When making decisions, agents tend to make use of decisions others have made in similar situations. Ignoring this behavior in empirical models can be interpreted as a problem of omitted variables and may seriously bias parameter estimates and harm inference. We suggest a possibility of...