∙ The Donald and Lang (2007) approach applies in the current setting by using finite sample analysis applied to the pooled regression (10). However, DL assume that the errors v_{gt} are uncorrelated across time, and so, even though for small G and T it uses small degrees-of-freedom in a t distribution, it does not account for uncertainty due to serial correlation in v_{gt}. 4. Individual-Level Panel Data ∙ Let w_{it }be a binary indicator, which is unity if unit i participates in the program at time t. Consider

y_{it } d2_{t } w_{it } c_{i } u_{it}, t 1, 2,

where d2_{t } 1 if t 2 and zero otherwise, c_{i }is an observed effect, and u_{it }are the idiosyncratic errors. The coefficient is the treatment effect. A simple estimation procedure is to first difference to remove

17

(11)