Bureaucracy at the boundary.

• Author: O'Connell, Anne Joseph
• Position: Nonexecutive administrative organizations creation and role in governance - II. Theories of Boundary Organizations A. Why Boundary Organizations Are Created 3. Putting Political Control and Competence Together through Conclusion, with footnotes, p. 884-927

3. Putting Political Control and Competence Together

   Even with just two dimensions (political control and agency competence) and two designers (the President and Congress), decisions regarding agency structure can become quite complicated. This subsection seeks to explain when these designers will choose to create boundary organizations. Figure 1 roughly illustrates the placement of a variety of entities, including boundary organizations, along these dimensions. The political control dimension, represented on the vertical axis, captures presidential control at the top and congressional control at the bottom. Entities toward the center face little congressional or presidential control, or can place both overseers in conflict to gain independence. (224) The competence dimension, represented on the horizontal axis, moves from structures more peripheral to the federal government or Executive Branch on the left to more central structures on the right.

   Some spaces in this two-dimensional model do not house organizations. Most notably, the Constitution and other bodies of law like the APA preclude some locations, and the following Section takes up legal constraints in more detail.

   By contrast, other spaces hold many organizations. Mixed public-private entities crowd along the horizontal axis, for example. The placements are relative. A line of public entities is at the right, in which Cabinet departments face the most presidential control. The courts and organizations at the boundary of the Legislative and Executive Branches face little or opposing amounts of presidential and congressional control. Congressional support agencies, such as the Congressional Budget Office (CBO), and entities at the border of the Legislative and Judicial Branches face more congressional control than presidential control. The preferences of the President and Congress are not marked; I consider them more systematically below.

   In certain contexts, the agency designers face tradeoffs between their preferences on political control and agency competence. Specifically, if the benefits to political control exceed the costs, Republican Presidents and Congresses confront a conflict between their desire for political control and their desire for private organizations. Democratic Presidents and Congresses, however, do not face such a conflict, as many assume they prefer public entities. Conversely, if the costs to political control outweigh the benefits, Democratic Presidents and Congresses face a choice between their desire to avoid control and their desire for public organizations. Republican Presidents and Congresses do not confront a conflict in this circumstance, as most people assume they prefer private entities. Those who are economically inclined can think of the tradeoff as a two-good consumer utility model. If the consumer is Republican, the vertical axis measures political control, and the horizontal axis measures the extent of privatization (so the origin is set at roughly no control and no privatization). The opposing party functions as the budget constraint, allowing only so much control and privatization.

   In addition to the connection between the two dimensions for a specific actor, one must consider how the preferences of the President and Congress interact in order to understand which organizational structures are chosen. Consider five cases: (l) a divided government, in which benefits to political control exceed costs for both actors; (2) a divided government, in which costs to political control dominate benefits for both actors; (3) a divided government, in which benefits to political control exceed costs only for Democrats; (4) a unified government, in which benefits to political control exceed costs for both actors; and (5) a unified government, in which costs to political control dominate benefits for both actors. Relevant party variations are considered in each case.

   In the first case, both Republicans and Democrats see the benefits to political control as exceeding the costs. Because of the divided government, the actors here are at odds--a Republican President and Democratic Congress (or a Democratic President and Republican Congress) will disagree on political control and agency competence. The outcome depends on bargaining strength but likely rests in the middle area of both dimensions. (225)

   In the second case, both Republicans and Democrats view the costs to political control as exceeding the benefits. The actors are less at odds than in the first case, but some conflict remains. A Republican President and Democratic Congress (or the reverse) have more similar views on political control but still clash on agency competence. The result is similar to the first case, probably with the agency somewhere in the middle of both of the two metrics. The difference is that the parties want to relinquish political control (as opposed to engaging in a zero-sum game).

   In the third case, Republicans see the costs of political control as exceeding the benefits, but Democrats view the benefits of political control as exceeding the costs. Neither party, therefore, faces any tradeoff on its own preferences concerning political control and agency competence. In the case of a Democratic President and Republican Congress, both favor presidential control but must compromise on agency competence. Thus, the organizations in the upper two quadrants of the figure (as determined by the figure, not the page) represent possible outcomes. With a Republican President and Democratic Congress, both institutions favor congressional control and must compromise on agency competence. The organizations in the lower two quadrants of the figure are possible outcomes.

   In the fourth case, as with the first, both Republicans and Democrats see the benefits to political control as exceeding the costs. Unlike the first case, however, the same party controls both the White House and Congress. Under Democratic control, one can expect a classic executive agency structure. Under Republican control, the desire for privatized structures tempers the desire for political control, likely resulting in boundary entities (perhaps with more presidential appointments).

   In the fifth case, as with the second, both Republicans and Democrats view the costs to political control as exceeding the benefits. Unlike the second case, however, in the fifth case the same party controls both the White House and Congress. Under Republican control, one anticipates an agency structure with considerable connections to the private sector. Under Democratic control, the preference for public structures and lack of control may yield boundary organizations within the federal government.

   In short, this analysis predicts that Congress and the White House will often establish boundary entities to perform tasks. Boundary organizations are therefore not marginal at all. These predictions could be tested by looking at changes in agency structures--which would keep agency functions relatively constant, in line with the model's assumption of a set task.

   2. The Consequences of Boundary Organizations

   The previous Section illustrated why politicians may want to establish boundary organizations. This Section briefly evaluates whether these entities are desirable in terms of social welfare and democratic legitimacy. Commentators often presume, explicitly or implicitly, some tradeoff between these two goals. (226) Just as boundary organizations balance the competence and responsiveness concerns of political actors, they also balance these normative objectives. Assuming these objectives are worthwhile and achievable, that balance will be optimal in a number of contexts, and suboptimal in others.

1. Social Welfare

   A social welfare planner cares about the public's economic interest. Compared to other entities, boundary organizations, particularly those at the public-private border, may promote that interest for a variety of reasons. First, such entities may be more flexible--both in terms of policy mechanisms and objectives--because of market pressure or other factors. (227) One reason may be the relative freedom that boundary entities enjoy from restrictions that other federal agencies face. (228) Alternatively, or in combination, flexibility may result from market or other pressures to innovate. Second, border agencies may be more insulated from the special interests of public or private actors (on either side of the boundary), (229) which often diverge from the public's interest. Thus, they may be able to function outside of arrangements that are dysfunctional. Third, if there are externalities or informational asymmetries in the market, boundary organizations may produce more efficient outcomes than purely private actors. (230) Finally, boundary organizations may have more expertise. (231) This expertise may derive from higher employee salaries, especially for entities unconstrained by federal hiring rules. (232)

   The social welfare story of boundary organizations is easy to tell. Indeed, it is the story told about many of these entities, particularly public-private ones. (233) But it may not be true in certain contexts, such as when organizations get stuck in particular modes of acting. Purely public or other central actors may be more entrepreneurial, boundary organizations may lack the necessary authority to implement their decisions, (234) or organizations may simply find it easier to operate on one side of the border than on both sides. Similarly, purely private entities may have even more flexibility than quasi-private ones. (235) Another context is when power vacuums that insulate border agencies produce opportunities for nonpublic behavior by such entities--either because they have special interests of their own or because other institutions are better able to capture them. (236) Semiprivate entities, for instance, may face little market oversight. (237) Market failure offers yet another...