The file FARM.csv contains data on output quantities (QL, QC and QO), input quantities (XK, XL, XA and XM) and environmental variables (ZD and ZR) for 11 states in the Northeast Farm Production Region of the U.S. for the period 1960-1989. The file also contains data on output prices (PL, PC and PO) and input prices (WK, WL, WA and WM). There are a total of T = 11 x 30 = 330 observations in the dataset. Economic theory suggests a relationship of the form: Model A: ln QL = b1 + b2*t + b3*ln ZD + b4*ln XK + b5*ln XL + b6*ln XA + b7*ln XM + b8* ln (QC/QL) + b9* ln (QO/QL) + e where t = 1, ..., 30 is a time trend and e is a random variable with E(e) = 0. Treat the 5 variables in question 6
Q6 ie. (A researcher suspects that the variables ln XK, ln XL, ln XM, ln (QC/QL) and ln (QO/QL) are correlated with the error term and decides to conduct a Housman test at the 5% level of significance using the 7 prices as instruments (n.b., prices, not log-prices). The value of the test statistic is F = Blank 1. If the null hypothesis is true then the test statistic follows an F-distribution with Blank 2 and Blank 3 numerator and denominator degrees of freedom (report the degrees of freedom as integers). The researcher should Blank 4 (reject/not reject) the null hypothesis.)
As endogenous and obtain IV estimates of the model parameters using the 7 prices as instruments (n.b., use prices, not log-prices). The estimate of b6 is Blank 1 with a standard error of Blank 2.