The following is the second post in a two-part series by Robin Rogers, associate professor of sociology at Queens College and the Graduate Center at the City University of New York (CUNY). For more about Robin and her first post, click here.
Before I jump into policy experiments, I want to reflect on the enthusiastic response that I received from last week’s Part One of Billionaire Education Policy. If I could summarize the response with one word, it would be relief.
A lot of people who work in education, philanthropy, and government are wary of the rise in billionaire policymaking, but are reticent in voicing their concerns. Perhaps this is fear of retaliation — what Edward Skloot calls the “Brass-Knuckles philanthropy”of the Gates Foundation. But I see another, more heartening piece to this puzzle. People in the philanthropic and advocacy communities don’t want to harm the mission of philanthropy. We fear that revealing the pitfalls of billionaire philanthropy might have some unforeseen effect on the good work that these foundations support.
Billionaire policymaking is the elephant in the room, but nobody seems sure how to approach it. I say that we should name the elephant, but we don’t have to shoot him. There is a middle road.
We’ve named the elephant – it is philanthro-policymaking. It is here to stay. A small, well-networked group of the super-rich will make and fund social policy globally. We don’t have to shoot the elephant, but we need to understand its nature and learn to live happily with it. Like any powerful institution, billionaire philanthropy needs checks and balances. Our task is to develop them.
Now, to education policy. If you’re not a policy wonk, wonkette, or even a wink, as my more politically savvy friends called me in college, stay with me. Once you get past the odd language of experimentation and evaluation, it’s all politics and human folly.
Testing new policy ideas is appealing. Why have a political battle over education reform, when you can experiment with a bunch on a small scale, and then pick the one that works best? In my last post, I mentioned the recent New York Times article “Policy-Making Billionaires” by Nicholas Confessore. In his coverage of Mark Zuckerberg’s controversial 100 million dollar donation to the Newark, NJ school system, Confessore wrote that NJ officials now plan use the money to “experiment” with education policy and find “what works” and then replicate the best programs with public money: “Whatever proves most effective [in the experiments] can then be rolled out on a larger scale.”
This approach to policy reform is not new. It was a central part of welfare reform in the 1990s. Testing and measuring are particlulary attractive to super-wealthy business oriented philanthropists – philanthrocapitalists. Philanthrocapitalist apply business models to philanthropy. They want to measure everything like money.
Social good is harder to measure than money. The official U.S. poverty line was changed this year after years of debate and controversy. We are struggling to even measure poverty. How do we measure student performance? Teacher quality? Our measurements are imprecise at best and meaningless and misleading at worst. Most educators, advocates, researchers, philanthropists, and policymakers are well aware of the problem of measuring complex outcomes. That awareness disappears when we talk about policy experiments. We act as if testing these programs will lead to some empirical, objective truth about what work bests.
Sociologists talk about manifest and latent functions – for all of you Sociologists, I am not suggesting a functionalist approach to education policy, the concept is illustrative. A manifest function is what something is supposed to do. For example, the manifest function of prisons is to incarcerate people. Things also have latent functions – effects that they have in addition to the stated objective. Prisons provide jobs, for example. That is a latent function.
Policy experiments are supposed to tell us empirically how good a program or approach is. They don’t do this very well. Randomized experiments are expensive, difficult, and rare. Most policy “experiments” aren’t really experiments. They are a trial run of a program with data collection. Even then, the data is often collected haphazardly or to highlight program success and minimize failures. Politics and research also operate in different time frames – solid evaluations often take years. In short, well-funded policy evaluations take too long to actually affect policy, and ad hoc evaluations don’t produce reliable findings. If you want to read more about these issues, I recommend Education Research on Trial.
If policy experiments don’t succeed in their manifest function, why are they still around? Because they are brilliant at their latent functions.
1) Building networks of people who support a particular reform and placing many of them in administrative positions.
2) Funding the intellectual development of a new policy.
3) Political advantage. If a program is in place, opponents can’t say the program is radical, impossible or to predict catastrophe — few social programs have immediate and obvious consequences.
4) Taking the debate out of the political realm — what should we do — where citizens play a role and putting them in the technical, “expert” realm — what works.
“Experiments” is not the correct word for this process. The scientific language of experimentation trips us up. Seeding is a more accurate description.
I’m not much of a gardener, but I know that I planted the plants that grow in my backyard, and I know that their success depends on what was planted there before, the quality of the soil, and the weather. Not everything that I plant grows. Some grow for a bit and then wither. Some flowers are hearty but ugly. But none are there because they’re empirically the best possible plants to be growing in my garden.
We need to think of experimental programs as planted seeds rather than clinical experiments. We learn which of the programs that we plant thrive and which fail. We can uproot the plants that are thriving but are poisonous to the plants around them. Rather than talking about outcomes and “yields” in some Sisyphean effort to find the thing that “works best,” we should talk about program results. We should talk about the actual plants, instead of pretending that our “experiments” will one day yield a perfect plant. We should talk about whether a specific goal was met. We should talk about how the goals relate to our values. And we should keep trying to get better measures for the outcomes we care about.
Talking about “policy experiments” as what they really are – seed programs for social policy – would help us see more clearly that billionaire philanthropists have become policy makers. The power of the economic elite currently hides behind the language of science, which seems to legitimize their actions and prevents us from asking questions. If Bill Gates is funding “research” and gathering “evidence” in “experiments” that he is funding, this seems normal. If he’s funding a seed program that will help the government set education policy, the privatization of policy becomes more obvious. We must not allow the language of science to obscure the power of the economic elite. Policy seeding is an effective political strategy