Billionaires are spending their fortunes reshaping America’s schools. It isn’t working.

Education

Story image for Education from Vox

Philanthropists need to take a step back from the American education system before it ruins them, they ruin it, or both.

Major philanthropies like the Gates, Walton Family, and Broad Foundations are spending hundreds of millions of dollars annually in an attempt to transform American K-12 education. Gates alone reported spending nearly $390 million in 2017; the Waltons spent more than $190 million. That’s a non-trivial chunk of the $67 billion all US foundations spent on all projects that year.

Such contributions have come under fire in recent years. The big foundations promote a particular set of K-12 education policies — including increased accountability for teachers, more school choice, and higher-stakes testing — that are profoundly controversial, and that teachers unions and skeptical education researchers have spent years questioning and resisting. The foundations’ use of billions in spending to change public policy on education raises troubling questions about democratic accountability and the role of money in politics (questions given new prominence when a major conservative education funder became US secretary of education).

Those are both valid lines of critique, but they’re not the ones I’m going to pursue here. (I am frankly more sympathetic to the Gates/Walton/Broad education reform agenda than a lot of my left-leaning friends.)

My beef, rather, is that improving the American education system, while important, is neither a neglected cause nor a tractable one. It is a system on which hundreds of billions of dollars are spent annually by diffuse governments whose policies are difficult and expensive to change, where matters of importance are intensely contested, and where interest groups tend to fight each other to a standstill.

And it’s a system where, even after investing millions if not billions in research, we still don’t have a lot of confidence as to which interventions are helpful and which are not. The views of key actors, notably the Gates Foundation, have tended to shift rapidly on those substantive questions.

If every issue in the world were as crowded and hard to make progress on as education in the US, then I’d understand why foundations like Gates and Broad keep chugging. But that’s not the case. There really are areas that are very, very important, and where progress is easier, because the political fights around them are less crowded and intense. Many of these foundations are already investing in a few of these causes.

That, if anything, makes their continued focus on US education more baffling. They know there are better causes. They should lean into them.

The framework for judging causes based on importance, neglectedness, and tractability (INT, for short) is not new. It originates with the Open Philanthropy Project, a San Francisco-based nonprofit focused on finding high-impact giving opportunities, and its predecessor organization GiveWell Labs.

Open Philanthropy and its sister foundation Good Ventures, which is funded primarily by the groups’ president Cari Tuna and her husband, Dustin Moskovitz, use the framework to guide most of their hundreds of millions of dollars in charitable contributions. The nonprofit 80,000 Hours, which offers career guidance to people looking to make a social impact in their job, uses the framework too, and the Oxford philosopher Will MacAskill has a useful breakdown of how it works in his book Doing Good Better.

It’s hardly perfect, but I find it very useful. A good baseline test for any philanthropist is that they work on issues that are genuinely important but that’s not enough of a filtering mechanism. A lot of stuff is important! Comparing relative importance is certainly possible; I think arts education is wonderful and important, but few would dispute that, say, funding field trips to art museums for rich US boarding school students is less important than making sure kids in extremely poor countries like Burundi, Afghanistan, or Haiti get quality instruction in math and reading.

But importance comparisons are often tricky and subjective, especially within given cause areas. If you’re interested in local US poverty alleviation, should you fund a soup kitchen or a homeless shelter? If you want to give to global health causes, should you donate to charities handing out insecticidal bednets to people in sub-Saharan Africa or to charities doing mass deworming projects?

That’s where thinking about neglectedness and tractability can help. Even when choosing among equally important causes, you probably should choose the cause that is more neglected, that has less money and fewer resources mobilized behind it than others. The basic reason is diminishing marginal returns: The first million dollars you spend on something is likely to have a much bigger impact than the second million, so it pays to look for causes where you’re closer to the first million spent than the second. Keeping neglectedness in mind can also prevent duplicating work that would’ve been done by another organization.

For instance, mass immunization campaigns targeting common diseases like polio, measles, and yellow fever are super-effective ways to prevent disease and save lives. And precisely because of that, the GAVI Alliance, UNICEF, and the World Health Organization spend billions every year on immunization programs, from a variety of funders ranging from governments to big foundations (Gates alone has donated billions to this cause). It makes sense for other funders to look at that situation and think, “They have this covered — why don’t I try something else?”

More to the point, even if that funder decided they ultimately did want to fund immunizations, they’d probably pay more per immunization than GAVI and other existing groups do. The existing firms are probably getting all the low-hanging fruit, forcing this hypothetical new funder to look for harder cases that are more expensive.

Tractability is the final criterion in this framework, and worth breaking down a bit further. For a cause to be tractable, funders don’t just have to know of cost-effective interventions that can help; they have to know ways to get those interventions adopted. If you, for some reason, supported reviving alcohol prohibition in America, there’s basically no way a temperance movement is going to succeed in the 21st century. Rather than wasting billions in an effort to revive the 18th Amendment, you should direct it somewhere it has better odds of effecting change.

K-12 education isn’t neglected

So let’s apply this framework to K-12 education. It is not a neglected cause. The US spent $668 billion on public elementary and secondary schools for the 2014-’15 school year. That’s more than the US spends on the military (about $583.4 billion in calendar year 2015), and in terms of total expenditures is probably only outpaced by health care and Social Security/pensions among government priorities.

That is not a bad thing! Good education costs money, and a few particularly rigorous economics studies in recent years have concluded that spending more on education typically improves student performance, as you’d hope. And under some circumstances, the heavy government investment in education would make a philanthropic investment more advisable, not less. If you can spend a few million dollars on think tanks, lobbying, etc. and change government policies affecting billions in annual spending, that’s pretty damn good leverage.

But the cause of trying to shift K-12 education resources in one direction or another is also extremely crowded. Teachers unions like those organized under the American Federation of Teachers or the National Education Association are not the all-powerful behemoths that some reformers paint them as being (and the Supreme Court’s ruling barring them from collecting fees from nonmembers promises to weaken them considerably). But they did spend $35 million in campaign contributions in 2016, and they spend millions more on lobbying annually.

If you consider their collective bargaining activities a form of political spending meant to shift public policy, their investment increases substantially. That spending coexists with the hundreds of millions that foundations like Gates, Walton, and Broad devote to the issue.

This spending is not additive; it’s clashing. And that’s a problem, from the standpoint of effectiveness. The most careful study of lobbying in America yet conducted — Lobbying and Policy Change by Frank Baumgartner, Jeffrey Berry, Marie Hojnacki, David Kimball, and Beth Leech — concludes that one reason the political system is biased toward preserving status quo policies is that groups lobbying on a given issue typically “faced organized opposition with roughly similar resources.”

That is, policy doesn’t change because two evenly matched sides fight each other to a draw. So when there’s a policy change that, say, Walton Family Foundation-supported reformers want for schools in DC, and the Washington Teachers Union wants to maintain the status quo, the two are likely to mobilize considerable resources to achieve a result no different from the status quo.

That’s, in a way, a win for the Washington Teachers Union, but the same dynamic plays in reverse if, say, WTU wanted to implement a change to DC education policy and Walton wanted to hold the line. The result is each spending millions to maintain a policy status quo that leaves neither fully satisfied.

That’s just how democratic politics, of all sorts, tends to work. Passionate interests on each side of an issue mobilize and fight and the stronger coalition, whether measured in money or grassroots support, ultimately prevails. And despite its considerable virtues, it’s a pretty wasteful process, especially on highly controversial topics like education in which much of the public is heavily invested. For philanthropists especially, it’s doubtful that investing heavily in a crowded, high-profile topic will turn out to be the most efficient possible use of resources.

K-12 education isn’t tractable

And then there’s tractability. The non-neglectedness of K-12 education in America itself makes the issue less tractable, for the reasons described above. But there’s also the problem that, despite massive investments in education research by Gates and other foundations, expert disagreement persists about what actually works. More than that, disagreement and shifting positions on interventions is common even within foundations.

Consider the small schools experiment. In the late 1990s and early 2000s, Gates, in conjunction with allied foundations like the Annenberg Foundation, Carnegie Corporation, Open Society Institute, and Pew Charitable Trusts, began offering well over $2 billion in grants to encourage school districts to break up large high schools into smaller schools. Gates was always eager to emphasize the schools were not merely small, but that breaking up schools into new institutions allowed those institutions to develop “high expectations,” “performance-based” cultures; The small size also allowed more personalized attention on individual students, or so the theory went.

The experiment, or at least Gates’s involvement, did not last very long. In 2005, an evaluation from the American Institutes of Research and SRI International suggested the intervention was falling short of expectations. His most vociferous critics alleged that the intervention was based on a basic statistical fallacy (inferring that small schools are good because they’re overrepresented among top high schools), a story that made its way into Daniel Kahneman’s pop science best-seller Thinking, Fast and Slow.

In his 2009 “annual letter,” Bill Gates conceded, “Many of the small schools that we invested in did not improve students’ achievement in any significant way … in most cases, we fell short.” The foundation mostly abandoned the small schools idea.

Its next project, the Measures of Effective Teaching (MET) project, pivoted from trying to promote a certain model of school to trying to figure out and promote a good model of teacher evaluation (the exceptional education reporter Dana Goldstein has an excellent paper looking at the MET investment in depth). Gates enlisted the well-respected Harvard education economist Thomas Kane, who conducted extensive randomized research concluding that evaluators could use a mix of standardized tests, student evaluations, and videotaped classroom sessions to identify teachers who cause more student learning to happen.

In addition, starting in 2009, the foundation began funding trials in three school districts and four charter groups, where they tested the new teacher evaluation methods they were developing. The hope was that using better evaluation methods would lead to more student learning or higher high school graduation rates.

No dice. A RAND Corporation report found that overall, the effects of the changes were minimal, at times even harmful.

“Several years into the initiative, there was evidence that it was helping high school reading in Pittsburgh and at the charter networks, but hurting elementary and middle school math in Memphis and among the charters,” Matt Barnum at Chalkbeat summarizes. “In most cases there were no clear effects, good or bad. There was also no consistent pattern of results over time.”

Allan Golston, who runs the US program at the Gates Foundation, recently suggested that the organization was ready to pivot again in response, saying, “We have taken these lessons to heart, and they are reflected in the work that we’re doing moving forward.” Even before the study’s release, Melinda Gates told the Associated Press that the program was a disappointment.

But over the past decade, as Gates abandoned small schools and pivoted to teacher evaluation work, a funny thing happened. Two randomized studies on small schools came in suggesting that Gates had been too hasty and they did improve student performance. In a series of reports released between 2010 and 2014, MDRC found that New York City students randomly sorted into small public high schools had higher graduation rates (71.6 percent versus 62.2 percent) and higher college enrollment (49 percent versus 40.7 percent) than those in traditional large schools.

Economists Atila Abdulkadiroğlu, Weiwei Hu, and Parag Pathak also evaluated the New York City small schools, exploiting the lottery used to determine enrollment in the schools, and found that they boosted college enrollment, reduced the use of remedial college courses, and improved test scores in math, English, science, and history. A few non-randomized studies reached the same conclusion.

But while he acknowledged the new evidence in a note last year, Gates has moved on from small schools. He operated at a swifter pace than that at which the evidence rolled in. It’s enough to make one wonder if the conclusion of the teacher effectiveness effort — whose RAND evaluation wasn’t randomized — will prove different as well, and subsequent research will redeem the approach.

Or maybe it won’t! Or maybe the small schools studies and their promising findings won’t replicate. My point isn’t to take one side or another. It’s merely to point at the Gates experience investing in education and ask: What good did they do here? They provided policymakers with some new evidence, for sure, but pilot programs might have done the same without costing hundreds of millions of dollars. But did we get closer to an effective school system because of the Gates investments? Did we move farther away?

Even if Gates does eventually settle on an approach to education policy that seems highly effective, will it be cost-effective relative to something dead simple, like handing out cash instead?

That might sound like a silly question, but the Brookings Institution’s Russ Whitehurst estimates that cash programs, like the earned income tax credit, do considerably more to boost student test scores than even education interventions generally known to be somewhat effective, like reducing class size or investing in pre-K. It’s a good benchmarking question to ask about any educational intervention, and it’s far from clear that there are improvements to US schools out there that exceed that bar.

Again, all of this sidesteps the question of whether the Gates Foundation could successfully spread effective policies once it identified them — and if it could do the policy spreading in a cost-effective way as well. Maybe it could! But it’s hard to do when you can’t even determine the effective policy you want to spread in the first place.

I don’t mean to argue that all of Gates’s spending has been wasted. He’s funded some good, important research. But when he announced another huge $1.7 billion education push in fall 2017, the obvious question was: How much of that money will bear fruit? And what are the odds that Gates will wind up abandoning this approach as well?

What education funders could fund instead

Gates’s lukewarm track record on US education hasn’t kept other billionaires out of the field. Quite the contrary: Education is only becoming more popular as a cause area.

The Chan-Zuckerberg Initiative, the charitable group founded by Priscilla Chan and her husband, Facebook CEO Mark Zuckerberg, has made education one of its top cause areas, pledging to “help every child access personalized educational experiences that can change the trajectory of their life.” Again, in theory and in a vacuum, an admirable and important goal; in practice, likely a quixotic one.

Indeed, a town in Connecticut wound up rejecting a CZI-sponsored personalized learning program when, New York magazine reports, “bizarre and sometimes inappropriate images appeared on their kids’ screens on third-party websites used as reading assignments: a pot plant, a lubricant ad, and then the coup de grâce, an ancient Roman statue of a man having sex with a goose.”

All charities face a learning curve, of course, and Cheshire, Connecticut, is alone in rejecting the program, while some 380 other school districts and charter schools still participate. But that raises its own questions: Why is CZI funding a program in hundreds of school districts without doing a small-scale, independently evaluated pilot to see, preliminarily, if it works? Why have Gates and Bezos signed on as funders for this program too?

The Bezos Family Foundation’s website, smartly, warns that “there is no silver bullet” when it comes to education — but is quick to add that “education is the silver bullet.” While pre-K is less crowded as a cause than K-12 education (or higher ed, which is its own whole mess of a situation), Bezos’s decision to spend billions funding his own line of preschools in the US without a comparable investment abroad betrays an implicit judgment that improving government services in the richest country on earth is higher-impact than doing that in, say, a poor country without universal education of any kind.

But to argue that foundations should pivot away from education, one must give them an alternative cause first. Gates is frankly the foundation about which I’m least concerned in this regard. Its record on global health is not spotless, but it’s far better than the foundation’s track record on US education. If you look at a promising cause in global health — wiping out malaria with the help of CRISPR, fighting smoking in low-income countries, trying to develop a universal flu vaccine — odds are that at least some Gates money is behind it.

But there are promising causes outside Gates’s scope too, that both his foundation and Bezos and Chan-Zuckerberg could pursue instead of giving deeper to education. To name just a couple:

  • As Flint, Michigan, has reminded us, lead poisoning continues to impose massive social costs in the US. One cost-benefit analysis suggested that $1 spent on lead abatement produces between $17 and $221 in social benefits. That’s just one study, but the overall evidence base for the intervention looks much firmer than most education interventions. I’ve found one lead-related study that Gates funded, but other than that, Gates, Bezos, and CZI are mostly absent on the issue.
  • Macroeconomic policymaking at the Federal Reserve influences trillions of dollars in economic activity globally but faces much less lobbying than policymaking in Congress, due in part to the frankly naive conceit that the Fed is “independent” from outside politics. As a result, there’s relatively little foundation money being spent ensuring that the agency that holds the fate of the global economy in its hands is making good decisions that prevent recessions and lower unemployment. It seems almost certain that the marginal lobbying dollar will do more good there than in education politics.
  • And then there are causes that have had next to no coverage. Suicides with pesticides are very common in the developing world and kill more than 100,000 people annually; banning certain highly lethal pesticides, as Sri Lanka has done, appears to dramatically cut suicide rates. I know of one highly effective group working on this problem, but it only works in a handful of countries, implying there could be room for more funders.

I’m pretty confident that those causes are more promising than K-12 education, but not 100 percent sure. I don’t have a large staff to do comprehensive cause selection research. Thankfully, billionaires’ foundations do! And given how difficult it’s proven to find good interventions in K-12 education, I strongly suspect those foundations can find more than enough projects elsewhere with a higher anticipated impact. They should start trying to do that.

[“source=businessinsider”]