The Journal of Public Policy

How Presidential Appointees (or Lack Thereof) Matter












By William Resh (@billresh), University of Southern California

As Gary Hollibaugh, Jr. and colleagues plainly stated on the LSE blog, “Presidential appointees matter.” Of course, this is of little question when these positions are filled. Incompetent appointees cause deleterious consequences for both citizens and presidents. But, what about those positions reserved for Senate-confirmed presidential appointees (PAS) that remain empty?

A report last year by Pro Publica grouses that the Obama administration has been subject to more vacancies than previous administrations—in both independent and executive branch agencies (see Figure 1). Yet, occupancy is often legislatively required in independent commissions and agencies before agency action can be triggered. Therefore, outcomes (or a lack thereof) are easier to identify during periods of vacancy in those positions (at least anecdotally), whereas the effects vacancies have on agency performance in executive branch agencies generally may be less evident. The substantive importance of filling these positions in executive branch agencies is not empirically established. In terms of functional performance, “the evidence linking appointee continuity and agency performance remains largely anecdotal” (Dull & Roberts 2009, 445).

Source: Government Appointees Project (Dull, Resh, & Roberts, Forthcoming)

There is certainly significant symbolism inherent in the ability of a president to fill appointee vacancies across agencies. Presidential appointees ostensibly offer a democratic dimension to leadership as emissaries of the sole elected office in the U.S. executive branch. Appointments are made for both patronage and policy, and seldom (at the PAS level) are the constructs wholly separate considerations. Presidents use patronage to engender support for their administration in a given policy area and to simultaneously “signal policy” by appointing prominent issue network actors.

In the research in which I am currently engaged, I argue that there are insights from both the traditional public management scholarship and the “bureaucratic control” scholarship emanating from political science that can potentially yield a great deal of insight both theoretically and empirically on the mechanism by which PAS vacancies affect how agencies perform. But first, we should address the relative strengths and weaknesses of both fields as they apply to an area as central to executive politics as presidential appointments.

Whereas management scholars tend to take a given operational goal as the objective starting point for their analyses, they often do not consider the inherently subjective value underlying that goal. But, as Herbert Simon (1997) puts it, goals are the “value premises that can serve as inputs to decisions” (152). The course of action chosen by a given organizational actor is a choice of values that are formulated by motives based on personal, organizational, resource, political, and environmental constraints. In the context of the U.S. executive branch, agencies are often delegated the responsibility to align goals that are premised upon disparate values representative of an assortment of constituencies, institutions, and attentive publics by which the agency is held accountable.

Thus, theories of bureaucratic politics from political science reject the value-neutral, technocratic conceptualization of public administrators in favor of bureaucrats as routinely allocating values. At the same time, much of this scholarship focuses on the congressional dominance of the bureaucracy, generates the presumption that the line between in-role and extra-role behavior is fixedly demarcated, and mainly makes “implied statements about bureaucratic policymaking… telling us little about executive action and, hence, about what bureaucrats actually do with the policy discretion they are granted by the legislature and the resulting policy outcomes” (Krause 2010, 534-535).

When introducing presidents (or executives) to this relationship, we can easily surmise that they are not so likely to generally leave the legislature’s delegation of value allocation to the career bureaucracy. In terms of the various, wide-ranging goals (re: value premises) that Congress has delegated to bureaucratic agents, it is likely that this value set is reflective of the inherent diversity of values that emanate from Congress’ pluralistic institution. The president, meanwhile, is likely to emphasize a more narrow range of values than Congress as a whole due to his institutional perspective.

As reflected in Figure 2, the president will balance the existing power structure of a given agency using an array of strategic tools—including appointments that are often founded on the level of ideological congruence and personal loyalty a nominee has to the president—to more precisely define and prioritize the agency’s value premises. The diversity or range of values represented in an organization is potentially narrowed by the president’s efforts toward executive cohesiveness and coordination.

Thus, one might posit that, as PAS initially enter the organization (PAS Placement), they will seek to narrow the value orientation of the agency according to their preferences. As their tenure increases, the value diversity of the organization will decrease further (PAS Continuity). As lower-level, unilateral appointments (PA) are made to the managerial and lower executive ranks of the organization (i.e., “politicization”), this too will further hone the value orientation of the agency. At the same time, this may also potentiate the inability to achieve comprehensive notions of performance in a pluralist/separation-of-powers system that promotes value diversity through a range of policy and institutional prerogatives.

In a current study, I examine the potential impacts that vacancies in presidentially-appointed positions have on the performance of federal agencies. I test how conditions of appointed leadership vacancies affect the extent to which an agency moves toward a state of organizational “goal entropy”—i.e., the value diversity of its goal environment. I argue that the range of values reflected in an agency’s set of stated operational goals can be conceptualized as the discretionary range delegated to an administrative agency and that this range can be measured empirically and is a function of appointed leadership continuity or the lack thereof. I also test whether the value diversity of an organization’s goal environment helps determine its ability to meet any given goal.

Combining two data sources—a detailed longitudinal roster of presidential appointment positions and the officials who occupy them, as well as U.S. agency Performance & Accountability Reports(PAR) (which provide strategic goals, specific annual performance targets based on the agencies’ strategic plans, and show whether the actual performance results meet the targets)  with administrative personnel records and budget data—I employ a multilevel time series logistic regression technique to model the temporal impacts of appointed leadership vacancies on goal setting, how appointees shape the value diversity of the organizations’ respective goal environments, and the likelihood that an organization meets any given goal as a function of these conditions. In other words, I devise a method of measuring and testing the extent to which executive attempts at political control supplement and narrow the legislatively delegated value structure of a given agency and thereby reveal the mechanism that leads to limited agency performance.

I also measure the persistence of PAS leadership vacancies over time. Under conditions of persistent vacancies, one might expect that it will be more difficult for any given organizational objective to be met. However, as goals are set ex ante for implementation and performance, one might also expect that the value diversity of the goal environment—i.e., the heterogeneity of values represented in an agency’s set of annual goals—will advance the agency’s efforts toward meeting its goals, generally, because the agency has more flexibility in prioritizing and defining the values that align best with its capacity. Following this logic, we should expect that appointed leadership continuity purposively narrows the value diversity of its goal environment, but that value diversity facilitates the ability of the organization to achieve those goals.

I employ a sample of over 7,000 goals drawn from over two dozen agencies that issued PARs between FY2000 and FY2011. Each agency issued a report in at least two fiscal years during this period, and the number of goals articulated by an agency in a given year ranged from 193 to 7. All goals were coded according to their predominant value or values as identified by three individual coders using the construct definitions provided in Table 1. The coders began with a relatively comprehensive list of values derived from a variety of public management research sources that focus on the concepts of public sector values and goal-setting.

While the various scholarly sources we encountered informed our list of predominant values, none of the sources attempted to attribute values directly to specified operational goals of public organizations, instead focusing on organizational mission statements. While values that underlie missions can also be fundamental to operational goals, operational goals tend to be more specific in their value orientations. For instance, the value of efficiency (maximizing output at an optimally minimal rate of input) may be absent as an explicitly stated, generalized value of the organization while simultaneously predominant in any of an organization’s stated operational goals. More elusive concepts such as “integrity” or “justice” tend to be explicitly manifest operationally through values such as “accuracy” or “procedural compliance,” respectively.

The actual values underlying goals are secondary to the main interest of this work. Rather, I am interested in how various human resources—namely politically appointed leadership—and the value diversity of the goal environment in which any given objective is embedded contribute to the likelihood that that objective is met. The four agency-level constructs of interest in this analysis are (1) the persistence of appointed leadership vacancies in the year during which an annual objective was to be met (Vacancy Index), (2) the persistence of appointed leadership vacancies in the year during which an annual objective was set (Vacancy Lag), (3) the layering of unilateral, non-Senate confirmed appointments through managerial and executive ranks (Politicization), and (4) the value diversity of an organization’s goal environment (Goal Entropy).

To capture the construct of “persistent appointee leadership vacancies,” I utilize a detailed longitudinal roster of PAS positions and the officials who occupy them (Dull, Resh, & Roberts 2013). This data collection examines appointments to full-time, civilian PAS positions in all departments, single-headed independent agencies, and Executive Office of the President organizations. Using this data, I derive a Vacancy Index that standardizes vacancies both within and across organizations. I include a vacancy index for the year in which the implementation toward goal attainment took place (V) and an index for the prior year in which the goal was originally set (l.V). The relative persistence of vacancies in presidentially-appointed leadership positions should also be a critical attribute of the institutional environment.

Borrowing from communications theory, I include Shannon’s (1949; 1951) measure of discrete entropy to capture the value diversity of the organization’s goal environment. Entropy is a measure that has been used across various settings in social science research. Here, I use this measure to capture the relative value diversity of an organization’s goal environment—what I call “Goal Entropy”—using discrete observations of the value types derived from the qualitative coding of PARs. Whereas , n is the number of discrete value premises observed in the qualitative coding process and pi are discrete probabilities that any value premise will be present within a given organization’s goal environment.  is a continuous and an increasing function of n. Its value ranges from 0 to log n( ~  max). As tends to log n, the value diversity of the organizational environment increases and becomes more heterogeneous (tending toward a diversity of value types being predominant across goals).

I begin by modeling Goal Entropy ( as a function of the Vacancy Index (V) and its lag, politicization (P), and other organizational-level variables using ordinary least squares (OLS). The varying intercept of the subsequent multi-level logistic regression is then modeled as a function of the predicted value of Goal Entropy (Ω) while controlling for potentially intervening organizational-level variables. Thus, . Whereas  is used as a key contextual-level characteristic in a multilevel logistic regression model with the binary outcome of whether a given stated organizational goal is met (Pr(METij=1|βj) = ϕij; log[ϕij/(1 – ϕij)] = ηij):

Level-1 Model

ηij = β0j +  Z*(Λ)

Level-2 Model

β0j = γ00 + γ01*(Ω j) + γ02*(Vj) + γ03* (Pj) + W*(ϒ) + u0j

To simplify explanation for the purposes of a blog post, Figure 3 provides a heuristic that helps clarify these conceptual relationships and how they bear out in the analyses. Highlighted variables are statistically significant, and corresponding signs indicate the direction of that relationship. To summarize, the findings support a theory of causal goal entropy. The predicted value of Goal Entropy (Ω) directly enhances the likelihood of goal attainment. And, this value is derived as a function of various human capital resources—namely presidentially-appointed leadership continuity (or conversely, vacancies). However, I find no support for the notion that vacancies preceding implementation (l.V) or during implementation (V) have any direct effect on goal attainment. Rather, vacancies have an indirect effect on goal attainment through the mechanism of goal entropy. In other words, vacancies that precede implementation (l.V)—i.e., as goals are set—increase the value heterogeneity of the goal environment, which in turn increases the likelihood of goal attainment. Moreover, politicization has a similar indirect effect on goal attainment by negatively impacting the value heterogeneity of the goal environment.

Gawthrop (1984, 6) contends that the “preeminent purpose of public management” is to accomplish the “integration and convergence of social values” in a pluralist system. To do so requires neither neutral nor responsive competence; it is a matter of providing institutional competence—the innate understanding of an agency’s history, programs, norms, constituents, and central organizational actors and how these characteristics can be employed toward specific actions. To merely be responsive to the entreaties of presidents and their staff risks not only the ethical dilemmas that concern Gawthrop, but also short-term political gains that may come at the expense of more holistic notions of effectiveness. In other words, the purported representative function of political appointments does not automatically ensure that the organization recognizes and adapts to its external environment while simultaneously achieving performance.

It’s not the experiential or technical competence of appointed leadership, per se, that might drive performance. Rather, it may be their propensity toward imposing a narrow vision of the organization. This is not to say that managerial or technical competences do not matter—because they certainly do, and they are very likely related to the relative breadth of organizational value perspectives. It is to say, however, that appreciation for the multiple value perspectives accommodated by the administrative state is important to performance. Appointees with parochial policy and value perspectives—regardless of experience or expertise—are likely to fail the organization overall.


Dull, M., Resh, W. G., & Roberts, P. S. (2013). Who isn’t running American government: Appointee vacancies in U.S. executive agencies, 1989-2009. Paper presented at the Midwest Political Science Association, Chicago, IL.

Dull, M., Resh, W. G., & Roberts, P. S. (Forthcoming). Government Appointees Project.

Dull, M., & Roberts, P. S. (2009). Continuity, Competence, and the Succession of Senate-Confirmed Agency Appointees, 1989-2009. Presidential Studies Quarterly, 39(3), 432-453. doi: 10.2307/41427373

Gawthrop, L. C. (1984). Civis, civitas, and civilitas: A new focus for the year 2000. Public Administration Review, 44(Special Issue: Citizenship and Public Administration), 101-111.

Krause, G. A. (2010). Legislative Delegation of Authority to Bureaucratic Agencies. In R. F. Durant (Ed.), The Oxford Handbook of American Bureaucracy (pp. 521-544). Oxford, UK: Oxford University Press.

Shannon, C. E. (1949). Communication in the presence of noise. Proceedings of the IRE, 37(1), 10-21.

Shannon, C. E. (1951). Prediction and entropy of printed English. Bell system technical journal, 30(1), 50-64.

Simon, H. A. (1997). Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations (4th ed.). New York: The Free Press.