Assume that the demand function takes the form q=p. Taking logs in the demand function we obtain the following estimating equation: log(q)=log()log(p)=~log(p) where q is per capita beer consumption, p is the total (taxed) price of alcohol p=ppre(1+). Notice that log(q)=~log(1+)log(ppre). Since we do not have access to prices, our estimating equation is: log(qit)=log(1+it)+it where i is the state and t is the year, =~E[log(ppre)] and it=E[log(ppre)]log(pitpre). That is, the error term equals the (unobserved) variation in pre-tax prices in different states. Write down the exogeneity condition for this estimating equation. Assume that the demand function takes the form q=p. Taking logs in the demand function we obtain the following estimating equation: log(q)=log()log(p)=~log(p) where q is per capita beer consumption, p is the total (taxed) price of alcohol p=ppre(1+). Notice that log(q)=~log(1+)log(ppre). Since we do not have access to prices, our estimating equation is: log(qit)=log(1+it)+it where i is the state and t is the year, =~E[log(ppre)] and it=E[log(ppre)]log(pitpre). That is, the error term equals the (unobserved) variation in pre-tax prices in different states. Write down the exogeneity condition for this estimating equation