Contains the data and code used in the "Something in the Air" working paper produced as my Economics Undergraduate Honors Thesis at UC Berkeley in Spring 2022.

In the paper, I analyze the effect of country policy on air quality outcomes to determine which types of national policies are most effective in decreasing air pollution.

• I develop a theoretical framework for teasing out the different effects of policy on "source behavior" (e.g. the amount of driving in a country) and "generation intensity" (e.g. how much pollution is emitted per unit of driving).
• I proceed by matching datasets on pollution and relevant control variables from OECD to policy data scraped from ECOLEX and downloaded from UNFAO.
• I generate regression estimates for the average effect of each policy keyword (e.g. "subsidy", "compliance mechanism", "tax credit") across a sample of the 38 OECD countries across two sectors: power generation and vehicle transport. These sectors account for a substantial proportion of all air pollution and have high data availability.
• I use a cumulative dummy approach for each policy keyword in each country to generate an average effect size for each keyword.

I use python with the xpath library to scrape the html from the ECOLEX database, and do all of the data cleaning and analysis in R.