Using remotely sensed night-time light as a proxy for poverty in Africa
2008

Using Night-Time Lights to Measure Poverty in Africa

Sample size: 338 publication Evidence: moderate

Author Information

Author(s): Noor Abdisalan M, Alegana Victor A, Gething Peter W, Tatem Andrew J, Snow Robert W

Primary Institution: KEMRI – University of Oxford – Wellcome Trust Collaborative Programme

Hypothesis

Can remotely sensed night-time light data serve as a reliable proxy for poverty measurement in Africa?

Conclusion

Night-time light metrics provide a cost-effective and accurate alternative to traditional asset-based poverty indices in Africa.

Supporting Evidence

  • Night-time light metrics distinguished between the most poor and least poor quintiles with greater precision.
  • The mean brightness of night-time lights had a Pearson correlation of 0.64 with the asset-based wealth index.
  • Spearman correlation for night-time light metrics was 0.79, indicating strong agreement with asset-based indices.
  • Overall, 2.2% of the total area of the 37 countries was covered by night-time lights.
  • Chad and Somalia had the lowest mean brightness of night-time lights.
  • Egypt had the highest percentage of area covered by night-time lights at 12.18%.
  • Administrative 1 units were ranked into quintiles based on asset indices and night-time light metrics.
  • NTL data can track changes in poverty levels over large scales.

Takeaway

This study shows that the brightness of lights at night can help us understand how poor or rich different areas are in Africa, just like counting people's belongings.

Methodology

Principal component analysis was used to compute asset-based poverty indices from household survey data, and correlations with night-time light metrics were examined.

Potential Biases

Potential biases in household survey data collection methods could affect the asset-based indices.

Limitations

The study relies on the accuracy of night-time light data and may not capture poverty nuances at smaller geographic scales.

Participant Demographics

Data was collected from 338 Administrative 1 units across 37 African countries.

Statistical Information

P-Value

<0.01

Statistical Significance

p<0.01

Digital Object Identifier (DOI)

10.1186/1478-7954-6-5

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication