US Decennial Census return rates: the role of social capital

Purpose - The purpose of this paper is to assess whether additional information about a community’s level of social capital can help to better predict a return rate from that area, in order to better target resources to improve mail-in responses. Design/methodology/approach - Two-sample two-stage least squares is used to apply determinants of six different measures of social capital from the Social Capital Community Benchmark Survey to observations in the Decennial Census (DC) which are then aggregated to the census tract level. The probability of a census tract having a high level of each social capital measure is estimated. Multivariate regression is used to identify the importance of high community social capital for predicting census mail-in return rates. Findings - The analysis reveals that a higher level of trust contributes the most to increasing return rates and a high level of political activism decreases return rates. Additionally, higher levels of sociability contribute negatively to DC return rates, which is consistent with sociability being linked to a more insular (i.e. family and friends) focus. Practical implications - While contributing statistically significantly to the predictability of census tract response rates, the cost of acquiring measures of social capital for each census tract may not to be viewed worth the gain in predictive power. Social implications - Higher levels of trust contribute positively to survey participation, suggesting that any social, economic or political environment that diminishes trust will undercut civic engagement. Political activism and (insular) sociability decrease participation. Originality/value - This paper combines non-public and public data to obtain measures of social capital along more dimensions than are typically studied, and finds that not all types of social capital are related to feelings of social integration in the same way.