Question

Following is an Estimated Multiple Regression for Cigarette consumption in the US. Based on the estimated parameters, and other statistics, Answer the following questions:

CigaConsm = 14.5 + 0.06LnInc – 0.65LnCigPr. + 0.025LnExcTax + 0.034Gender

T-stats:            (2.90) ( 1.30)         (-2.25)                (2.40)                   (1.67)

Where CigaConsm represents cigarette consumption in millions of boxes per year in a given state; Inc is median household income of the State; Cigpr is cigarette price per pack; Exctax is Excise tax per pack of Cigarette, and Gender is the Dummy variable designating if the state has 50% or more female.

R-Square = 0.92          T-Critical: 1.96 at 46 degrees of freedom; n (sample size) = 51.

1. Interpret the Estimated parameters

2. If price of Cigarette were to increase by 10%, some may claim that the consumption will go down by 20%; Are they right? Why or Why not?

3. Suppose it is argued that Female in general smoke less than Male, Do you agree or disagree? Why?

4. If the State of Wyoming decides to increase the Excise Tax rate by 20% from its present rate, $1.50 per pack, then their Tax revenue from Cigarette alone will go up by 25%. Do you agree or disagree? Why? What would be the correct amount of Revenue increase given that the price per pack of Cigarette in Wyoming is $6.50 and they sell about 5 million packs a year.

5. Is there any problem in this estimated Regression? If so, name a few.

Answer 1

A) Interpretations:

1. For 1% increase in the median household income of the State the cigarette consumption in millions of boxes per year in a given state increases  by 0.006 (0.6%)units keeping all other factors constant.

2. For 1% increase in the cigarette price per pack the cigarette consumption in millions of boxes per year in a given state falls by 0.0065(0.65%) units keeping all other factors constant.

3.For 1% increase in the Excise tax per pack of Cigarette the cigarette consumption in millions of boxes per year in a given state increases by 0.00025(0.025%) units keeping all other factors constant.

4. If the state has 50% or more female then the  cigarette consumption in millions of boxes per year is 14.534 units higher than if its not the case keeping all other factors constant.

B) NO, reason explained in part a as 1% - 0.65% fall

10%- 6.5% fall

C) YES , I do agree.  

But the regression results shows opposite , one of the issue in the model.

D) Incorrect

E) Incorrect coefficient signs