conditional bias term which may not vanish at a rate faster than root-N when more than one continuous variable is used for … conditional bias, matching estimators with a fixed number of matches do not reach the semiparametric efficiency bound for average … treatment effects, although the efficiency loss may be small. Third, we propose a bias-correction that removes the conditional …

can lead to over- as well as under-estimation of the coefficients of interest. Critical for determining the bias is the … error of the optimal prediction error form and show that such methods may in fact introduce bias. Finally, we present some …

almost all these papers ignore the bias in the estimated standard errors that serial correlation introduce4s. This is … serially correlated, which will exacerbate the bias in standard errors. To illustrate the severity of this issue, we randomly …

Empirical researchers routinely encounter sample selection bias whereby 1) the regressor of interest is assumed to be …

and contrast our results with those of other decompositions. We find that, over a 5-year period, the measurement bias …