You're not wrong.

It is about where the CEO was conceived and born. From the article's introduction:

>Our approach addresses these issues by exploiting prenatal exposure to pollution from Superfund sites as an exogenous source of variation in executive risk-taking behavior. [...] we control for a wide range of fixed effects, including firm, year, industry-year, CEO birth year, birth county, and headquarters state, ensuring that comparisons are made among otherwise similar CEOs. [...] we show that selection mechanisms in promotion can amplify behavioral traits shaped by early-life conditions, even when firms are unaware of those traits.

Personally, I think it sounds like hogwash -- "statistically significant" findings that have little bearing on reality.

More likely, those born in heavily industrial areas are more likely to have careers in industry.