2004年-世界发展银行全球_International_Economic_Activities____________and_the_Demand_for_Skilled_labor__Evidence_from_Brazil_and_China_43页_381kb

Summary of "International Economic Activities and the Demand for Skilled Labor: Evidence from Brazil and China"

Core Content

This paper examines the impact of international economic activities—such as the use of imported inputs, exports, and foreign direct investment (FDI)—on the demand for skilled labor in Brazilian and Chinese manufacturing plants. The study is based on plant-level data collected by the World Bank and aims to determine whether these activities increase or decrease the relative demand for skilled workers, particularly in developing countries with a relative abundance of unskilled labor.

The paper investigates two competing hypotheses:

• Trade Hypothesis: Increased international integration leads to specialization in unskilled-labor-intensive sectors, thereby reducing the demand for skilled labor.
• Skill-Biased Technological Change (SBTC) Hypothesis: International activities facilitate the diffusion of skill-biased technologies, increasing the demand for skilled labor.

It is argued that in industrialized countries, these two factors reinforce each other, but in developing countries, they may operate in opposite directions. The paper seeks to determine which effect is dominant in Brazil and China, two developing countries that have experienced significant international integration.

Main Findings

• In Brazil, increased international integration (imported inputs, exports, and FDI) is associated with a greater demand for skilled labor.
• In China, the opposite is observed: increased international integration is associated with a reduction in the demand for skilled labor.
• This contrast suggests that the effect of international activities on skilled labor demand depends on the country and may be influenced by the absorption of skill-biased technologies.

Key Variables and Measures

Skilled Labor Measures

• Nonproduction workers: Sum of management, professionals, and other nonproduction workers.
• Engineering, technical, and managerial occupations: Sum of management and professionals.
• College-educated workers: Workers with at least some college education.

Absorptive Capacity Measures

• R&D expenditures: Dummy variable indicating whether a plant engages in R&D.
• New product line introduction: Dummy variable indicating whether a plant introduced a new product line.

International Economic Activities

• Exporting: Dummy variable for plants that export, and share of sales exported.
• Imported inputs: Dummy variable for plants that use imported inputs, and share of inputs imported.
• FDI: Dummy variable for foreign-owned plants, and foreign ownership share.

Methodology

• The paper uses a translog cost function to estimate the relative demand for skilled labor.
• It distinguishes between wage bill shares and employment shares of skilled workers.
• To address endogeneity and unobserved heterogeneity, the authors:
  • Include managerial education as a proxy for manager quality.
  • Control for R&D activities and new product line introduction as indicators of absorptive capacity.
  • Use instrumental variables and weighted regressions to ensure robustness.

Empirical Results

• In Brazil, the use of imported inputs, exports, and FDI is positively associated with increased demand for skilled labor.
• In China, the use of imported inputs is associated with decreased demand for skilled labor, while the effects of exports and FDI are less clear.
• The magnitude of the effect varies depending on the definition of skilled labor:
  • For college-educated workers, the effect of imported inputs is the largest in Brazil (36% increase in employment share).
  • In China, the negative effect is most pronounced for nonproduction workers (15% decrease in employment share) and engineering, technical, and managerial workers (8% decrease).
• The effect on wage bill shares is less pronounced in Brazil than in employment shares, suggesting that relative wages of skilled workers may increase in plants using imported inputs.
• In China, wage bill shares are not significantly affected by imported inputs, but employment shares are significantly reduced.

Robustness Checks

• The results are robust to various estimation techniques, including instrumental variables and different specifications of international activities.
• The authors account for differences in industries and regions by including dummy variables.
• They also consider different measures of skilled labor and find that the overall pattern remains consistent.

Conclusion

The study highlights the complex relationship between international economic activities and the demand for skilled labor in developing countries. While the trade hypothesis predicts a reduction in skilled labor demand, the SBTC hypothesis suggests an increase. The findings show that in Brazil, international activities tend to increase the demand for skilled labor, whereas in China, they tend to decrease it. This implies that the direction of the effect is not uniform across developing countries and that skill-biased technology diffusion may play a significant role in shaping labor demand outcomes. The paper underscores the importance of absorptive capacity and managerial quality in determining the impact of international integration on skilled labor demand.