Vigilance Required in Governmental Laboratories

Donald F. Kettl
Donald Kettl
Donald F. Kettl Donald F. Kettl is a Nonresident Senior Fellow at the Brookings Institution as well as Professor Emeritus and former Dean of the University of Maryland School of Public Policy

October 12, 1997

From welfare to health reform, national policy makers are relying on the cherished old saw: the states are America’s “laboratories of democracy.” For better or worse, the states will be where the action is for at least the next decade.

The “laboratories of democracy” idea holds that the states—smaller, more lively, more varied, closer to the people than the national government—have the ability and drive to cook up sharp, new ideas for making government work better. A strong consensus has emerged that the national government too often has failed and that the states can do a better job.

My wife and I, though, have recently returned from Edinburgh, Scotland, where we stayed just around the corner from “The Jekyll and Hyde Pub.” The pub’s name was a reminder of an important lesson about laboratories: Well-meaning experiments don’t always turn out well, and experimental failures can sometimes be downright terrifying.

As the states move into their own laboratories, how can we figure out whether they are producing kindly Dr. Jekylls or scary Mr. Hydes? The debate will hinge on four questions.

First, when problems occur what is their cause? It would be unrealistic to expect all states to tackle these tough problems flawlessly. After all, the federal government has not—and no new experiment works perfectly the first time or every time.

But as the states work through welfare reform, they’re likely to find it impossible to place all welfare recipients in productive jobs and keep them there. Some welfare recipients will have inadequate skills, health problems or transportation problems and will, therefore, not be able to get and keep jobs. When the inevitable problems arise, will they stem from impossible goals managed remarkably well? Or from reasonable goals managed poorly? How will we able to tell the difference?

Second, if there are problems, how do we track down the culprits and make sure the problems don’t recur? All of the really interesting programs rely on a long chain of management. Welfare reform, for example, operates through a federal program devolved to the states, which often in turn pass responsibility to local (mostly county) governments, which in turn are contracting with for-profit and not-for-profit organizations that will actually deliver the services.

Such contracting systems make eminent sense. The alternative, after all, is building big new state and local bureaucracies to manage these programs, and that is precisely what these initiatives seek to avoid. But these long chains are very complicated. Where will the problems lie? With poor performance by contractors, or poor oversight by government managers two steps removed from operations? And even more important, how will we be able to tell?

Third, just what results should we reasonably expect? It’s probably not reasonable to expect all welfare recipients to find jobs. But what is reasonable? Should we shoot for 98 percent? Or 95 percent? Or 80 percent?

Or in land use planning, a subject that has bounced with great volatility between state and local governments for generations, how much agreement can we really expect on decisions that deal with competing uses—and where decisions tend to spill over local government borders and affect neighbors profoundly?

The problem is that we so often convince ourselves of the inescapable wisdom of our policy decisions that we are unprepared for dealing with the inevitable failures. As a result, we overreact in devising new solutions.

Finally, since the newly devolved programs are taking state governments boldly where no government has gone before, neither past experience nor analytical smarts can guide us through these tough questions. In the old days, ideology would come riding to the rescue. If life became complicated, being a liberal or conservative, a Republican or a Democrat would invariably provide clear answers to complex puzzles.

But there are two complications in falling back on ideology today. One is that devolution pushes past the answers that ideology can provide. The most outstanding of today’s governors and mayors are supreme pragmatists, not ideologues. They focus on solving problems, not living by sacred tracts.

The states and local governments are hard at work on some of the most important and most complicated problems of our time. They are bound to produce some huge successes as well as some stumbles, if not outright failures.

Any good lab will try to figure out what works, what doesn’t and why. The four forces—the tension between management and policy, the long chain of government management, the problem of defining success, and the decline of ideology—provide a guide for separating some states’ Dr. Jekyll from others’ Mr. Hyde.