The analysis is based on wave I and wave IV of Add-Health. We exploit the recently released variable indicating a polygenic score of educational attainment for 1,886 individuals. 
Educational attainment is measured in years of education and it is based on self-reported educational attainment at Wave IV, where respondents were 26-32 years old.

Neighbourhood characteristics

Neighborhood characteristics are based on contextual characteristics at Wave I, reported as census block group level. Census blocks are defined as “a small, usually compact area, bounded by streets and other prominent physical features as well as certain legal boundaries.” A block group is a cluster of census blocks within a census tract. It is the lowest level of geography for which the Census Bureau publishes sample data. For the 1990 Census, block groups averaged 452 housing units, or 1,100 people.5 A typical census tract contains 4 or 5 block groups.
The following information at the block group level (measured in Wave I) are used in the preliminary analysis:

Empirical strategy

Our empirical strategy investigates whether there is interaction between genetic predisposition for higher education and neighborhood characteristics. Using OLS regression models, we estimate the following equation:
\(\begin{equationarray} \text{EduYears}_{ij}=\alpha+\theta*X_{i}+\beta*\text{PGS}_{i}+\gamma*C_{j}+\delta*\text{PGS}_{i}*C_{j}+\varepsilon_{ij} \\ \text{College}_{ij}=\alpha+\theta*X_{i}+\beta*\text{PGS}_{i}+\gamma*C_{j}+\delta*\text{PGS}_{i}*C_{j}+\varepsilon_{ij} \\ \end{equationarray}\)