Social Interactions and Breast Cancer Prevention Among Women over 50 in the U.S

This paper examines whether a woman’s reference group has an influence on her preventive behavior against breast cancer. I focus on women over 50 and their decision to undergo mammography. I analyze two age groups: women 50-74 years old, to whom mammography recommendations directly apply, and women over age 75, for whom screening is optional. I define reference group as people living in the same geographical area at the MSA, county, and state levels. The size of the social multiplier is calculated by comparing the effects of a variable at the individual level and at the group level. The model exploits a variant of methods proposed by Glaeser and Scheinkman (2001), fixed effects, and a split sample instrumental variable to address the endogeneity of group level mammography rates. My findings support the existence of multipliers in age, education and among black women. The age multiplier suggests that a higher proportion of older women in the Metropolitan Statistical Area (MSA) – a phenomenon referred to as the “Graying of the MSAs” - may increase an individual woman’s likelihood of engaging in preventive health behavior, including having a screening mammogram. Moreover, living in age-restricted housing developments such as retirement communities, where an increase in face-to- face interactions is likely to occur, may generate large social multipliers in preventive health behaviors among the residents. The multiplier associated with education suggests that mammography promotion efforts that focus on raising awareness of breast cancer will be effective, since women influence other women’s decision through knowledge dissemination, role modeling, and experience sharing. Moreover, mammography promotional efforts that target black women are likely to achieve larger results in the long run, than is predicted by the usual models studying determinants of individual compliance with mammography recommendations. In addition, I find that the multiplier among women over age 75, for whom mammography is optional, is at least as large as for the group of women 50-74 year old. Knowledge of the magnitude of the effect of social interactions is important from the health policy perspective, since the cost of achieving public goal for screening rates may be much smaller than is predicted by the standard models