2015年-IMF国际货币组织全球_Fiscal_Policy_Implications_for_Labor_Market_Outcomes_in_Middle_24页_822kb

报告摘要

Summary of "Fiscal Policy Implications for Labor Market Outcomes in Middle-Income Countries"

Core Content

This working paper by Ara Stepanyan and Lamin Leigh examines the long-term effects of public employment on labor market outcomes in middle-income countries (UMICs). The study explores whether public employment can reduce unemployment and how it affects private employment.

Main Viewpoints

Public Employment and Unemployment: The paper finds no evidence that public employment reduces the unemployment rate in the medium to long term. Instead, it suggests that public employment may lead to job destruction in the private sector.
Substitutability and Rents: The impact of public employment on private employment depends on two key factors:
- Substitutability of public and private production: If public jobs are substitutable with private jobs, they negatively affect private employment.
- Size of public rents: Higher public sector rents increase the crowding-out effect on private employment.
Channels of Impact:
- Public employment increases wage pressure in the private sector, leading to reduced private employment.
- Public employment can also reduce the relative price of private goods, negatively affecting private wages and labor-force participation.
Policy Implications: The study suggests that reforms to reduce public rents and public employment may improve labor market performance. It also advises against using public employment as a long-term tool for reducing unemployment.

Key Information

Sample and Data

The study uses data from 24 upper-middle-income countries (UMICs) between 1995 and 2011.
Data sources include the International Labor Organization (ILO), LABORSTA, and country statistical offices.
A narrow definition of public employment is used, excluding employment by state-owned enterprises.
Time-dependent macroeconomic variables are averaged over three-year periods to remove cyclical fluctuations.

Stylized Facts

Public employment accounts for an average of 13% of total employment across the sample.
In 60% of countries, the share of public employment increased over time.
The volatility of unemployment is higher than that of public employment, with Albania, South Africa, and Mauritius showing the highest unemployment volatility.
Ukraine has the highest volatility in public employment, likely due to government downsizing in 2000.

Empirical Results

Public employment is statistically significant in explaining unemployment and private employment in most models.
In OLS models, public employment has a negative impact on private employment.
In GLS and fixed effect models, public employment is still significant, but its effect on unemployment is less clear.
Labor market institutions (e.g., cooperation in labor-employer relations, rigidity of employment, flexibility of wage determination) also play a role in explaining unemployment and private employment.
Endogeneity bias is a concern, as public employment, private employment, and unemployment are jointly determined.

Methodology

The paper uses three-stage least squares (TSLS) and simultaneous equation regressions to address endogeneity.
The theoretical model assumes that public employment affects private employment through wage pressure and relative price changes.
The urbanization rate, population density, and trade openness are used as proxies for the valuation of public goods in the policymaking process.

Conclusions

Public employment does not reduce unemployment in the long run.
It crowds out private employment, especially when public and private production are substitutable and public rents are high.
Reforms aimed at reducing public rents and employment could improve labor market outcomes.
Policymakers should avoid relying on public employment as a long-term strategy to reduce unemployment.

Key Tables and Figures

Table 1: Shows the share of public employment in total employment and its correlation with unemployment across 24 UMICs.
Table 2: Presents OLS and GLS estimates of unemployment based on labor market institutions and public employment.
Table 3: Provides OLS and GLS estimates of private employment based on labor market institutions and public employment.
Figure 1: Displays the average relationship between public employment and unemployment rate for the period 1995–2011, showing a negative but statistically insignificant correlation.