Exercise 2. Suppose that the population model determining y is y = Bo + Bx + Bx + Bzx3+U And this model satisfies CLRM assumptions. However, we estimate the model that omits x3. Let Bo, and be the OLS estimators from the regression of y on x and x. Show that the expected value of B is n i=1 Pixi3 E(B)=B + B3 where f; are the OLS residuals from the regression of x on x. Hint: In the model 9 = Bo + x + x, B can be expressed as residuals from the regression of x on x. yn where fi are the OLS Plug y = Bo + Bx + Bx2 + 3x3 + u into this equation. After some algebra, take the expectation treating x3 and f; as nonrandom.