Haotian Xue 2025.10

The data used in this study are primarily drawn from three sources: the China Labor-force Dynamic Survey (CLDS), the Global Summary of the Day (GSOD) meteorological database from the U.S. National Centers for Environmental Information (NCEI), and the China City Statistical Yearbook. An integrated panel dataset was constructed, combining individual-level microdata on income, city-level meteorological conditions, and socioeconomic characteristics.
For meteorological variables, daily observations from GSOD stations were spatially interpolated using the inverse distance weighting (IDW) method to generate gridded data at a 0.1° × 0.1° resolution. These data were then aggregated to derive city-level daily mean, maximum, and minimum temperatures. Annual averages of these temperature measures were computed and used as core explanatory variables. Annual precipitation at the city level was also controlled for.
In terms of economic and individual data, the study used the annual total income of individuals from the CLDS as the dependent variable. A set of control variables was incorporated, including individual characteristics (age, age squared, and educational attainment) and city-level socioeconomic indicators (government expenditure as a share of total output, economic agglomeration, secondary industry share, and GDP growth rate). The data processing procedure involved standardizing coding rules, removing outliers and invalid samples (e.g., individuals outside the working-age range or with extreme income values), applying price deflation adjustments, and taking the natural logarithm of income variables.
The final sample consists of 29,629 valid observations. Descriptive statistics show that the mean of the logarithm of annual individual income is 9.773, the average annual temperature is 16.425°C, and the average age is 44.117 years. The average educational attainment corresponds approximately to a high school level. Other city-level economic variables also exhibit reasonable distributions. This dataset effectively integrates micro-level individual heterogeneity with macro-level climate and economic information, providing a solid empirical foundation for identifying the effect of temperature on labor income and exploring its underlying mechanisms.