It appears Milwaukee’s School Choice program produced somewhat better outcomes at much lower taxpayer expense than the public schools

See At Least as Good at a Fraction of the Cost? Some “Flop”! by Neal McCluskey of Cato.

"A lot of well-intentioned people think it is not enough for families to be able to choose schools. They have to choose “good” schools. Those people often do not think private school choice programs that give parents a lot of control over which schools they select are up to par. Fine. But just because you don’t like something doesn’t make it a “clear flop.”

Writing at The 74, Richard Whitmire warns that we should beware Trumps bearing school choice gifts. He argues that President-elect Trump’s proposal to spend $20 billion on school choice could be dangerous not because of, say, federal rules that might be attached to unconstitutional largesse, but because the money might not be restricted to “great” schools. “Great,” presumably, should be defined by legislators or bureaucrats. After all, you don’t want to replicate the Milwaukee voucher program:

Those in the school reform movement learned the hard way that choice alone does not produce more seats in great schools. If that were the case, we’d all be praising the early voucher program in Milwaukee and the widespread charters in Ohio and Michigan. But in all those cases, choice alone produced nothing.
In Milwaukee, for example, which I visited repeatedly while researching my book On the Rocketship, about the creation of one best-in-class charter network, the more-than-two-decade-old voucher experiment proved to be a clear flop. (Note that I didn’t say unpopular. Who objects to free tuition for their kid’s parochial schools?)

Set aside the first evidence that Milwaukee’s program isn’t a “clear flop”: It is popular, indicating that the people it is supposed to serve are at least getting something they want. What about other important measures, including test scores, graduation rates, competitive effects, and costs? According to researchers at the University of Arkansas’ School Choice Demonstration Project, who intensively studied Milwaukee:

Our main findings included that the program had a positive effect on a student’s likelihood of graduating from high school and enrolling and persisting in a 4-year college. We found little evidence that the Choice program increased the test scores of participating students, though our final analysis revealed a positive effect of the program on reading scores when combined with high stakes testing. There was no evidence of program effects on math scores. Competition from the Choice program appears to have boosted the test scores of students who remained in Milwaukee Public Schools (MPS), but those systemic effects of the program were modest in size. Because the maximum value of the voucher…is substantially less than what the government pays to educate students in MPS, the state saves over $50 million per year from the operation of the program. 

Is the choice program transformative? No. But a flop? It appears to have produced somewhat better outcomes at much lower taxpayer expense than the public schools. It is also nowhere near a free market, with regulations constraining admissions policies, hours of instruction, and testing. And freedom is the key to unleashing competitive pressures, specialization, and innovation.

The Milwaukee voucher program is not a flop, and making policy based on the idea it is would be a mistake."