Marcin Balcerzak/

Rule #1: Policy Design Starts With Operational Capacity

What happens when policy wonks make implementation someone else's problem.

Sir Jonathan Ive, the head of design for Apple, obviously thinks seriously and analytically about design. In February, in a long profile of Ive in The New Yorker, Ian Parker observed that Ive thought “bad industrial design often starts in ignorance of what a material can and cannot do.”

Clearly there are things that an industrial material—be it steel, silicone, or Styrofoam—can do. But there are also things that each of these materials cannot do. To be effective, a designer, as Ive emphasizes, needs to understand such differences.

The same might be said of the challenges of designing any public policy: “Bad policy design often starts in ignorance of what an organization can and cannot do.”

Too often, the policy wonks, legislative analysts and elected officials who design policy are ignorant of what an organization can and—even more importantly—cannot do. Before public officials ask an organization to do something, they need to ask themselves two key design questions:

(1) Can the organizations to which we plan to assign this responsibility do this?

(2) What additional operational capacity do these organizations need so they can actually do what we ask?

After all, as the policy is initially designed, there might be no organization that can do what is asked.

Certainly the designers of public policy have a diversity of alternatives. Lester Salamon of Johns Hopkins University has emphasized: In addition to direct government, there exist a variety of “tools” with which to accomplish public purposes. Government can contract out, regulate the private sector (for either economic or social purposes), purchase services from nongovernmental organizations, and offer tax breaks, grants, or loans.

Will the assigned organizations, however, have the operational capacity to achieve the purpose behind any of these designs? Even if government is creating incentives for others to implement the policy, it is relying on people—be they in a corporation or a family—to have the competences necessary to convert the policy into a results-producing reality.

Maybe, despite a policy’s explicit incentives, the people in the organization—be it large or small—are simply uninterested. After all, the required tasks might not mesh with what Morton Halperin, now of the Open Society Foundation, once called their “organizational essence.” This is “the view held by the dominant group in the organization of what the missions and capabilities should be.” If not, the organization won’t really try.

Or maybe these organizations are not capable of doing what is required. They may simply not have people with the required operational capacity.

Or maybe the policy is a product of “disjointed incrementalism” with different features of the policy added sequentially by different designers. The disjointedness of such “incrementalism” might take place over years, as different policy designers contribute their own pet accessory. (Think Apple does it this way?)

Or maybe the policy was designed in such a way that no normal human organization could implement it successfully. It might be so rigidly specified to conform to some cherished theory or to satisfy a key stakeholder, that the assigned organizations will be unable to experiment, learn, adapt and then learn some more.

Or maybe the definition of success was left so vague—even contradictory—that it is impossible to motivate individuals and teams to focus on accomplishing anything specific.

Or maybe the organizational arrangements are so complex—designed perhaps to give a number of favored organizations their own desired role—that too many people have the ability to veto any activity.

Or maybe the policy design diffused responsibility for success so broadly that collaboration is impossible.

Or maybe an overhead agency has the ability to overrule details that are necessary for implementation.

Or maybe no available organization has the necessary operational capacity. Nevertheless, the policy design does not include the resources necessary to create these competences.

Those who design public policies are rarely the ones charged with implementing it. The designers know how to maneuver their ideas through the policy-making sausage machine. They understand how to negotiate the inevitable compromises. After that, implementation is someone else’s problem.

Yet, general ignorance of what organizations can and cannot do is often the cause of bad policy. And specific ignorance of what the assigned organization can and cannot do is simply a specific case of the same rule.

Only if policy designers develop the knowledge about what different organizations can and cannot do, can they wisely design any policy.

“Bad administration can destroy good policy,” observed Adali Stevenson, “but good administration can never save bad policy.”

Robert D. Behn, a lecturer at Harvard University's John F. Kennedy School of Government, chairs the executive education program Driving Government Performance: Leadership Strategies that Produce Results. His book, The PerformanceStat Potential, was recently published by Brookings. (Copyright 2015 Robert D. Behn)

(Image via ArtcoMarcin Balcerzak/