∙ As discussed by BDM, a common way to estimate and perform inference in (5) is to ignore v_{gt}, so the individual-level observations are treated as independent. When v_{gt }is present, the resulting inference can be very misleading. ∙ BDM and Hansen (2007b) allow serial correlation in v_{gt }: t 1, 2, . . . , T but assume independence across g. ∙ If we view (7) as ultimately of interest, there are simple ways to proceed. We observe x_{gt}, _{t }is handled with year dummies,and _{g }just represents group dummies. The problem, then, is that we do not observe _{gt}. Use OLS on the individual-level d a t a t o e s t i m a t e t h e g t , a s s u m i n g E z i g t ′ u i g t 0

and the group/time period sizes, M_{gt}, are reasonably large.

11