What Can the Federal Government Do Well?
By Jal Mehta
Last week I was in D.C. for a conference hosted by Rick Hess and AEI that aimed to derive lessons from 50 years of federal government involvement in schooling. The adjective in the title is "sobering" (sobering lessons), and the tone of the conference was fairly bleak. Speaker after speaker, from the left as well as the right, talked about the inability of the federal government to generate on the ground improvement in schooling. You can read an overview and links to the papers here.
As a contrarian, I started to note the examples that contradicted the thesis, and found that the speakers had actually offered a fairly lengthy list of things the federal government could do well amidst their overall pessimism. Here's a scorecard:
1) Keep states and districts from doing awful things (IDEA)
2) Provide political cover for people lower in the system who want to do things but are afraid of political backlash
3) Fund R and D (we could be much smarter about it, but it is a function of the feds)
4) Use the bully pulpit
5) Fund discrete things directly (magnet schools were cited)
6) Directly fund CMOs and other external actors who might be able to help provide better schooling
7) Create common data and force some degree of transparency
Feds might be able to:
1) Use incentives to get states and districts to do valuable things, especially things they were marginally inclined to do anyway.
2) Help build an infrastructure of NGOs that could help schools do things better
1) Create quality at the level of schools/force schools to get better
2) Make people do things they don't want to do well (i.e at a high level of quality)
3) Foster innovation (maybe they can, but they haven't done much yet).
Four observations about this list. The first is that what we think of as what government does -- issuing lots of regulations that seek to force lower powers to do their bidding -- is just one of many tools at the government's disposal. That particular mechanism works really badly at producing quality, but the government has a number of other arrows in its quiver.
The second is that the bully pulpit is really important. With only 10% of the overall funding, the federal government has been able to drive the conversation--first around high poverty students in the 1960s, then through A Nation at Risk (a commission sponsored by the DOE), and most recently through NCLB.
Third, there may be some things where there are just not good answers. One of the most interesting debates at the conference was between equity liberals like Kati Haycock who were defending the spirit of NCLB on the grounds that if the federal government didn't stand up for the most disadvantaged students, who would? On the other side were a number of us who were arguing that if we were moving from floors towards ceilings, regulatory mechanisms weren't going to get us there. But Kati's point was also right -- she was arguing that if we left it to states or schools, there were some states and schools that have just terrible records of treating poor and minority kids well. I find both sides of this argument persuasive -- leaving well enough alone or incentivizing the leaders means that you won't reach some kids who need it most; but trying to use regulatory mechanisms to "order" all of the nation's schools is also not likely to yield the results that you want. Perhaps the answer is more capacity building, or better mechanisms of selecting or training teachers, or turnarounds for the most disadvantaged schools, or some other policy fix, or perhaps this is just a tension with no really good answers.
Fourth, we're only one country in one period of time. What we have and haven't done well to date is not a measure of what we could or could not do well in the future. More on that tomorrow.
In any case, to justify the above list would require much more thorough documentation; think of it as a provocation to start the conversation. If the question is comparative advantage, what do you think the federal government is particularly positioned to do well? And, just as importantly, what does it not do well?
Jal Mehta is an Assistant Professor at the Harvard Graduate School of Education.