GPS2....

4. A city proposes to build a sanitary landfill. A researcher obtains a random sample of house selling prices in the city after the proposed location is known but before the landll is constructed. The researcher wishes to estimate how the announcement of citing the landfill affects house values. The researcher assumes the population model ln(houseprice) = [in + [31 ln(feet) where the variable houseprice represents selling price of a house ( measured in dollars) and the variable feet represents the distance ( in feet) from the house to the landll. The researcher obtains the following estimates: ln(houseprice) = 9 .8 + 0.521nlfeet) A. What is the precise interpretation of the estimate of 51? 0.52% increase in price per 1% increase in distance from proposed landll B. Suppose that we took the original variable feet and instead measured it in thousands. That is, suppose we did the following in Stata: gen feetwldll = feet/1000 gen lfeet_l|300 = inrfeet_1000) regress lhouseprioe lfeet_1000 Given the information provided in this question, is it possible to determine the slope coefcient the numerical valuefor this regression? if so, give the numerical value and briey explain how you arrived at this answer. if not, what is the minimum additional information that is required to determine it? That is, is there anything short of having the entire data set and estimating another 015 regression that would allow you to determine the slope coefficient in this regression? Aln(houseprice) Aln(fee:_1000)' Changing the units of the regressor to hundreds of feet doesn't change that Aln(feet _1000) is a proportional change. The slope still equals 0.52. The slope in the equation is still an elasticity, C. Again in the situation described in part B and given the information provided in this question, is it possible to determine the constant in this regression? if so, give the numerical value and briey explain how you arrived at this answer. if not, what is the minimum additional information that is required to determine it? That is, is there anything short of having the en tire data set and estimating another 0L5 regression that would allow you to determine the estimate of the constant in this regression? Note 5,, = ln(y) 0.52 ]n(x) = 9.8. After the transformation, the OLS estimate of the constant will be I W Bile = W 51(ln(I)in(1000))= ln(y) 31ln(y)+ almanac) = 5'0 +elin(1000) = 9.8 + .52ln(1000) m 13.4 2. Recall GPSZ Question 4: A city proposes to build a sanitary landll. A researcher obtains a random sample of house selling prices in the city after the proposed location is known but before the landll is constructed. The researcher wishes to estimate how the announcement of citing the landll aects house values. The researcher assumes the population model ln(houseprice) = o + A ln(jeet) + u where the variable houseprice represents selling price of a house (measured in dollars) and the variable feet represents the distance (in feet) from the house to the landll. The researcher obtains the following estimates: ln(h0useprice) = 9.8 + 0.52 ln( feet) Another researcher, however, believes that this model, which we refer to as model (1), is biased because it omits the variable age which represents the age of the house in years, i.e., this researcher believes that the population model is instead (2) 1n(houseprice) = o + A ln(feet) + ,b'zage + u Further, understand that the proposed location of the sanitary landll is near an older section of the city containing older houses and that, other things constant, consumers prefer newer houses to older houses. Assume the second researcher is correct. Should we have reason to believe that the slope estimate of 0.52 from model (1), which we now denote ] , is biased upwards (EWJ > A) or biased downwards (E [E 1]