A two-stage estimator for probit models with structural group effects

Does immigration induce ‘native flight’ from public schools into private schools?

“…We therefore estimate a random effects probit model to obtain efficient estimates in two stages. See Borjas and Sueyoshi (1994) for a thorough discussion of the technique, and for proof that the estimation technique provides asymptotically unbiased inferences if the number of observations per group (in our case, MA / year combination) is sufficiently large.…”

“…As suggested in Borjas and Sueyoshi (1994), we estimate the second-stage regression using generalized least squares (GLS). Specifically, we use the weights on each MA in the second-stage regression.…”

“…(1) The sample consists of native-born children in the 1980 and 1990 Census who are currently enrolled in school and who are not living in group quarters. The table presents coefficients from a second-stage estimate based on the method of Borjas and Sueyoshi (1994). In the first stage a probit for student's enrolment in private vs. public school is run, including dummies for each MA and for each MA interacted with a dummy indicating whether the observation is from 1990.…”

Why Are Some Public Officials more Corrupt Than Others?

“…9 Borjas and Sueyoshi (1994) show that when the first stage is a probit, the weights necessary to recover the coefficients that would have obtained in a one step procedure are more complicated than the inverse of the squared standard errors. I show that the results are not sensitive to the weights.…”

Participation for What? A Policy-motivated Approach to Political Activism

“…I employ a two-stage random effects logit developed by Borjas and Sueyoshi (1994). First, separate logit models are estimated for each quarter from 1980 to 1994.…”