Constraints on decision-making authority, which are intended to prevent erroneous interpretation or application of policies, have an unintended consequence of paralyzing the bureaucracy – ensuring very inefficient implementation of virtually all policies. Moreover, the same constraints and their implied message regarding politicians’ priorities have a more important negative consequence – destruction of opportunities for significant learning from program-level experience. Too often, it is forgotten that, although policies carry legal authority, they are merely theories about which actions will best achieve goals for social welfare (Aligica and Boettke 2009: 119). Fundamental problems of public organizations are that outcomes are not subject to direct control and expenditures do not automatically produce intended results. Where this is admitted and, even more rarely, when we have evidence, politicians are far more likely to blame bureaucrats than to question the policy (unless it can be attributed to the opposing party). Unfortunately, most officials who realize the impotence of their efforts react by attempting to manage more vigorously – creating even more barriers to coordination of knowledge and activities that could otherwise improve performance. Decision-making in such an environment revolves around who holds which cards, not which choice will have a positive effect on public outcomes. The debate is not academic; it is political – an ongoing power struggle between the legislative and executive branches of the federal government. They argue about resource allocation, not outcomes, because spending is the key to the realization of every politician’s goals, whether s/he is an elected or appointed official. To Congress, “flexible” means ceding control to the Executive’s agencies, and they react cynically to pleas that it’s necessary to manage for results – like some voters react to their own claims that spending votes are “for the children.” Agency executives, who have direct access to the data on program operations, fret about inability to control outcomes for many established programs. Members of Congress are understandably skeptical that increasing the scope of executive authority to manage resources will fundamentally transform the interventions. From the legislative perspective, the President has plenty of flexibility already – to organize the agencies and hire talented managers. Why give him the opportunity to redirect funds toward his own favorite constituencies, especially when appropriations supposedly represent passionate negotiation (logrolling) on Capitol Hill? Rivalry between the branches of government is a familiar subject of debate and forms the basis of most economic models that treat government as an agent of the citizen/taxpayer (Mueller 2003: Chs 16-17). These models correctly identify budgets and power as the spoils of continuous war between politicians and bureaucrats. However, dominance models ignore the qualitative differences – competing interpretations of what constitutes the public good – that serve as public justifications for apparent obstinacy. Majority and minority, legislative and executive agents routinely (and hypocritically) accuse their opponents of selfish ambition. It is safe to say that they are all right and wrong to some extent. The battle is clearly worth fighting for agents and the principals they represent. But none of them can credibly claim to represent all the principals and to know
what will work, or not. Yet they expend all resources at their disposal to obtain control over the prize – presumably to do a marginally better job than their rival – reducing the resources and political capital available for implementation of whatever policy and management decisions survive. The game is a typical prisoner’s dilemma, in that agents’ choices to fight rather than cooperate reduce overall payouts. Principal/agent models add fuel to the fire by perpetuating a view of government as an ideological battleground. Ideology has its place in policy formulation, but it can lead to bad government when it interferes with execution. In government, everyone is an agent with unique views and knowledge, all of which are valid, prospectively. At some point, though, action is required – implementation of policies that do not have unanimous support. Agents are responsible for helping us confront the possibility that we are wrong. In the process, we avail ourselves of opportunities to improve matters and to learn from the experience. The system can be used to exercise power, but it can also transmit knowledge; emphasizing instrumental value at the expense of the experimental prolongs factional rivalries that squander resources and paralyze the system. Analysis of political tactics avoids the bigger issue of whether we may be our own worst enemies. Accountability is not a blame game; it is shared responsibility for seeking the truth. It is time to set aside prejudice and commit to two principles of accountability: (1) Policies deserve to be tested and (2) Bureaucrats deserve to be trusted. My Kaleidic Hyperstructure (KH) model demonstrates that relationships, not controls, are the key to effective policy implementation. C, the idealized citizen/ taxpayer, is the only reasonable candidate for principal. But because it is absurd to suggest that a single set of policy preferences would be meaningful, the model represents citizens by special interests (C1, C2, etc.) that align with distinct agencies and programs. In addition to oversimplifying a complex system of relationships, principal/ agent models skirt the knowledge problems by assuming that we know what it is the people want, that it can be provided by a government agency, that we know how to do it and that we will know if it’s done or not. Typically, voters are deemed to elect “responsive” officials who strike optimal political bargains on their behalf and keep unelected bureaucrats in line by tying their hands with laws and public hearings. Control of spending by elected officials is deemed critical to ensure productivity because bureaucrats with freedom to choose will waste resources by working less and pursuing policies at odds with the law. The major flaw in this approach is the implicit freezing of knowledge in authorizations and appropriations. Implementation of the law as enacted is presumed “good.” Unknown details, as well as those admittedly impossible or inconvenient to monitor, are delegated. However, premises regarding cause/effect and resource allocations are treated as conclusions, not testable hypotheses. What emerges from these lines of inquiry are institutional control strategies based on static representations of legislative/executive relationships. Such models fail to provide insight as to how imperfect policies and processes might be improved after being blessed by both Houses of Congress.