The following model can be used to study whether campaign expenditures affect election outcomes:
voteA = β0 + β1 log(expendA) + β2 log(expendB) + β3 prtystrA + u
where voteA is the percentage of the vote received by Candidate A, expendA and expendB are campaign expenditures by Candidates A and B respectively, and prtystrA is a measure of the party strength for Candidate A (the percentage of the most recent presidential vote that went to Candidate A’s party).
(i) True or False: If Candidate A’s expenditures increase by one percent, voteA will increase by approximately β1 percentage points.
True False
(ii) True or False: In terms of the parameters, the null hypothesis that a 1% increase in Candidate A’s expenditures is offset by a 1% increase in Candidate B’s expenditures is: H0: β1 = −β2.
True False
(iii) Use OLS to estimate the following model: voteA = β0 + β1 log(expendA) + β2 log(expendB) + β3 prtystrA + u.
The estimated equation is: voteA^ = + log(expendA) + log(expendB) + prtystrA.
True or False: As both the coefficient on log(expendA) and the coefficient on log(expendB) are statistically significant (at the 1% significance level), campaign expenditures by Candidates A and B have an effect on the percentage of the vote received by Candidate A.
True False
Leave a Reply