Are National Family Surveys Biased toward the Happy Family? A Multiactor Analysis of Selective Survey Nonresponse

Virtually, all large-scale family surveys in the United States and Europe have yielded a positive view of family ties in contemporary societies. The present study examines whether surveys like these are affected by selective nonresponse. Are people with negative family ties less likely to participate in surveys, and if so, to what extent does this yield a biased descriptive view of family solidarity? Using a novel multiactor design with matched register data, we examine the determinants of nonresponse of the parents of adult children aged 25-45 in the Netherlands. Our analysis reveals significant effects of the strength of parent-child ties on parental nonresponse, especially for fathers. Moreover, we find negative effects of divorce on father's participation and this effect is stronger when family ties are weak. While these findings support the hypothesis of selective nonresponse, the magnitude of the effects is small and descriptive findings on family ties change only modestly when correcting for selective nonresponse.