It didn't. The Washington Post reported late yesterday:
Students in the nation's only federally funded school voucher initiative performed worse on standardized tests within a year after entering D.C. private schools than peers who did not participate, according to a new federal analysis that comes as President Trump is seeking to pour billions of dollars into expanding the private school scholarships nationwide.
The study, released Thursday by the Education Department's research division, follows several other recent studies of state-funded vouchers in Louisiana, Indiana and Ohio that suggested negative effects on student achievement.
The more the evidence mounts against the idea, the more Team Trump is left to argue that they want a policy that plainly doesn't work.
And before conservatives argue that this D.C. study is only one report, consider that there's ample evidence pointing in the same direction. The New York Times reported in February, "[E]ven as school choice is poised to go national, a wave of new research has emerged suggesting that private school vouchers may harm students who receive them. The results are startling -- the worst in the history of the field, researchers say.... It's rare to see efforts to improve test scores having the opposite result."
For federal policymakers, the question is simple: does real-world evidence matter more than ideological goals or not?
Postscript: Just as an aside, note that the Obama administration considered scrapping the D.C. voucher system, but then-House Speaker John Boehner (R-Ohio) made a direct appeal to the Democratic president, saying it was personally important to him that the voucher program continue. Obama, trying to be conciliatory, agreed to leave it alone.