2013年-IMF国际货币组织全球_Structural_Transformation_and_the_Volatility_of_Aggregate_Output_in_OECD_Countries_29页_945kb

Summary of "Structural Transformation and the Volatility of Aggregate Output in OECD Countries"

Core Content

This working paper investigates the relationship between structural transformation—the shift of labor from agriculture and manufacturing to the service sector—and aggregate output volatility in OECD countries. The author argues that as labor moves from high-volatility sectors (agriculture and manufacturing) to low-volatility sectors (services), the overall volatility of aggregate output decreases. This conclusion is reached through a combination of empirical analysis and a calibrated dynamic general equilibrium model.

Main Findings

• Negative Relationship: There is a negative correlation between the employment share of the service sector and the volatility of aggregate output in OECD countries. This suggests that as the service sector becomes a larger part of the economy, output volatility decreases.
• Sectoral Volatility Differentials: Labor productivity is more volatile in agriculture and manufacturing than in services. This implies that as labor shifts towards services, the overall volatility of output is reduced.
• Quantitative Contribution: Structural transformation accounts for a significant portion of the decline in output volatility. Specifically, shifts in the service sector explain about 30% of the volatility decline in the U.S. since the 1950s, and nearly half of the decline after the mid-1980s.
• Counterfactual Analysis: By constraining labor mobility across sectors, the author finds that the volatility of aggregate output decreases when labor is reallocated towards the service sector. This supports the idea that structural transformation plays a key role in reducing output volatility.
• Role of Financial Development: The level of financial development is controlled for in the analysis, and the results remain significant even after accounting for this factor.

Key Methodological Points

• Indivisible Labor Model: The paper uses a model where labor is indivisible, meaning individuals either work full-time in one sector or not at all. This is different from models that assume divisible labor.
• Calibration to the U.S.: The model is first calibrated to the U.S. economy to estimate preference parameters and then applied to other OECD countries.
• Decomposition of Volatility: The author decomposes output volatility into permanent (structural transformation) and transitory (exogenous shocks) components. By holding the permanent component constant, the impact of structural transformation on volatility can be isolated.
• Robustness Checks: The results are robust across different measures of volatility, including varying the rolling window and using de-trended series.

Empirical Evidence

• Volatility Trends: The volatility of aggregate output is lower in countries with a larger share of employment in the service sector. This is observed across both panel data and cross-sectional data.
• Correlation Analysis: The correlation between service sector employment share and output volatility is -0.68 across countries and -0.56 in the panel, indicating a strong inverse relationship.
• Development Proxy: The correlation between output volatility and GDP per capita is -0.54 across countries and -0.45 in the panel, but it is weaker than the correlation with the service sector employment share, suggesting that structural transformation explains more of the volatility reduction than development alone.

Model Structure

A. Economic Environment

• The economy consists of three sectors: agriculture, manufacturing, and services.
• Each sector uses linear technology with labor as the only input: $ Y_i = A_i L_i $.
• $ A_i $ is the sector-specific labor productivity, which is exogenous and captures factors like capital and local institutions.

B. Preferences

• Households have preferences over consumption of all three goods.
• The model incorporates non-homothetic preferences and low elasticity of substitution between goods, which are key drivers of labor reallocation.
• The utility function is specified as:
  $$
  U(c_{a,t}, c_t, h_t) = \omega \log(c_{a,t} - \bar{a}) + (1 - \omega) \log(c_t) + \phi \log(1 - h_t)
  $$
  where $ \bar{a} $ is the subsistence level of agricultural goods.

C. Analytical Solution

• The model is solved in two stages:
  1. For a given labor share in agriculture, the planner determines the optimal labor shares in manufacturing and services.
  2. The planner then optimizes the labor share in agriculture based on the results from the first stage.
• The optimal reallocation path is given by:
  $$
  \frac{\pi_s + \bar{s}/A_s}{\pi_m} = \left(\frac{1 - \theta}{\theta}\right)^{\varepsilon} \left(\frac{\log(1 - \bar{h}_m)}{\log(1 - \bar{h}_s)}\right)^{\varepsilon} \left(\frac{A_s}{A_m}\right)^{\varepsilon - 1}
  $$
  where $ \varepsilon < 1 $ indicates low substitution elasticity.

D. Calibration

• The model is calibrated using U.S. data to estimate key parameters.
• The calibration allows the model to replicate the sectoral labor reallocation observed in the U.S. and other OECD countries.

Conclusion

• The secular shift of labor towards the service sector in OECD countries during 1970–2006 was volatility-reducing.
• Structural transformation is a continuous process that contributes to the stabilization of output.
• The paper contributes to the broader literature on business cycle patterns across countries and highlights the importance of sectoral labor allocation in shaping macroeconomic volatility.

Key Contributions

• Introduces a dynamic general equilibrium model with indivisible labor to analyze the volatility-reducing effect of structural transformation.
• Extends previous U.S.-centric studies to a panel of OECD countries, offering a more comprehensive view of the phenomenon.
• Demonstrates that structural transformation is a key determinant of output volatility, even after controlling for financial development and GDP per capita.