What happens to local communities when manufacturing disappears? I examine changes in associational density over nearly two decades as a proxy for social capital in US labor markets. Exploiting plausibly exogenous trade-induced shocks to local manufacturing activity, I test whether deindustrialization is associated with greater or lower organizational membership. I uncover a robust negative relationship between the two variables, particularly acute in rural and mostly-White areas. My findings, however, are sensitive to measurement: There are no clearly discernible effects of deindustrialization on social capital when I consider alternative proxies for the outcome. To reconcile these results, I present evidence suggesting that economic adversity may induce a qualitative, rather than quantitative, change in social capital.