Neighborhood and Individual Factors Associated with Survey Response Behavior: A Multilevel Multinomial Regression Analysis of a Nationwide Survey in Japan

This study attempts to identify neighborhood and individual characteristics associated with survey responses in Japan. While the response rates of social surveys in Japan have declined in the past decades, factors contributing to non-responses have not been sufficiently examined. Therefore, this study is intended to empirically investigate the factors that differentiate individuals’ responses by analyzing a nationwide survey, ‘A Trend Survey on Japanese National Character 2012 (TS-JNC 2012)’. For this purpose, multinomial multilevel techniques are employed to assess if neighborhood and individual characteristics are related to non-response behaviors (i.e. non-contact and two types of refusal). The analyses reveal that both neighborhood and individual variables relate to how individuals respond to the national survey, indicating that non-responses are not random among individuals and neighborhoods; a higher degree of urbanization indicated by four neighborhood variables relates to non-responses. More specifically, the findings reveal that individuals in neighborhoods with higher population densities and crime rates are more likely to be ‘non-contact’ and ‘refusal’, even when other neighborhood and individual characteristics are controlled. In addition, cross-level interaction effects indicate that females and individuals who live in large residencies are more likely to refuse participation when in neighborhoods with relatively higher crime rates.