A Censored Maximum Likelihood Approach to Quantifying Manipulation in China’s Air Pollution Data

A-Tier
Journal: Journal of the Association of Environmental and Resource Economists
Year: 2020
Volume: 7
Issue: 5
Pages: 965 - 1003

Authors (3)

Dalia Ghanem (not in RePEc) Shu Shen (not in RePEc) Junjie Zhang (Duke Kunshan University)

Score contribution per author:

1.341 = (α=2.01 / 3 authors) × 2.0x A-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

Data manipulation around cutoff points is observed in economics broadly and in environmental and resource economics in particular. This paper develops a simple and tractable censored maximum likelihood approach to quantify the degree of manipulation in China’s air pollution data around the “blue-sky day” cutoff. We construct annual measures of manipulation for 111 Chinese cities. For Beijing, we estimate 4%–16.8% of manipulation among reported blue-sky days annually, which translate to an estimated total of 208.1 manipulated blue-sky days between 2001 and 2010. For the remaining cities reporting pollution data over the 10-year period, we estimate a 93.9 average for the total number of manipulated blue-sky days with a 395.9 maximum. Using LASSO shrinkage, we examine the relationship between manipulation and local official characteristics, and find a positive correlation between manipulation and having an elite-educated party secretary, robust to numerous checks. Further empirical analysis suggests that promotion considerations may help explain this finding.

Technical Details

RePEc Handle
repec:ucp:jaerec:doi:10.1086/709649
Journal Field
Environment
Author Count
3
Added to Database
2026-01-29