Derivation of the logit model

This lecture focuses on the logit model, examining its key components: the choice set C_n, the error term epsilon_in, and the deterministic part V_in of the utility function. The assumption of independent random variables for epsilon_in is discussed, following a distribution with location parameter 0 and scale parameter mu. The importance of understanding the independence assumption across alternatives i and individuals n is highlighted, with the parameter mu being the same for all individuals and alternatives. The challenge of identifying the scale parameter mu from data is addressed, emphasizing the need for normalization to one during model estimation.