April 7, 2010 | The Atlantic | Original Article

Personality and the Census

The Census Bureau's participation map made quite a splash last week. More than half of Americans (56 percent) had completed their forms by April 5, but there was tremendous variation across the 50 states. Wisconsin topped the list with 69 percent of Wisconsinites sending in their forms. Midwestern states did well across the board with more than two-thirds of Iowans and Minnesotans completing theirs. At the opposite end of the spectrum, less than half the residents of Alaska, Texas, New Mexico, and Louisiana completed theirs; 50 percent of New Yorkers had filled out their forms and 51 percent of D.C. residents.

But what factors might drive census participation? With the help of my colleague Charlotta Mellander, I decided to take a quick look. We examined the correlations between census participation and economic factors, demographic characteristics, and state personality traits. This analysis is based on simple correlations which identify associations between variables but do not specify causality. I also spoke to my colleague and collaborator, the Cambridge University personality psychologist Jason Rentfrow about these results.

The short answer is that in terms of who fills out census forms, personality is the only thing that seems to matter. We found no correlation between census participation and key economic and demographic variables like income, economic output, education level, or type of jobs. But we found some considerable correlation between states where more people filled out their census forms and three of what psychologists term the Big Five personality types -- especially in states with high concentrations of agreeable people, extroverts, and open-to-experience types.

States with large concentrations of agreeable people were most likely to complete their census forms (with a correlation of .43). Rentfrow says: "Agreeableness reflects a tendency to avoid conflict and to cooperate, so it makes sense that regions with large proportions of agreeable people would do what was asked of them and return their census forms on time."

States with large concentrations of extroverts were also more likely to complete their forms (with a correlation of .37). "One aspect of extroversion is the tendency to be expressive and reveal information about oneself," notes Rentfrow. "So perhaps people in extroverted regions are more comfortable revealing personal information about themselves as compared to places where people are introverted and private."

Conscientiousness comprises traits like responsibility, discipline, and obedience, so one would think that states with high concentrations of these more diligent types would be more likely to fill out their forms. But that's not what we found. The correlation between conscientious personality types and census participation was positive (in the range of .2)  but not statistically significant.

States with high concentrations of open-to-experience people were less likely to have their forms filled out. The correlation for all 50 states is negative but not significant in statistical terms. When we remove Alaska, which is an outlier, from the analysis, the correlation is stronger and statistically significant (-.34). Openness is a disposition toward creativity and originality. So in highly open regions, substantial numbers of people may be so focused or preoccupied with generating their next new idea -- writing their latest essay, making that short film, forming a new band, or just chasing after the latest, greatest experience -- that they put off their census forms. Or, Rentfrow adds, "People high in openness tend to be unconventional and to challenge the status quo, so one possible explanation for this finding is that more people in open regions may question the legitimacy or utility of the census."

That leaves the neurotics. Census participation is negatively associated with neurotics, but the relationship is not statistically significant.

Ultimately, when it comes to completing the census form, our correlations show that it's not economics, education, or types of employment of a state that seem to matter. Personality ends up being a bigger factor, at least according to our provisional analysis. Rentfrow provides some useful context: "Completing the census form is not just about reporting one's age, sex, race, income, education, or occupation. It asks people to reveal something about who they are, and not everyone is comfortable or compelled to do so. Research in psychology provides some clues about which people would be likely to feel comfortable or obliged to complete their forms, and the state personality data provides a map of where those people live."