sample period.

• Data and Methodological Considerations:

  • Data is sourced from the Annual Survey of Industries (ASI), covering registered manufacturing firms.
  • The study uses real gross output and real value added as measures of productivity.
  • The Hall corrections are applied to account for increasing returns and deviations from marginal cost pricing.

• Infrastructure Growth Trends:

  • Road capacity (national and state highways) and electricity-generating capacity grew significantly.
  • There was a high correlation between infrastructure variables and time, suggesting the need to address multicollinearity in regression analysis.

Key Information

• Infrastructure Types Studied: Paved roads (national highways, state highways, district roads) and electricity-generating capacity.
• Time Period: 1972–1992.
• States Analyzed: 16 Indian states.
• Methodology:
  • Uses a production function model with the form $Q_{i,t} = A_{i,0} e^{\lambda_i t} B_{i,t}^{\gamma_i} F(K_{i,t}, L_{i,t}, M(B_{i,t}))$.
  • Estimates the Solow residual to measure total factor productivity growth.
  • Applies Hall corrections to account for non-competitive pricing and increasing returns.
• Key Results:
  • Total productivity in India grew at an average rate of 0.5% annually.
  • Infrastructure spillovers contributed significantly to this growth.
  • TFP growth was more conventional at around 2.2% annually.
  • Convergence in productivity levels was observed among states, especially those with initially lower productivity.
  • Multicollinearity was a concern due to the strong correlation between infrastructure variables and time.

Table Highlights

• Table 1 shows that materials were the primary driver of output growth, with capital also contributing significantly.
• Table 2 highlights that TFP growth was more substantial (around 2.2%) than total productivity growth (0.5%).
• Table 3 provides regional productivity growth estimates, showing that Maharashtra had the highest initial productivity, while Himachal Pradesh had the lowest. The bottom five states showed the fastest growth in both output and productivity, indicating catch-up dynamics.

Conclusion

The study concludes that infrastructure investment in India generated substantial spillover externalities, significantly contributing to productivity growth in the manufacturing sector. This contrasts with findings in the U.S., where such effects were not observed. The results support the idea that infrastructure is a key driver of economic development, especially in low-income economies, and that regional differences in infrastructure can have a measurable impact on productivity across states.