Why Government Fails So Often: And How It Can Do Better (37 page)

CHAPTER 8

Implementation

T
he public policy world brims with interesting, provocative, often plausible ideas for improving social welfare. These ideas emerge from executive branch policy shops, congressional staff work, independent think tanks, lobbyists, economics and political science departments, public policy programs, law schools, business organizations, “public interest” groups, and other sources. Even judges sometimes get into policy-making-in-the-large when they find constitutional violations by government bureaucracies and then fashion unusual remedies—that is, those that go beyond simply awarding monetary damages or a simple prohibitory injunction—with the goal of rectifying those violations.
¹
A very small percentage of these ideas successfully run the marathon political gauntlet and make it into the United States Code.

At this point, an even greater challenge looms for these new laws: they must be implemented in the real world outside Washington, D.C. The field consists of complex social, political, fiscal, market, legal, intergovernmental, bureaucratic, and institutional conditions that even the most seasoned policy makers in Washington cannot fully anticipate. Worse, even if the policy makers could foresee these conditions, they would have relatively little control over them.

Political scientists have spilled a lot of ink describing the recurrent kinds of implementation obstacles that impede a large number of federal programs. In the policy studies field, the leading study of this phenomenon is
Implementation
, a celebrated book by Jeffrey Pressman
and Aaron Wildavsky, whose droll subtitle summarizes what they found:
How Great Expectations in Washington Are Dashed in Oakland; Or, Why It’s Amazing That Federal Programs Work at All, This Being a Saga of the Economic Development Administration as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation of Ruined Hopes
.
²
The book recounts in great detail the process by which, beginning in 1966, a grant from the Economic Development Administration (EDA) in the U.S. Department of Commerce of more than $23 million for public works projects in the port and city of Oakland, California, was implemented. The grant aimed to create infrastructure and provide three thousand new jobs to inner-city minority residents of the high-unemployment Oakland area. It was greeted with the great fanfare and high hopes that accompany such initiatives. Three years later, only forty-three jobs had been created as a result of the government’s $1.08 million in business loans, and the Oakland antipoverty agency was deriding it as a “pretty big disaster.”
*

In chronicling this particular policy disappointment in Oakland, Pressman and Wildavsky did not merely recount what happened at each stage of this program’s development as EDA officials sought to implement it at the local level. As distinguished political scientists deeply rooted in both theory and empirics, they had a larger objective: to draw from their detailed account of the program’s failed implementation more general lessons about
why
so many promising, well-intended programs suffer the same fate. I now set the stage for these lessons by summarizing their rich, exhaustive chronicle of the Oakland Project.

Having explored the lessons of this case for policy implementation more generally, the chapter then turns to the empirical evidence on how a wide variety of other kinds of policies have been implemented,
and with which effects. All of the policies examined here seek in one way or another to alter how particular markets function, thus extending the analysis in
chapter 7
. I shall discuss these policies under eight general rubrics, differentiated according to what the policies
aim to do to or with markets
: (1) perfect them; (2) supplement them; (3) suppress them; (4) simplify them; (5) subsidize them; (6) redirect them; (7) reintroduce them; and (8) recruit them for regulatory purposes.

THE IMPLEMENTATION OF THE OAKLAND PROJECT

The EDA’s principal mission was to aid the economies of rural areas and small towns, not cities like Oakland. An EDA internal evaluation noted, however, that the agency’s leader was determined to use the agency to set up employment programs “where the action was,” in cities, and he decided to concentrate its efforts on one city, apparently because he had to spend program funds quickly before the appropriation expired, and this could be accomplished by doing it in one city rather than dividing the funds among more of them. Oakland was selected for three main reasons: the EDA had experience and contacts there; it hoped that a jobs program there would head off a feared riot by unemployed blacks; and the city’s Republican mayor could not go over the administrator’s head to complain to the Democratic White House if something went wrong with the program.

The EDA, however, did not fully appreciate the obstacles to policy implementation posed by the nature of the Oakland city government, which would have to make key decisions if the program were to succeed. But the city government was fragmented into a number of municipal agencies that city hall could not control, in part because Oakland had a council-manager form of government in which the elected, part-time mayor was only one of nine city council members, and the full-time appointed city manager effectively ran the administration with staff, information, and financial resources largely under his control. Also, the lack of politically oriented interest groups
and party activity meant that elected leaders could not mobilize relevant information or political support for policies they might wish to propose.

The EDA’s program developed problems not only with its Oakland “partners” in the city government but within its own ranks and with other federal agencies. The administrator assembled a very talented group of young policy experts within the EDA to run the program for him, largely from Washington. Proud of their innovative energies and antibureaucratic values, and pressed for time by their other Washington responsibilities, they took shortcuts by disregarding prescribed bureaucratic channels. This, along with the priority that Oakland was demanding, aroused resentment in EDA headquarters, which had traditionally focused on rural development.

Other mission conflicts soon appeared. Business interests in Oakland wanted EDA funds to go to commercial development, which they predicted would create more jobs. But city officials doubted that the new jobs would go to the long-term and minority unemployed, so EDA officials designed an overall employment plan under which a board dominated by poverty group representatives would allocate program funds. Because the EDA wanted to get the funds committed quickly before the appropriation lapsed (which might cause Congress to reduce future funding), there was little time to conduct an extensive search for the best public works projects. Under these pressures, the EDA selected the Port of Oakland, the institution most adept at preparing federal applications. An EDA official explained, “EDA had to spend its funds by June of 1966. The port had the projects and the others didn’t.” The EDA ranked specific projects by expected job creation, and the funds were allocated to the ranked projects until the available funds ran out. The port, arguing that the EDA’s stringent requirement that each project must recruit, train, and hire the local long-term unemployed would scare off potential developers, resisted the requirement but ultimately accepted it. Key EDA officials, consultants, and city personnel came and went. A new mayor took over in Oakland. The top EDA official and cheerleader for the Oakland projects resigned because, he said, the White House had decided that the
EDA should not be spending money in cities, and Vietnam was claiming more and more resources.

As program problems and danger signals appeared, officials eager to get the funds out tended to sweep the difficulties under the rug, saying that they were just glitches that did not affect developers’ willingness to hire the long-term and minority unemployed. Some of these glitches, however, turned out to be far more consequential. For example, the port claimed that it could not proceed without advanced funding from the EDA for interim development and construction financing, advances that the EDA could not provide. A dispute over the quality of the fill materials on a terminal site arose, and the terminal’s prospective tenant said it could not meet the program’s stringent hiring conditions. The U.S. Navy’s engineers complained that the construction would create both serious navigational hazards and threats to airplane safety, which brought the Federal Aviation Administration (FAA) into the dispute. When the port sought to begin dredging, the Army Corps of Engineers entered the fray, as did the Bay Area Rapid Transit District (BART), which claimed that the dredging would impinge on its operations. When the port requested changes in the plans (a larger restaurant), the EDA said this was major enough to require a new project application. The U.S. General Accounting Office (GAO) questioned the EDA’s grants and loans, and the San Francisco Department of Public Works complained that the projects would interfere with water contact sports in the bay. The authors summarized the situation thus: “As 1968 ended and a new administration prepared to come to power in Washington, the future plans of the terminal lay in doubt. In the areas of dredging, filling, financing, design, and relationships with other governmental institutions, the technical problems surrounding construction of the marine terminal had proved to be formidable. What had appeared initially to be a relatively straightforward program now involved new and unforeseen participants—the navy, the GAO, local governmental bodies—whose agreement was necessary if the program was to continue.”

The other major public works project—the aircraft maintenance hangar—was beset by even greater obstacles: a cost overrun of
almost 50 percent in two years, delays caused by new rounds of bargaining among the various interests, divisions within the EDA about how to respond, and reports indicating that few new jobs were being created.

As the program bogged down, the EDA’s leadership constantly changed, and bureaucratic units within the agency with differing perspectives jockeyed with each other for power. Meanwhile, the political situation in Oakland was changing, “becoming more strident, more polarized, more hostile” as black leaders sought to organize their constituents by attacking the mayor, who looked in vain for program support from the EDA—particularly when the EDA refused to pay for the huge cost overrun. With a new administration installed in Washington, new EDA policy makers coming on board, and both components of the Oakland Project—public works and jobs for the minority unemployed—floundering in delay, disarray, and unrealized promises, the EDA tried tightening the screws on the local players but to little avail.

Pressman and Wildavsky’s verdict on the Oakland Project was a profoundly dismaying one. Five years after Congress had appropriated the funds and the EDA had committed them, projects and employment plans had been approved, but major obstacles to implementation remained. No final plan had been reached for the most important construction project, which was supposed to create twelve hundred jobs. The new small business loans program—targeted at inherently bad risks, replete with conflicts of interest, and competing with existing businesses—failed abjectly to create new jobs. The training programs that were to prepare the hard-core unemployed for jobs had become snarled in intergovernmental and interagency wrangling. And the EDA was having trouble enforcing its employment and affirmative action requirements even for those construction projects that were completed or underway. No one could be sure how many new jobs had been created, nor was it clear how many were filled by minorities and the hard-core unemployed, or how long they would last. (On the brighter side, one relatively small project, the West Oakland
Health Center, seemed to have created 160 jobs, almost all of them for minorities.)

SOME GENERAL LESSONS ABOUT IMPLEMENTATION OBSTACLES TO POLICY SUCCESS

The failure of the Oakland Project implies that other programs are likely to fail for similar reasons, although the specifics of their failure will of course differ from Oakland’s and from each other’s. (See, for example, Derek Bok’s brief account of the failure of the Comprehensive Environmental Response, Compensation, and Liability Act of 1980, commonly known as the Superfund.
³
) This implication is amply warranted because, as Pressman and Wildavsky emphasize, the Oakland Project was actually far easier to implement than most federal programs. The EDA sought “to avoid the institutional fragmentation, multiple and confusing goals, and inadequate funding that had characterized previous federal-city programs.” It included

just one federal agency in one city; there would be only one major local recipient, the Port of Oakland, whose tenants’ performance could be bound by a written and signed employment plan; and there would be an immediate commitment of $23 million…. We are unlikely to get many programs that will be simpler in that they involve fewer participants or less complicated arrangements. Few programs would specify as clearly the benefits to the participants—cheap money, jobs, political credit—than this one. The difficulties that arose here, therefore, can be expected to afflict almost any new program. If we understand why this program ran into difficulty, we can hope to get at some of the underlying factors that make programs fail when there is no apparent reason why they should…. The apparently simple and straightforward is really complex and convoluted [because of] the number of steps involved, the number of participants whose preferences have to be taken into account, the number of separate decisions that are part of what we think of as a single one. Least of all do we appreciate the geometric growth of interdependencies over time where each negotiation involves a number of participants with decisions to make whose implications ramify over time.
⁴