2. A researcher wants to estimate the effect of campaign spending on campaign vates. Assume that each election has two candidates. Let voted be the percentage of the vole for Candidate A. let exmeads be campaigning expenditures by Candidate A, Iet expendB be campaigning expenditures by Candidate B, and let rotexpend be total campaign expenditures; the latter three variables are all measured in dollars. Suppose that the model is specified as voted=0+1expendA+1expendB+3totexpend+u. a) What is the problem with this model? b) How would you fix this problem? c) Now suppose that we use as explanatory variables expendA, expend, and shareA, where share A=100 (expendA/totexpend) is the percentage share of total campaign expenditures made by candidate A. Would this model suffer from the same problem as model above