Matt Chingos and Paul Peterson have a new study based on a random-assignment experiment of the effects of privately funded vouchers on college enrollment. They followed the students who had received a private school scholarship in New York City. Earlier research found significant achievement benefits from that program for African American students but there was some controversy over whether those findings were robust to reasonable alternative specifications.
Now the students are old enough for college and Chingos and Peterson went back to see if achievement gains translated into higher college enrollment. They did. African American students who received a private school scholarship were 24% more likely to enroll in college. The improvement was more dramatic in the chances that African American students would attend private 4 year colleges and even selective private colleges.
No significant effects were observed for Latino students nor for the very few white students in the sample. It appears that choice has the biggest effect on those whose options would have been the worse in the absence of a choice program. It is also worth remembering that the scholarship was good for only half of private school tuition, so it is possible that a more generous program would have a broader effect.
You can read Chingos and Peterson’s summary of their findings in the Wall Street Journal.
Jay P. Greene's Blog 
!!!!BOOOOOOOOOOOOOOOOOOOOOOOOM!!!!
I challenge anyone to come up random assignment evidence of a bigger impact for a mere $4200 than more than doubling the percentage of Black students attending selective universities.
Well, since there was very little apparent impact of the voucher itself, I’d say you would have a hard time collecting. Since out-of-school factors are much greater than in school factors, the throw-away lines about the class-size costs were just that. Throw-aways. To provide real significant impact, try focusing on out-of-school factors. For example, books in the home are a great place to start. Average gap between students having between 0 and 10 books in the home, and those having over a hundred, is nearly two years on NAEP 4th grade reading. Costs of providing the home library of high-quality books geared to the 0-5 year old children in the home? About $250 per child in a three child family. Again, the impact was not really based on $4200 at all… the vouchers were “half-vouchers” so families had to supplement, and even the Catholic schools admit that tuition needs to be supplemented by what they call a “tuition supplement that in the 90’s was estimated at 180% of tuition.
Damn, I’m never going to get a steak dinner out of you if you keep offering bets that you can’t lose!
I will buy a steak dinner for Matt Ladner if he offers up a steak dinner bet that he can lose.
C’mon, Matt – I did it for Jay Mathews!
[…] among African American students who were awarded and used vouchers to attend private schools. As Jay Greene notes: No significant effects were observed for Latino students nor for the very few white students in […]
[…] among African American students who were awarded and used vouchers to attend private schools. As Jay Greene notes: No significant effects were observed for Latino students nor for the very few white students in […]
Jay, it is actually MUCH more likely that any effect found (the confidence intervals, when reported for an estimate of 20% increase in college enrollments for merely being offered a voucher range from 1% to 39%!) is due to the wide differences between the supposedly “random” assignment of students to receive the vouchers, vs. those applying for but NOT receiving the vouchers. Go back to the research document and read it more clearly! It might also be that any effect disappears completely if we actually did have random assignment to the different groups!
See a full critique of this study, plus a back and forth with the authors, here: http://nepc.colorado.edu/thinktank/review-vouchers-college
[…] back in 2000. Never mind that the positive findings she challenged were later vindicated by research showing significant increases for voucher students on high school graduation and college at…. It appears that one of the perks of being a NYT reporter is that you don’t have to get the facts […]