In a hypothetical data set with 400 adults in the New York metropolitan area, 250 own their homes and 150 do not. Of the home owners, 245 own a car. Of the 150 non-home-owners, only 25 own a car. A researcher runs a regression of expenditures on movie entertainment on the dummy variable OwnHome and OwnCar. She finds that at the 5 percent significance level she cannot reject the two individual null hypotheses that the parameters multiplying these coefficients are zero, but that she can decisively reject the null hypothesis that both parameters are jointly equal to zero (P-value = 0.2%). Can you explain why she might have obtained such a result?