Lung cancer is associated with smoking and is characterized by low treatment rates and lower research funds with respect to other cancers. Research shows that patients with lung cancer often internalize societally biased beliefs on the effectiveness of treatment and the accompanying stigma, which may deter them from seeking treatment and, thus, hinder the diffusion of innovative therapies. We investigate the impact of social effects on treatment rates and innovation adoption using administrative data on advanced lung cancer patients in Ontario (Canada). We estimate a structural model of treatment choice where patients base their own decision on the treatment decisions of their reference group. Identification rests on the exogenous variation in the treatment propensity of physicians. We find that biased beliefs and stigma deter access to treatment: placing all patients in a neighborhood characterized by low social discrimination increases treatment rates by 4 percent and the use of innovative therapies by 3 percent. Social effects account for around 2 percent of the gap in research funding for lung cancer, which amounts to $7 million every year in US public funding alone.