世界发展银行-September-2019-PovcalNet-Update-_-What_s-New_12页_315kb

Summary of Global Poverty Monitoring Technical Note 10: September 2019 PovcalNet Update

Core Content

The September 2019 PovcalNet update from the World Bank includes revised survey data and methodological adjustments that result in minor changes to global poverty estimates. The update focuses on four countries—Argentina, Egypt, Ethiopia, and Mauritius—and introduces new metadata on the comparability of poverty estimates over time within countries.

Main Points

• Global Poverty Estimates:

  • The global $1.90 headcount ratio for 2015 increased slightly from 9.94% to 9.98%.
  • The number of people living in poverty increased from 731.0 million to 734.5 million.
  • These changes are relatively small compared to previous updates.

• Data Revisions:

  • Argentina (2003–2017):
    • Survey weights were revised to use the 2001 Census instead of the 2010 Census due to irregularities found in the latter.
    • Non-response imputation methods were standardized to match the hot-deck methodology used in earlier years.
    • This ensures consistent sampling and weighting across all years.
  • Egypt (2010/11, 2012/13):
    • The World Bank updated the welfare aggregates using a methodology similar to the 2015 CAPMAS adjustments.
    • The 2015 survey included a hedonic model for housing services and a market reference price for food items purchased via smartcards.
    • These changes were not applied in earlier surveys, leading to differences between PovcalNet and official CAPMAS aggregates.
    • The updated series now includes households from Helwan and 6th of October governorates, which were previously omitted.
  • Ethiopia (2015):
    • Incorrect survey weights were corrected, leading to a more modest decline in poverty.
    • The headcount ratio for $1.90 poverty decreased from 33.5% to 30.8%, instead of the previously estimated 27.3%.
    • The Gini coefficient also increased more moderately, from 33.2 to 35.0, instead of 39.1.
  • Mauritius (2012):
    • The consumption aggregate was slightly revised by Statistics Mauritius.

• National Inequality Measures:

  • For China, India, and Indonesia, inequality estimates are based on a combined distribution.
  • A coding error was corrected in the merging of distributions, which slightly affected national inequality estimates but had no impact on poverty estimates.
  • The largest Gini difference was 0.17, while most changes were smaller than 0.1.

• Other Changes:

  • The country name Macedonia, FYR was updated to North Macedonia in line with World Bank naming conventions.

Key Information

• The coverage and population of the estimates remain unchanged from the March 2019 update.
• No new surveys were added or removed.
• The CPI and National Accounts Data vintage used in the estimation has not been revised.
• The comparability of poverty estimates is now documented in the metadata database, which assigns a comparability value to each survey point.
• A value of 0 indicates the oldest comparable series, while a 1 indicates a break in comparability.

Structure of Comparability Metadata

• Each survey point (country, year, welfare, data type) is assigned a comparability value.
• Within the same country, survey points with the same comparability value are considered comparable.
• When comparability is broken, the value changes to 1 for that year and continues until the next break.
• The most recent comparable poverty series for each country is associated with the highest comparability value.

Appendix 1 Highlights

• Table 1 in the Appendix shows the changes in key poverty and inequality estimates from March 2019 to September 2019.
• The changes are generally small, with the largest impact observed in Ethiopia due to the correction of survey weights.
• For Indonesia and India, the changes in the Gini index are too small to be displayed in the table.

Conclusion

The September 2019 update reflects methodological improvements and corrections in survey data for several countries, leading to small but meaningful adjustments in global poverty and inequality estimates. These revisions enhance the accuracy and consistency of the data, supporting more reliable poverty analysis and monitoring.