Are there civic returns to education?
Dee asks if increased educational attainment result in higher amounts of adult civic engagement?
Like Currie and Moretti (2003), Dee is interested in measuring the positive externalities of education. His implication is that if civic involvement can be shown to be a causal effect of education, policy makers would need to weigh that in our decisions about whether and how much money to spend on education.
Dee suggests that this causal effect may happen in two ways:
- Education may reduce the effective costs of civic participation by making it easier to make sense of complex political information.
- Education may shape individual preferences for civic activity by instilling democratic values or altering the peer groups and shared social norms.
Dee presents two analyses. The first includes data from the High School and Beyond dataset which includes longitudinal data from high school sophomores in 1980 with followups into the 1990s. The second data sets from the 1972-2000 General Social Surveys.
He looks at outcomes form HS&B that include answers to questions on whether respondents were: (1) registered to vote, (2) voted in last 12 months, (3) voted in 1988 Presidential election, or (3) volunteered in last 12 months.
He looks at GSS data on (1) voting in the presidential election, (2) newspaper readership, (3) group membership, and (4) and tolerance of the speech of people with unpopular viewpoints.
Dee's identification strategy includes instrumental variables on geographic proximity to 2-year colleges that exploits cross-sectional variation in the availability of these schools colleges which he argues represents an exogenous force that drive college attendance among students who are otherwise equal in expectation in the first study. In the second study, he instrument son variation in child labor laws which influences amount of schooling at the secondary level and which he argues is exogenous to attitudes on civic participation.
In both cases, he finds that educational attainment has large and statistically significant effects on subsequent voter participation and support for free speech. In particular, he finds in both studies that about ~2.5 more years of schooling increased voter turnout by 16 to 17 percentage points about 7% per year of schooling) and that education is associated with a increases in essentially all of his measures.
Important threats to validity that Dee addresses include: (1) the unobserved traits of communities near colleges (e.g., high socioeconomic status) could simultaneously encourage both higher educational attainment and increased civic participation and (b) the availability of colleges may promote a more politically aware and engaged culture that supports higher amounts of civic engagement of teens independently of its effects on educational attainment.