An analyst is interested in how education affects wage. The analyst specifies the economic model The analyst specifies a statistical model wage Bo+Beduc + e. = wage; Bo+Beduci + ui = unexper; +u. The analyst assumes SLR.1-SLR.3 hold, but SLR.4 may not. Specifically, the analyst fears Elu | educ] 0. The analyst regresses wage on education to obtain B. i. (10 points) Why may the analyst fear Efu | educ] 0? ii. (10 points) Suppose that Elu' | educ] = 0. The analyst thinks 10. Given this assumption, do you expect the OLS estimator B to be close to B in a large sample? If not, do you think , is systematically above or systematically below B. Please explain.