Theorists have proposed that a value of close friends and family—strong ties—is the ability to confide in them when facing difficult issues. But close relationships are complicated, and recent studies report that people sometimes avoid strong ties when facing personal issues. How common is such avoidance? The question speaks to theoretical debates over the nature of “closeness” and practical concerns over social isolation. We develop an approach and test it on new, nationally representative data. We find that, when facing personal difficulties, adult Americans are as likely to avoid as to talk to close friends and family. Most avoidance is not actively reflected on but passively enacted, and, contrary to common belief, is not limited to either specific network members or particular topics, depending instead on the conjunction of member and topic. Building on Simmel, we propose that a theory of the fundamental need to conceal and reveal helps account for the findings. We suggest that there is no more empirical justification for labeling strong ties as those who are trusted than for labeling them as those who are avoided. In turn, isolation might be less a matter of having no intimates than of having repeatedly to avoid them.

Studying causal effects of continuous treatments is important for gaining a deeper understanding of many interventions, policies, or medications, yet researchers are often left with observational studies for doing so. In the observational setting, confounding is a barrier to the estimation of causal effects. Weighting approaches seek to control for confounding by reweighting samples so that confounders are comparable across different treatment values. Yet, for continuous treatments, weighting methods are highly sensitive to model misspecification. In this article we elucidate the key property that makes weights effective in estimating causal quantities involving continuous treatments. We show that to eliminate confounding, weights should make treatment and confounders independent on the weighted scale. We develop a measure that characterizes the degree to which a set of weights induces such independence. Further, we propose a new model-free method for weight estimation by optimizing our measure. We study the theoretical properties of our measure and our weights, and prove that our weights can explicitly mitigate treatment-confounder dependence. The empirical effectiveness of our approach is demonstrated in a suite of challenging numerical experiments, where we find that our weights are quite robust and work well under a broad range of settings.