联合国西亚经济社会委员会-被占领巴勒斯坦领土的多层面贫困（英）-2021-36页_900kb

Summary of "Nowcasting Multidimensional Poverty in the Occupied Palestinian Territory"

Core Content

This report presents a methodological framework for nowcasting and forecasting multidimensional poverty in the occupied Palestinian territory (oPt) using a dynamic model based on the 22 indicators grouped into seven dimensions of poverty. The model is designed to capture changes in the multidimensional poverty index (MPI) and headcount (MPH) by analyzing the evolution of individual indicators and their interdependencies.

Main Views and Key Information

1. Multidimensional Poverty Index (MPI) and Headcount (MPH)

• The Palestinian MPI is constructed based on 22 indicators across seven dimensions: Education, Health, Employment, Housing conditions and access to services, Safety and use of assets, Interpersonal and state violence, and Personal freedom.
• Each indicator has a deprivation cutoff and a weight that contributes to the overall index.
• The MPI is calculated as the weighted average of households that are multidimensionally poor, while the MPH is the proportion of such households.
• The poverty cutoff is fixed at 0.20, and the index is sensitive to this choice, which is tested in the empirical section.

2. Theoretical Model

• The model treats the multidimensional poverty index as a function of the multidimensional cumulative distribution function $ H_t(\mathbf{x}) $, which can be expressed using Sklar's Theorem as a 22-dimensional copula $ C_t(\cdot) $ and univariate marginal distributions $ F_{X_{k},t}(x_k) $.
• The counterfactual multidimensional poverty index is reconstructed using the estimated copula and simulated changes in the marginal distributions of the indicators.
• The modeling approach allows for the decomposition of changes in MPI and MPH into changes in marginal distributions and copula structure.

3. Data and Methodology

• The baseline model uses data from the Palestinian Consumption and Expenditure Survey (PECS) for 2011 and 2016.
• Twelve indicators are available for both years, while ten indicators are only available for 2016.
• The dynamic of the deprivation incidence for each indicator is modeled using a two-parameter logistic function:
  $$
  F_{X_k}^M (z_k) = \frac{1}{1 + e^{-\alpha_k + \beta_ky_t}}
  $$
  where $ y_t $ represents an index (e.g., time, production, or political indices) that is linked to each indicator.
• The political and security indices (demolition index, occupation severity index, and hygiene and cleaning supply index) are used to capture the impact of conflict and occupation on multidimensional poverty.

4. Regional Disaggregation

• The model is applied separately to Gaza and the West Bank due to their distinct experiences with occupation.
• The demolition index is used for the West Bank, while the hygiene and cleaning supply index is used for Gaza to reflect the unique impact of the Israeli blockade.
• The occupation severity index is a weighted geometric average of the demolition index and the injury index.

5. Impact of the COVID-19 Pandemic

• The pandemic is modeled as a shock to the economic and social indicators, particularly affecting aggregate production and health outcomes.
• The World Bank's economic update (October 2020) is used to estimate the GDP contraction of 8% in 2020 and a 2.5% recovery in 2021.
• The baseline simulation incorporates these assumptions to reflect the pandemic's impact on the multidimensional poverty index.
• The reduction in injuries in Gaza during 2020 and 2021 significantly influenced the MPI trend, leading to a decrease in poverty in that region despite an increase in the West Bank.

6. Simulation and Forecasting

• The model simulates the MPI and MPH for all years between 2011 and 2021.
• The simulated values are compared with estimated values from the 2016 survey to validate the model.
• The results show that multidimensional poverty can move in opposite directions in the West Bank and Gaza due to different dynamics in economic, political, and security conditions.

Conclusion

• The proposed nowcasting framework is a valuable tool for policy-making in the absence of regular surveys.
• It allows for flexible modeling of indicator changes and counterfactual projections.
• The impact of the pandemic and occupation is clearly captured, highlighting the importance of non-monetary factors in assessing poverty in the oPt.
• The regional disaggregation provides a more nuanced understanding of poverty trends in Gaza and the West Bank, which are influenced by distinct socio-political contexts.