The strange death of education reform, part three

www.slowboring.com

The strange death of education reform, part three

Matthew Yglesias

14 - 18 minutes

This is the third of an ongoing series; part one is here and part two is here. And a related guest post is here. Also — if you’re in or near Chicago tomorrow, come hang out at an informal Slow Boring happy hour starting at 5:30 p.m. at Radio Room, 400 North State Street.

One question I got from a lot of conservatives following the initial installments in this series is why am I overcomplicating things instead of just saying that teachers unions are bad.

The reason is that I think what we’ve learned from charter schools tends to debunk the strong critique of teachers unions.

Now to be clear, the charter experience does force what I would call the narrow critique of teachers unions — that in specific cases where doing something bad for the union would almost certainly be good for students, the union (like almost any institution) instead does what’s good for the union.

But the strong criticism of teachers unions is that collective bargaining is such a pernicious force in the public education system that simply sweeping it aside could dramatically improve performance. Charter schools are a critical testing ground for this claim because they have an enormous amount in common with traditional public schools. Most notably, a charter school cannot institute a competitive admissions process the way a private school or a “good” college or graduate school program would. Charter schools, like public schools, have to accept eligible students who want to attend. Charters are allowed to prevent over-enrollment by means of a lottery, which gives us lots of good opportunities to compare outcomes for students who win the lottery to outcomes for those who lose. This is excellent for social science research — and it shows that some charter schools and, notably, some charter networks do deliver consistently above-average outcomes.

But in terms of the strong critique of teachers unions, it also shows very clearly that on average, charter schools freed from the shackles of collective bargaining agreements are no better than traditional public schools. In many cases, they are worse.

Teachers unions support Democratic Party politicians, so if you favor tax cuts, banning abortion, and making it easy to buy a gun, you have very good reason to support any proposal to weaken them. But charter schools are a good test of the proposition that ditching unions would make schools obviously better, and they show pretty clearly that it doesn’t work. This is why there’s been a tremendous backlash against charters not only on the left (see “Abbott Elementary” and Jessica Winter’s lavish praise of it in the New Yorker) but on the right as well, where the current fad is for “universal” choice on the model of higher education.

My view is that this is misguided. The charter school experiment’s successes and failures are, I think, genuinely informative about education policy, and we should be mining them for insights, policy improvements, and expectation-setting. Instead, we’re moving toward an equilibrium that’s going to be dangerously bad for the most disadvantaged students.

Charter schools are a slightly odd idea, and because K-12 education is highly decentralized in the United States and the charter school situation varies from place to place (much like the baseline public school situation), there’s a lot of confusion about exactly what they are and how they work.

The whole country is divided up into a bunch of school districts, and there’s very little rhyme or reason to how these districts work. The biggest school district in America is New York City Public Schools, which serves kids who live in New York City. Number two is the Los Angeles Unified School District, which serves kids who live in the city of Los Angeles, but also some adjoining areas of LA County. Those are the two biggest cities in America, so you expect to see Chicago third on the list, but #3 is actually Miami-Dade Public Schools, the school district for all of Miami-Dade County, not just the City of Miami. Then Chicago is #4 because it’s only for the city of Chicago and doesn’t cover the surrounding Cook County. And the whole list is full of these quirks — Dallas Independent School District covers most but not all of Dallas County in Texas.

The districts themselves also operate differently. In the rural northeast, a small town is typically its own school district and only operates one school. In other parts of the country, there are tons of small-but-not-that-small districts like Kerrville Independent School District over where my in-laws live, which runs four elementary schools and two high schools. But KISD is still pretty small. There’s one school district for all of Clark County, Nevada, but Kerr County, Texas contains eight separate school districts.

Because school district boundaries often don’t line up with the exact borders of municipalities or counties, the districts are governed by an independently elected school board. But in cities where they do perfectly match the city boundary, they’re often under “mayoral control” like the police department or any other municipal function. “County control” — when a school district perfectly lines up with a county — is less common. Sometimes you have a case like San Francisco where the city is also a county and also a school district, but nonetheless the school system is a separate entity run by its own elected board.

In virtually all cases, a student has the right to attend a school in the district where he lives but not in nearby districts where he doesn’t live. When a district contains multiple schools, the most common situation is that you have the right to attend a specific school based on your address but not to attend different schools in different neighborhoods. But a large public school district can offer a range of options, including some public schools with selective admissions criteria. In Boston public schools, admission for all students is determined by a kind of lottery. Students in D.C. can lottery into out-of-boundary schools, but they have an absolute right to attend their in-boundary school, so it is very hard in practice for out-of-boundary kids to get into the in-demand schools west of Rock Creek Park since the people who live there take up the bulk of the spots.

So what’s a charter school?

    First and foremost, a charter school is not owned or administered by a public school district.

    In practice, this means that charter schools are rarely covered by collective bargaining agreements between teachers unions and public school districts.

    Charter admissions are normally limited to residents of a particular school district, but not limited to residents of a particular neighborhood within that district.

    A charter school gets public funding via a formula determined by the number of students it enrolls, but it can also raise money separately from the public system.

    A charter school cannot charge tuition.

    A charter school cannot use selective admission criteria; if enrollment exceeds capacity, admission has to happen by lottery.

Charters operating in liberal areas tend to characterize themselves as “public charter schools” because they are trying to win political fights on left-skewing terrain. Critics, though, will point to (1) to argue that these are not public entities at all. But fans will point to (3) and especially (6) to note that public charter schools are open to the public in a way that traditional public schools are not. The best public middle school in D.C. is Alice Deal Middle School, and access to that school is auctioned via the real estate market. If you want your child to attend, you need to buy or rent a house in one of the city’s most expensive neighborhoods. One of the more in-demand public charter middle schools in D.C. is BASIS, and to get in there you need to win a lottery in which everyone — rich or poor — has essentially equal odds.

In practice, though, the reason charters are controversial is (2) and (4). For labor unions representing the public school workforce, students and money bleeding out of the system and into a parallel school system is very bad. Alternatively, if you’re convinced that teachers unions are wrecking public education and schools freed from the shackles of collective bargaining agreements would perform much better, then charter schools seem like a great option.

If you don’t care at all about education but do care a lot about labor unions, then the main thing you need to know is that the union haters’ optimism about charter schools is basically wrong.

Just saying “hey, here’s a school without a teachers union!” does not miraculously generate huge learning gains. There’s now a long list of charter school evaluations conducted by CREDO at Stanford, and they are all over the map. Note that one cool thing about charters is that the lottery admissions method is good for empirical research. You can compare lottery winners to lottery losers and get a good estimate of the difference. And the studies show that the winners are better off in some states but not in others.

It’s also worth saying that in most cases, these effect sizes are pretty modest and the big exceptions to that are in small states (Rhode Island, D.C., Nevada), which raises the possibility that random chance is partially driving the results. Now note that this measures the gap between a state’s traditional public schools and its charter schools, not the quality of the charter schools as such. Traditional public school NAEP scores are higher in Texas than in California, so if the charters are equally good in both states, that would register as charters being beneficial in California and harmful in Texas.

So that’s bad news for monomaniacal union haters.

That said, there are some clear success stories. The charter schools in New York, Massachusetts, D.C., and New Jersey all seem to be clearly beneficial. And when researchers zero in on those states, they find that there are particularly large charter school benefits in urban areas. A natural thing in those circumstances would be for charter schools to expand at the margin and enroll a larger share of students. In New York, this is tied up in endless real estate wrangling (what in New York isn’t?) with the unions always on the side of making it harder for the charters to expand. In Massachusetts it’s worse: there’s a firm statutory cap on the number of charter schools. Reformers put together a ballot initiative to lift the cap, and the union — along with the state’s top Democratic Party politicians — mobilized to defeat it. So in summary:

    Non-union schools aren’t automatically better than union ones — in some cases, they are worse.

    When non-union schools are better, unions mobilize to stop kids from attending them anyway.

I think all this research leads to some pretty straightforward policy recommendations. States where charters are showing substantial benefits should let their charter sectors expand. States where charters are showing substantial harms should become much more vigorous about shutting down the worst-performing schools. States that are in the middle should try to do a little bit of both. And everyone should be enthusiastic about improving educational outcomes while tempering expectations about exactly how much sweeping social change can be effected just by improving K-12 schools.

Instead, the Democratic Party has taken a hard tilt in favor of the union position, while Republicans have turned against the regulatory constraints that, in my view, make charter schools good.

The issue on the left is pretty straightforward. Unions have always hated charter schools for pretty rational reasons, and they don’t like good charter schools any more than bad ones. As long as Barack Obama was president, the charter movement had a powerful ally, but Joe Biden has never been a charter schools guy. Biden owns the “moderate Dem” brand, so with him siding with the left on this, the pro-charters side is on the defensive. There are two good Jon Chait articles on this if you want details. I’d just add that status quo bias is a big deal, and despite anti-charter sentiment ruling the roost in Democratic Party politics, it’s not as if charter-heavy cities are getting rid of existing schools that people like.

Meanwhile, because the left has been so hostile to charters all along, not much attention has been paid in progressive circles to the extent to which conservatives have given up on this idea. But as Corey DeAngelis — who’s affiliated with an overlapping network of right-of-center education groups — bragged last week, the move to what they call “universal school choice” is on the march now in a lot of red states.

What does universal school choice mean in practice? I think this explainer from the Arizona Department of Education is probably as good of a concise and neutral explanation as you’re going to find:

    What is an ESA? An Empowerment Scholarship Account (ESA) is an account administered by the Arizona Department of Education (ADE) and funded by state tax dollars to provide education options for qualified Arizona students.

    An ESA consists of 90% of the state funding that would have otherwise been allocated to the school district or charter school for the qualified student (does not include federal or local funding). By accepting an ESA, the student's parent or guardian is signing a contract agreeing to provide an education that includes at least the following subjects: reading, grammar, mathematics, social studies and science. ESA funding can be used to pay private school tuition, for curriculum, home education, tutoring and more.

    Can my student participate in the ESA Program and remain his/her public or charter school? In almost all situations, students on ESA cannot be enrolled in a public school (this includes district, charter and public online programs). This is a violation of the ESA contract. Additionally, the student must be withdrawn from the public school at the time that the ESA contract is signed. If the state is paying for an ESA student's education at the same time as the student is receiving ESA funds, the state is distributing the student's education funding twice. You may be able to pay for certain public school programs utilizing your student's ESA funds, please contact us for more details.   

I completely understand why this idea appeals to conservatives more than charter schools do — homeschoolers and people who want to send their kids to religious schools are important elements of the conservative base, and this helps them whereas charter schools don’t. Charter schools were part of the education reform moment when everyone was focused on the idea that we needed to improve learning outcomes for low-income students, especially urban Black and Hispanic students. Universal choice comes out of a post-reform moment when conservatives are focused on stuff conservatives care about more deeply, like supporting religious people. But I still think the problems of low-income people, including those living in cities, are important, and I don’t think this move to universal choice is going to be good.

But that critique is for another edition.