David Brooks has a new column making the case for pre-K investment, and I agree with much of the substance — particularly his emphasis on the importance of decentralization and a test-and-learn approach. Below I’ve highlighted a few areas of agreement and disagreement that I’ll discuss in reverse order:
Obama is trying to significantly increase the number of kids with access to early education. The White House will come up with a dedicated revenue stream that will fund early education projects without adding to the deficit. These federal dollars will be used to match state spending, giving states, many of whom want to move aggressively, further incentive to expand and create programs.
But Washington’s main role will be to measure outcomes, not determine the way states design their operations. Washington will insist that states establish good assessment tools. They will insist that pre-K efforts align with the K-12 system. But beyond that, states will have a lot of latitude.
Should early education centers be integrated with K-12 school buildings or not? Should the early childhood teachers be unionized or certified? Obama officials say they want to leave those sorts of questions up to state experimentation. “I’m just about building quality,” Education Secretary Arne Duncan told me. The goal is to make the federal oversight as simple as possible.
That’s crucial. There’s still a lot we don’t know about how to educate children that young. The essential thing is to build systems that can measure progress, learn and adapt to local circumstances. Over time, many children will migrate from Head Start into state programs. [Emphasis added]
(1) “The essential thing is to build systems that can measure progress, learn and adapt to local circumstances” — this is one of Jim Manzi’s core principles, and I believe that it should be integrated into every institution that delivers taxpayer-financed public services, whether public or private. The idea is not to fund big randomized controlled trials and then roll out the lessons in centralized fashion. Rather, it is to constantly conduct small-scale trials designed to assess all kinds of small, incremental tweaks, recognizing that a result that works well in one environment might not work well in another.
(2) While I appreciate Secretary Duncan’s apparent indifference to whether early childhood teachers are unionized or certified, I’m skeptical that the federal Department of Education will be “indifferent” in practice, as evidenced by the outsized role given to “teacher buy-in” in the Race to the Top (RTT) initiative. Moreover, there is a relationship between funding levels and unionization.
(3) If it were true that Washington’s main role were to measure outcomes, Washington would have to be less prescriptive. Demanding that states establish good assessment tools makes sense, as it will promote transparency and the diffusion of knowledge. Yet establishing that pre-K programs will align with the K-12 system is potentially very prescriptive.
(4) And the biggest problem, as we’ve discussed before, is the structure of the federal match, which creates an incentive for states to spend more, not more efficiently. Say we determine that a high-quality, cost-effective preschool program can be funded for $9,450 per pupil. The federal government, under the CAP proposal, will match spending up to $10,000. So if a state spends $4,725 of its own money, it will receive a matching federal grant of $4,725. But if it spend an additional $5,275, it would receive a matching federal grant of $10,000. So why bother finding the most cost-effective intervention? Indeed, state officials who want to implement the most cost-effective intervention will face intense lobbying efforts from pre-K providers who want to maximize the federal matching grant, as it will mean the difference between $20,000 in spending and $9,450. This pot of money will attract intense interest from public employee unions, which will insist that early childhood teachers be certified and unionized, lest uncertified, non-union teachers pose a danger to young children. Given the number of children involved, there will be scare stories that will occur in pre-K facilities, whether unionized or not.
So what is the alternative to a federal matching grant? The federal government could exclusively pursue the RTT strategy: states that construct good data systems and roll out innovative pre-K programs will be eligible for temporary federal funding. One big difference with RTT is that states would have to have programs up and running for, say, two years before they’d become eligible. This is far more stringent, yet it would force states to demonstrate that these pre-K programs are valuable enough that they’d be willing to fund them independently. Moreover, states would have to show not just that they can cook up elaborate plans to submit to the federal Department of Education, but that they can actually implement real-world projects.
The big reason this modified RTT-style approach is unattractive is that it would be harder for a sitting president to capture the political dividends, and there would be a real risk that the next president would abandon the initiative. That is a fair point. But it would help ensure that the program doesn’t lead to overspending while still encouraging the diffusion of knowledge.
UPDATE: More details are being released about the president’s proposal, and I will have a follow-up shortly.