世界发展银行-September-2020-PovcalNet-Update---What_rsquo_s-New_41页_6mb

Global Poverty Monitoring Technical Note: September 2020 PovcalNet Update Summary

Core Content

The September 2020 PovcalNet update introduces several key changes to global and regional poverty estimates, primarily driven by the adoption of revised 2011 Purchasing Power Parities (PPPs), updates to national accounts and population data, and the inclusion of new survey data. These revisions aim to improve the accuracy and consistency of poverty estimates used in the forthcoming Poverty and Shared Prosperity Report 2020.

Main Changes

1. Annual Regional and Global Poverty Estimates

• For the first time, PovcalNet now provides annual poverty estimates for all regions, making it easier to track poverty trends over time.
• Population coverage rules have been revised to ensure better representativeness of data for global and regional aggregates.
  • Coverage now includes survey data from three years before and after the line-up year.
  • The regional population coverage threshold was increased from 40% to 50%.
  • The global poverty coverage threshold requires data from at least 50% of the population in low-income and lower-middle-income countries (LIC/LMIC), where most of the poor reside.

2. Revised 2011 PPPs

• The update adopts the revised 2011 PPPs from the International Comparison Program (ICP) released in May 2020.
• These PPPs are used to adjust poverty estimates for relative price differences across countries.
• The global extreme poverty rate increased by 0.24 percentage points (17.7 million more people) in 2017 using the revised PPPs.
• The international poverty line (IPL) remains at $1.90 per person per day (with an updated underlying estimate of $1.87).
• The $5.50 poverty line increased slightly by $0.15 due to the revised PPPs, while the $3.20 line remains unchanged.
• The World Bank decided to keep all global poverty lines unchanged, including the Societal Poverty Line (SPL), to avoid frequent revisions that could affect comparability.

3. Surveys Added

• Nigeria 2018/2019 survey data was added, significantly impacting poverty estimates for Sub-Saharan Africa.
  • The survey covered approximately 22,000 households and is representative at national, zonal, state, and rural/urban levels.
  • However, Borno state was excluded due to inaccessibility during the survey.
• Canada 2017 data was added as binned data (400 quantile groups) from the Luxembourg Income Study (LIS).

4. Revisions to Welfare Aggregates

• US data in LIS was updated, affecting disposable household income (DHI) estimates.
• India, Indonesia, and China had errors in their national Gini coefficient estimates, which were corrected using the revised 2011 PPPs and proper application of spatial price adjustments.
  • For India and Indonesia, the urban/rural spatial price adjustment was incorrectly applied in the March 2020 update.
  • For China, the Gini coefficient changes were minimal due to the unchanged urban-to-rural PPP ratio.

5. Changes to National Accounts Data

• National accounts data used to adjust survey data to reference years was updated using the May 2020 version of WDI.
• Supplementary data was obtained from the April 2020 WEO for countries where WDI data were missing (e.g., Angola, Djibouti, Iran, South Sudan, Syria, and Taiwan, China).
• The pass-through rate for India is 67%, used to adjust HFCE growth from national accounts to survey-based consumption data after 2015.

6. Changes to Population Data

• Population data was updated using the July 2020 version of WDI, including Eritrea which was previously excluded.
• This ensures more accurate representation of the population in poverty calculations.

Key Information

• Global poverty estimates for 2017: 689 million people live below the IPL of $1.90 per day, with a global poverty rate of 9.2%.
• Sub-Saharan Africa has the highest regional poverty rate at 41.0%, accounting for over 60% of the global poor population.
• South Asia has less than 50% survey coverage, so its poverty estimates are not displayed.
• The new coverage rules allow for more granular poverty estimates, especially in regions like Latin America and the Caribbean, where poverty estimates are now reported for 2018.
• The annual line-up series is noisier than previous estimates, especially before 1990, due to the shift from income to consumption data at the country level.
• The new PPPs have minimal impact on global poverty estimates compared to previous PPP revisions (e.g., 2005 to 1993).

Conclusion

The September 2020 PovcalNet update enhances the accuracy and timeliness of global and regional poverty estimates by incorporating revised PPPs, new survey data, and updated national accounts and population data. These changes provide a more comprehensive and up-to-date view of poverty trends, while maintaining the integrity of the poverty measurement framework.