2013年-IMF国际货币组织全球_Income_Mobility_and_Welfare_28页_700kb

Summary of "Income Mobility and Welfare"

Core Content

This paper develops a quantitative framework to analyze individual income dynamics, mobility, and their implications for social welfare. The authors focus on intra-generational income mobility, which refers to the movement of individuals within their lifetimes across the income distribution. The study uses individual income data from Mexico to illustrate the framework and its empirical applications.

Main Points

• Income Dynamics and Components: Individual income is modeled as a stochastic process with two unobserved components:

  • A persistent component, modeled as an AR(1) process, reflecting long-term changes in income.
  • A transitory component, which includes measurement error and temporary shocks, assumed to be i.i.d. (independent and identically distributed).

• Income Mobility Definition: The authors define income mobility using the Hart Index, which measures the complement of the correlation between log incomes at time 0 and time $t$:
  $$
  m_t = 1 - \operatorname{corr}(\ln y_{i0}, \ln y_{it})
  $$
  This index captures the extent to which individuals' incomes change over time.

• Welfare Implications: The paper highlights that:

  • Measurement error and transitory shocks have minimal welfare effects as they do not influence long-term consumption.
  • Persistent income shocks (income risk) have significant negative effects on social welfare.
  • Catching-up of low-income individuals with high-income individuals (welfare-enhancing) also affects social welfare, though to a lesser extent than income risk.

• Key Findings from Mexico Data:

  • Most of the measured income mobility in Mexico is due to measurement error and transitory shocks, which are welfare-neutral.
  • A small portion of mobility is due to persistent income risk and catching-up, both of which have important welfare implications.
  • The auto-correlation coefficient ($\rho$) is a key determinant of mobility and welfare. A higher $\rho$ reduces mobility and welfare, as it implies greater persistence in income shocks.

Key Parameters and Their Implications

The framework estimates four key parameters:

• $\sigma_{\omega_0}^2$: Variance of initial income.
• $\sigma_{\eta}^2$: Variance of transitory shocks.
• $\sigma_{\epsilon}^2$: Variance of persistent shocks.
• $\rho$: Auto-correlation coefficient of the AR(1) process.

These parameters are used to derive analytical expressions for:

• Income mobility as a function of the persistence of shocks and their variances.
• Social welfare based on a consumption-saving model with labor income risk and incomplete markets.

Methodological Contributions

• The paper proposes a tractable analytical framework that bridges the gap between empirical income mobility analysis and welfare-theoretic evaluation.
• It introduces a two-step estimation approach:
  1. First, estimate a Mincer earnings regression to obtain residuals.
  2. Second, use these residuals to estimate the mobility parameters using equations (4') and (6'), which incorporate both the cross-sectional variance and covariance of residual income over time.
• The authors use a non-linear seemingly unrelated regressions (NLSUR) model to estimate the parameters, which allows for efficient estimation by combining information from both equations.

Empirical Application in Mexico

• The data used comes from the Encuesta Nacional de Empleo Urbano (ENEU), a rotating panel survey that has been conducted since 1987.
• The survey includes demographic and employment data for individuals aged 12 and above, and provides quarterly observations over 18 years (72 quarters).
• The results show that:
  • The autoregressive parameter $\rho = 0.977$, indicating high persistence in income shocks.
  • The variance of transitory shocks $\sigma_{\eta}^2 = 0.202$ is much larger than the variance of persistent shocks $\sigma_{\epsilon}^2 = 0.015$, suggesting that measurement error plays a significant role in observed income mobility.
  • The variance of initial incomes $\sigma_{\omega_0}^2 = 0.104$ is also estimated, and the results are robust across different sub-samples.

Conclusion

• Decomposing income mobility into its fundamental components (persistent vs. transitory shocks) is crucial for accurate welfare evaluation.
• The study underscores that most observed income mobility is welfare-neutral, while persistent income risk and catching-up have significant welfare implications.
• The framework is generalizable and applicable to other contexts, particularly where social insurance schemes are absent or incomplete.

Key Equations and Expressions

• Income process:
  $$
  \log y_{it} = \omega_{it} + \eta_{it} + \mu
  $$
• AR(1) process for the persistent component:
  $$
  \omega_{i,t+1} = \rho \omega_{it} + \epsilon_{i,t+1}
  $$
• Income mobility (Hart Index):
  $$
  m_t = 1 - \frac{\rho^t \sigma_{\omega_0}^2}{\sqrt{\sigma_{\omega_0}^2 + \sigma_{\eta}^2} \sqrt{\rho^{2t} \sigma_{\omega_0}^2 + \sigma_{\eta}^2 + \frac{1 - \rho^{2t}}{1 - \rho^2} \sigma_{\epsilon}^2}}
  $$
• Variance of residual income over age:
  $$
  \operatorname{Var}[v_{iz}] = \sigma_{\eta}^2 + \rho^{2z} \sigma_{\omega_0}^2 + \frac{1 - \rho^{2z}}{1 - \rho^2} \sigma_{\epsilon}^2
  $$
• Covariance of residual income over time:
  $$
  \operatorname{Cov}(v_{iz}, v_{i,z+1}) = \rho^{2z + 1} \sigma_{\omega_0}^2 + \frac{1 - \rho^{2z}}{1 - \rho^2} \rho \sigma_{\epsilon}^2
  $$

Policy Implications

• The findings suggest that public policy should focus on reducing persistent income risk and enhancing catching-up mechanisms to improve social welfare.
• The importance of decomposing income mobility into its components is emphasized for more accurate welfare assessment.