Political Targeting in Democracies

Abstract

Democratic erosion—the undermining of republican government by a leader with authoritarian tendencies—depends on the improper use of the state apparatus of the state against opponents (“political targeting”). Because political targeting sometimes falls into a legally gray area, and because officials have some maneuvering room in how to respond to the orders they receive, their preferences matter. In the United States, officials’ behavior seems to be most influenced by a) the professional risks of refusing improper orders, b) normative obligations to uphold the rule of law and to act ethically, and c) attitudes toward the leader. These factors are, in turn, largely a function of 1) how officials are selected and 2) the extent of oversight and procedural checks they face. These findings have potentially broad implications for democratic erosion.

Harassment of political enemies by the state requires subordinate officials to execute orders that are improper and often outright illegal. In this Reflection, I discuss when officials in an established democracy will engage in such behavior (“political targeting”). I thus aim to provide the micro-foundations for a theory of democratic erosion.

Obviously, political targeting does not exhaust the list of unconstitutional or improper behaviors in which an elected leader might engage. Other such actions include obstruction of investigations into wrongdoing by the leader; protection of allies and accomplices from regulatory or law enforcement action; preferential treatment of constituencies based on their perceived loyalty to the incumbent; exploitation of legitimate authorities for corrupt purposes, such as abusing pardon power (Pfiffner and Florence 2019); encouragement of supporters to disrupt constitutional processes through mob action; enlistment of foreign allies, state or local governments, and the private sector to punish adversaries; and the use of privately financed units to conduct dirty tricks. However, directives that single out adversaries for harassment by agents of the state arguably constitute some of the most troubling orders from the perspective of democratic continuity.

The next section discusses the existence of legally gray areas and bureaucratic discretion, which make officials’ motives relevant to political targeting. The third section describes the context in which officials operate in one country: the United States. The fourth section speculates about the factors that influence officials confronting improper orders, based on discussions with officials in the national security and law enforcement apparatus. Aside from idiosyncratic factors, three things appear to loom largest in officials’ willingness to participate in political targeting: a) the professional costs of defying an order from superiors, b) normative commitments to legalism and integrity, and c) attitudes toward the president. These factors are, in turn, shaped by how officials are chosen and monitored.

I focus empirically on the United States because it is an important country in its own right and because the fate of democracy there is likely to influence politics in other countries. Furthermore, there is unprecedented debate among political scientists about the fragility of American democracy (inter alia Bernhard and O’Neill 2020, Lieberman et al. 2019). Unfortunately, this debate is sometimes based on international comparisons (see inter alia Kaufman and Haggard 2019) that are unlikely to produce accurate point predictions. A nuanced understanding of how the American system actually operates in practice is crucial to assessing the potential for democratic “backsliding” (Bermeo 2016) in general and political targeting in specific.

Notwithstanding this empirical focus on the United States, my theoretical ambition is broader. The final section thus returns to a more general discussion of political targeting. The lessons of the U.S. case, I argue, have little to do with the sort of large-scale institutional reforms often discussed by political scientists (see inter alia Howell and Moe 2021, Adserà and Boix 2007), nor even with the (valuable) debate over constitutional “bright lines” (Carey et al. 2019). Rather, they concern a) the thoroughness of internal governmental procedures that provide guidance to bureaucrats and b) the selection processes for officials in positions most likely to be implicated in targeting opponents of the incumbent.

The Agency of Bureaucrats

In the real world, bureaucrats in established democracies are rarely confronted with clear-cut instructions to commit flagrant misdeeds. Rather, problematic orders are often couched in such a way as to give them a veneer of propriety, to diffuse legal accountability, and to create a permissive environment for wrongdoing. Because agents of the state must obey legitimate directives, what they are supposed to do when confronted with such ambiguous situations is not obvious.

Even if the impropriety of a directive is clear, the seriousness of this impropriety may vary. Some actions entail only minor violations of agency policy, whereas others constitute more significant transgressions, and still others involve criminal acts (themselves of varying degrees of seriousness). These two dimensions (ambiguity and seriousness) are shown in figure 1. The most egregious actions—those that are undeniably criminal—fall into the upper right quadrant.

Figure 1 Political targeting: examples of specific actions by agents of the state

As an example, a law enforcement or intelligence officer might be instructed to “look into” a political opponent of the government (see the top pane of figure 1). If this official breaks into the opponent’s house in the middle of the night and steals damaging information with which to blackmail her—the top-right corner of the upper pane in figure 1—he has committed multiple criminal offenses. If he instead works to drum up criminal charges against her based on flimsy evidence (see the bottom right corner of the upper pane in figure 1), his violation is also extremely serious but less clear-cut, given the extent of discretion that investigators typically have. By contrast, if our investigator merely adopts an insulting tone with the opponent, his actions will be subject to interpretation and—even if judged to be improper—will constitute only a minor violation of policy (see the bottom left corner of the top pane of figure 1). If while questioning her he unsnaps the strap on his holster as a prelude to drawing the firearm—“breaking leather”—he has certainly committed an infraction, but that breach is less serious than a black bag job. A range of other actions fall elsewhere along the dimensions of ambiguity and seriousness, such as checking the ownership of a vehicle parked in her driveway without legal predicate, pulling over her car for a minor violation that might not normally merit a traffic stop, tailing her in an official vehicle, and so forth.

The bottom pane of figure 1 focuses on harassment of a political opponent as she returns to the country from abroad. Government search powers tend to be very broad at international borders, and customs or immigration officers typically enjoy considerable discretion in deciding whether to search travelers or their belongings. An official could readily engage in low-level political targeting while credibly claiming to have acted in good faith. The least serious and most ambiguous violation would be for the customs officer to drag out the process of checking the target’s documents, thereby delaying her slightly and perhaps signaling that the state has its eye on her. More egregiously, an official might conduct a more intrusive physical search than is warranted; claim that an item found in the opponent’s possession (e.g., a pack of cigarettes) is contraband and then hold her up while the status of that item is verified; or even take her in custody without proper grounds and keep her confined beyond the allowable period (i.e., unlawful arrest and imprisonment).

The impropriety of all these actions varies to some extent from one democracy to another, as well as within particular countries over time. Legislation and procedural requirements (e.g., requiring sign-offs for certain actions or prohibiting specific activities) can thus reduce the ambiguity that agents of the state face. For instance, extralegal black bag operations for intelligence purposes were standard practice in the United States during the 1930s–1970s and explicitly or tacitly approved by at least five presidents (Weiner 2012; Kessler 2008; Jeffreys-Jones 2007); however, such activities would be thoroughly illegal today.

Nevertheless, even the strictest rules and most intensive oversight inevitably permit bureaucrats some flexibility. For instance, law enforcement officers and prosecutors must be allowed some discretion in drawing on their contextual knowledge to determine whether certain actions are suspicious and in allocating finite law enforcement resources. If officials are sympathetic to the incumbent leader, they may zealously pursue directives to target political opponents—perhaps even exceeding what was asked of them. If officials dislike an order, however, they may find ways to resist. For instance, they can insist that the order be transmitted through proper channels and in writing; ask their superior for clarification on the order and its rationale; consult with lawyers at their agency to determine its legality; “forget” that the order was given; arrange to be preoccupied with other business rather than execute it; delay its execution until it is moot; interpret it in such a way that what is actually carried out is legal; threaten to resign rather than carry it out; report what they were asked to do to an oversight body; etc. Indeed, bureaucrats may be quite adept in the art of slow-rolling orders of which they disapprove. Thus, an investigator told by his boss to “come down hard” on a political opponent may argue that he has complied with the order if he merely interrogates her extensively, and customs officials instructed to “let her know she is not welcome” could finesse the order by simply making a few extra inquiries (which is well within their discretion). In this context, officials must often draw their own lines, according to their own lights. Both their own views and the flexibility they have to act on them influence the likelihood that they will participate in political targeting.

An Example: The American System

Determining how these factors operate requires of a more granular understanding of security agencies than appears in most analyses of democratic backsliding. For instance, characterizing the United States as a presidential system with a bicameral legislature, an independent judiciary, and a professional civil service presents an accurate but a radically incomplete depiction of the landscape that officials inhabit. Selection procedures for officials in internal security posts and oversight mechanisms (including internal checks that limit bureaucratic discretion) are essential to understanding the potential for political targeting.

Selection of Officials

Simplifying considerably, officials in the civilian side of the U.S. executive branch come in three main flavors: a) senior political appointees in the bureaucracy, b) White House staff, and c) career civil servants. They operate in different contexts, with implications for their likely willingness to engage in harassment of political opponents.

Political appointees occupy the top several rungs in almost all of the federal bureaucracy. These individuals serve at the pleasure of the president and are normally appointed on the basis of loyalty to his party and to him. In total, the bureaucracy includes approximately 9,000 Plum Book positions (Plum Book 2021), comprising both senior and junior policy-related political appointments, and 4,000 presidentially appointed positions (the great bulk of which are listed in the Plum Book). The tenure of these appointees is short; on average they hold a given post for less than two years (Lewis 2011; Maranto 2005; Dull and Roberts 2009; Tenpas 2020).

Several characteristics of senior political appointees are relevant for their likely willingness to engage in political targeting. First, they tend to have successful careers outside of government. Although many political appointees expect to earn more money after leaving public service—based on the increased prominence, insight, and connections that a stint in a senior federal position brings—most are also attracted to the inherent appeal of those posts. Second, those who end up at the Assistant Secretary level or above will typically bounce in and out of government multiple times in their careers, returning when their party is out of power to a career in law, the private sector, or the “quasi-academic sphere” (Lawson 2014). Because successful Washington careers span multiple administrations, involvement in legally dubious activity would not only make it difficult for them to secure a future post but could also render them “radioactive” (as one former official put it) to the white-shoe law firms, political consultancies, and government contractors that commonly hire former officials.

Although political and partisan vetting is a crucial part of the selection process for senior political appointees, the larger “filtration process” (Mukunda 2012) tends to weed out individuals with an uncertain commitment to the Constitution, especially in security posts. To begin with, political appointees in national security roles have often had lengthy careers in non-partisan positions, such as the military, foreign service, or intelligence community. Both selection into these careers and the professional experiences they bring strongly incline security officials to patriotism and constitutionalism. Fidelity to black-letter law and court rulings has likewise been heavily inculcated in those with a legal background, which most individuals in the Department of Justice have (see Weiner 2012, chapters 38-39; Kessler 2008).

Senate confirmation is another important element of the filtration process for political appointees in the bureaucracy, and constitutional provisions tell only part of the story. In practice, Senatorial courtesy plays a role in filling some of the positions in the Justice Department that might be useful for targeting political opponents, such as most of the ninety-three U.S. Attorneys (USAs). (“Senatorial Courtesy” 2014). Furthermore, commitment to the rule of law figures prominently in Senate vetting of nominees for most security posts, based on collective memories of J. Edgar Hoover’s directorship of the Federal Bureau of Investigation (FBI), violations of the Central Intelligence Agency (CIA)’s charter into the early 1970s, and Watergate. Bipartisanship is common when approving these nominees.

The situation for the second set of officials—political appointees in the White House, including the West Wing—is rather different (see Bose and Rudalevige 2020). First, White House staff tend to be chosen for enthusiastic support of the president’s agenda. In many cases, they have been through the intensely partisan experience of a presidential election campaign together. Second, very few members of the White House staff are subject to Senate confirmation. Third, White House staffers operate in a context where tremendous attention is focused on pleasing the president, and professional advancement involves demonstrating fidelity to him and his political agenda. Finally, many senior staffers interact directly with the president, meaning that he can exploit personal loyalty and the trappings of office to influence them. Even when White House appointees have similar backgrounds to political appointees in the bureaucracy, the environment in which they work is very different. Collectively, these factors make it likely that political targeting would originate with or be transmitted to the bureaucracy by a politically reliable senior staffer in the White House.

By far the largest category of U.S. officials are career civil servants, hired and promoted through a highly formalized, ostensibly merit-based system, rather than through political patronage. They remain in government even when administrations change; they cannot be dismissed at will by political superiors; and they are not allowed to engage in partisan activities on the job, with restrictions being tighter for officials in security and law enforcement agencies (see Office of Special Counsel 2018). The bulk of civil servants are on the General Schedule (GS) or its equivalent at their agency, with a much smaller number in the elite Senior Executive Service.

Most of the positions central to political targeting are reserved for civil servants, almost always at the Senior Executive Service level. (Section A of the online appendix discusses the main posts.) Appointments into these positions are typically made by a cabinet secretary or agency head, typically in consultation with the White House and the chairs of the relevant congressional committees. In practice, seniority also plays a major role in some agencies, as at the FBI and the Secret Service. In addition to these posts, the highest-ranking civil servant in an agency is called upon to “act” in the position above him whenever a political appointee is not in place. Given the sluggish and contentious nature of the appointments process (Golden 2002), this situation occurs with some frequency. Who will be appointed to an acting position normally follows a pre-set order that is not subject to presidential manipulation.

All told, the posts most relevant for punishment of political adversaries comprise a small fraction of the federal bureaucracy—no more than a few hundred senior positions, of which almost all are reserved for civil servants or require Senate confirmation. (Refer to section A of the online appendix for a detailed discussion.) Of these posts, around two dozen would be crucial to any concerted effort at political targeting by the administration; all of them are either Senate-confirmed (e.g., the Attorney General, or AG) or reserved for civil servants (e.g., the Deputy Director of the FBI).

Oversight

One crucial element of the American system is legislative scrutiny of Executive Branch operations (Aberbach 2002; Balla 2000). Even given a “unilateral presidency” (Waterman 2009; Durant and Resh 2009), Congress frequently writes specific procedural requirements into authorizing legislation, appropriation bills, and other laws (inter alia Shipan 2005; MacDonald 2010; Gailmard 2002). It has its own bureaucracy with which to investigate agencies (the Government Accountability Office); it claims the authority to subpoena executive branch officials; and bureaucrats are accustomed to meeting with congressional staff, briefing Members, and testifying on Capitol Hill. Because it is rare for both chambers and the presidency to be controlled by the same party for any length of time, oversight of the administration by political opponents of the incumbent is common (see Kriner and Schickler 2016). Oversight is far from perfect, especially when it comes to lower-level Plum Book posts and positions on the White House staff, and legislative appetite for active surveillance of the national security bureaucracy is low. Nevertheless, senior civil servants and political appointees in the bureaucracy are well aware of pressures from Congress (Furlong 1998).

A related oversight mechanism, which affects both career civil servants and politically appointed officials in the bureaucracy, is the post-Watergate system of Inspectors General (IGs). IGs now number over seventy and usually have their own law enforcement agents (see Council of the Inspectors General 2014). Although IGs are appointed through the usual process of Senate confirmation and report through the Executive Branch chain of command, in practice most remain in place when administrations change. In addition, Congress has imposed a number of special reporting requirements on IGs aimed at encouraging whistleblowing by officials and informing Congress of potential malfeasance. ¹

Internal procedures that limit bureaucratic discretion also tend to be well developed in the United States, with legislation being supplemented by executive orders and often painstakingly detailed agency policies (inter alia Kagan 2001). In particular, a dense thicket of laws and rules governs the behavior of security agencies. ² For instance, in response to Executive Order 12333, the Intelligence Community issued “linear feet of regulations” codifying internal processes (Slick 2014, 2). Multiple sign-offs are required for actions that would constitute political targeting.

The longstanding division between intelligence and law enforcement agencies—as designated by statute and reinforced by rules of evidence that normally preclude prosecutors from relying on information that might compromise intelligence sources and methods—effectively constitutes another check on political targeting. Intelligence agencies seeking to harass a political opponent would have to hand off the information they obtained to law enforcement entities, who would in turn have to collaborate with prosecutors, who would then have to deal with an independent judiciary and citizen-juries. Of note, most other democracies lack these checks. For instance, few European countries draw a bright line between intelligence and law enforcement or guarantee jury trials in all criminal cases, much less civil actions brought by the state.

In addition to Congressional oversight and internal checks within the bureaucracy, judicial review of bureaucratic operations is comparatively well developed in the United States (Whittington 2009). This role for the judiciary can have a potent effect on the willingness of all types of officials to engage in improper activity. As one senior intelligence officer put it: “Nothing concentrates the mind and dampens excess so wonderfully as the imminent prospect of explaining one’s actions to a Federal judge” (Magnet 1996, 50).

Beyond these formal oversight mechanisms, the “media regime” (Lawson 1999) shapes the context in which American officials operate. Traditions of investigative journalism are well established in the United States and leaking to the press by officials in the Executive Branch is common (especially by officials in the West Wing of the White House, the Justice Department, and the FBI). The media environment both increases the odds that improper orders will be exposed and that other players (such as election-seeking legislators or civil society groups) will be motivated to take action in response.

One final aspect of the American system that shapes officials’ operating environment is the fixed presidential term and term limit. The fact that administrations will change every four to eight years means that a civil servant’s decision to align herself too closely with the incumbent will only bring benefits for a time, after which she will have to pay the piper.

Again, oversight and control mechanisms do not apply in the same way to different types of officials. For instance, political appointees in the White House usually cannot be compelled to testify before Congress. These facts again suggest that White House staff are most likely to be involved in political targeting.

Political Targeting in Action

To illustrate how political targeting might operate in practice, consider four hypothetical scenarios: a) initiation of a criminal investigation against an opponent without sufficient predicate, b) harassment of an opponent returning to the country from abroad, c) illegal electronic surveillance of a citizen through “reverse targeting,” and d) unlawful search of a journalist’s cellphone. (Refer to section B of the online appendix for a discussion of just how hypothetical these scenarios really are.) The ultimate outcome of these scenarios obviously depends on the behavior of a number of actors; the objective here is simply to show how orders might travel within the executive branch and identify the sorts of officials who would be involved.

In the case of a criminal investigation into a political opponent, the easiest path would probably be for the AG (acting on White House instructions) to reach out directly to one of the ninety-three regionally based USAs, bypassing the Deputy AG to minimize the number of people involved. (Bypassing the Deputy AG is atypical but not improper.) The USA in question would then attempt to enlist a receptive career prosecutor from within his office and a member of a federal investigative agency who could line up law enforcement resources—presumably the relevant Special Agent in Charge of an FBI field but potentially an analogous official at another federal investigative agency or, within a large FBI field office, a squad leader.

In the case of harassment at a port of entry, two routes are most plausible. First, the directive might come from the president or White House chief of staff to the secretary of Homeland Security (perhaps orally, during a meeting on an unrelated subject), and thence to the Senate-confirmed commissioner of U.S. Customs and Border Protection (CBP). Second, a senior person on the White House staff might contact the commissioner directly. Because the normal chain of command would then involve numerous steps within the civil service, involving more people, the commissioner might reach out directly to the head of the relevant field office—say, Miami International Airport. The head of the field office could then give orders directly to front-line officers, again bypassing some officials. Although a few of the steps in the chain of command could legitimately be skipped, far too many are in this case for the order to be fully proper. (See the left side of figure 2, in which solid lines indicate legitimate chains of command and dotted lines indicate a likely path for improper orders.) As one former official noted, “there is almost never a good reason” for a phone call from the White House to the Commissioner of CBP requesting a specific action.

Figure 2 Possible pathways for political targeting orders

If the president wished to surveil a domestic political opponent electronically (the third scenario), one option would be to order the FBI to do so. However, following indictments of FBI officials for illegal wiretapping in the 1970s, such an overt approach is difficult to imagine; savvy political operators would seek a work-around. Through “reverse targeting,” which is illegal but hard to prove, the government collects information on an American citizen “incidentally” by legally eavesdropping on a foreigner (who enjoys far fewer privacy protections) with whom the citizen has contact. Electronic surveillance of a foreigner residing in the United States who is suspected of espionage requires a warrant from the Foreign Intelligence Surveillance Court, as requested by lawyers from the Department of Justice based on derogatory information uncovered by investigators in the Counterintelligence Division of the FBI’s National Security Branch or other parts of the intelligence community. The simplest way to obtain a warrant would be for someone in the White House to ask a trusted agent within the intelligence community to find (or fabricate) derogatory information about the foreigner in question. However, the warrant would still need to be signed off by the relevant Special Agent in Charge, before receiving a sign-off from FBI headquarters and then being delivered to career lawyers in the Justice Department—all of whom are supposed to vet the information in it (Rangappa 2017). To move on to the true target (i.e., the political opponent), this process would have to begin anew, with greater scrutiny and more sign-offs, including from the FBI’s director himself. Although the Foreign Intelligence Surveillance Court almost always grants requested warrants, the process ensures that multiple officials would need to be made complicit before a warrant could be requested.

In the final case, the White House seeks to learn whether a political opponent inside the government has been leaking damaging information to the press. The specific plan is to exploit emergency national security authorities to surreptitiously access the list of contacts on a journalist’s cellphone, on the legally tenuous basis that his sources are somehow connected to terrorism or that the journalist has been party to leaks of classified information. The idea for this operation might begin with a call from a senior White House aide to the (Senate-confirmed) Under Secretary of Intelligence and Analysis at the Department of Homeland Security, thus bypassing the Department’s leadership. Thence the Under Secretary might go directly to a civil servant (e.g., a GS-13) responsible for breaking the phone’s encryption and downloading the contact information, skipping several levels along the way. (See the right side of figure 2, in which solid lines indicate legitimate chains of command and dotted lines indicate a likely path of improper orders.) Given the technical issues involved, however, it is likely that other civil servants would also need to be involved.

Some of these cases are obviously less challenging than others for an authoritarian-minded incumbent, and other acts of political targeting might be easier to carry out. However, in each case several officials must be involved—typically one or more people in the White House, one or more political appointees in the bureaucracy, and several civil servants. The involvement of many people increases the chance that an order will be rebuffed, reported to oversight bodies, or leaked to the media. Efforts by those issuing an improper order to cut people out of a process that was deliberately designed to include them is inherently suspicious; directives that lack the right sign-offs and do not come through the normal chain of command are prima facie improper. Again, most democracies will have such internal checks within the bureaucracy, even though the extent of oversight and the types of individuals involved in the process vary.

What (American) Bureaucrats Consider

Systematic testing of which factors might lead officials to participate in political targeting, even in one country, is beyond the scope of this Reflection. However, informal conversations with several dozen current and former officials, as well as a half-dozen structured interviews in which various scenarios were presented (refer to section C of the online appendix), offer some hints about the sorts of considerations that influence bureaucrats’ decision-making.

As might be expected from actual humans, rather than computers or philosophers, informants rarely approached the issue of political targeting with a specific set of well-formulated considerations in mind. Their starting point was sometimes a highly personalized take on the situation: the idiosyncrasies of the office in which they were then working, their relationship with the person issuing an improper order, etc. As often happens in ethnographic work, such responses must be distilled and presented in a way that is both systematic and recognizable to the informants themselves.

All types of officials—White House staff, political appointees, and civil servants—expressed significant concern about the professional costs of crossing a superior when given a direct order, even if the order was improper. Career civil servants fretted primarily about alienating senior civil servants within their own agency, though they also sought to avoid irritating political appointees in their chain of command and (to a lesser degree) the organizationally more distant White House. Senior political appointees were somewhat less concerned about retaliation from their politically appointed superiors, mainly because they had their own political connections. Although Senate-confirmed appointees deeply desired good relations with the White House, they were also aware that they could be scapegoated to protect the president if an investigation into potential wrongdoing began.

To the extent that officials saw a positive benefit of complying with an improper directive, it took the form of goodwill. Goodwill might help them accomplish other goals, including a promotion that resulted in a) greater status (relevant for all three types of officials), b) a sense of forward progress in their career (mainly relevant for civil servants), and c) greater influence over policy (mainly relevant for political appointees). However, for all types of officials, the downside of refusing to execute an improper order seemed to weigh far more heavily in their thinking than any potential upside of currying favor.

Officials systematically expressed an intense normative aversion toward improper activity. All invoked their obligation to uphold the rule of law—both in the general sense of being a law-abiding person and in the specific sense of executing their official duties. Almost all mentioned (without prompting) the oath that officials swear to uphold the Constitution.

Closely related to these considerations was officials’ desire to behave ethically (see Zacka 2017). The more egregious the act, the more such ethical considerations figured in officials’ thinking. Almost all informants identified lines they would not cross, with actual criminality being a common redline no matter what the countervailing circumstances. Legalism also appeared to serve as a mental refuge for many officials when competing considerations became difficult to weigh, as they frequently were. As a result, “sticking to procedure” or “following the law” emerged as the default course of action, especially for civil servants.

Normative considerations aside, the prospect of punishment in a system with considerable oversight also militated against complicity in improper activities. For all three types of officials, the risk of becoming the target of a lengthy, costly, reputation-destroying investigation was deeply unpalatable. The possibility of prosecution for acts that involved criminal offenses—even if slight—constituted an even more powerful deterrent. Three former political appointees specifically referenced Nixon Administration aides who went to prison as a result of the Watergate scandal, and one senior civil servant noted that his obligation to his family would preclude doing anything that could lead to ignominy or incarceration.

One final set of considerations that some officials mentioned—usually after prompting—were ideological and political. Positive attitudes toward the administration disposed political appointees to push the limits of the system, presume that questionable orders were legal, and punish those who unfairly attacked the administration. At the other end of the attitudinal spectrum, several officials who served in the Trump administration expressed disgust with the president’s foreign policy and what they perceived as his casual attitude toward the rule of law; those sentiments left them disinclined to do anything improper on his behalf. That unique situation reflected the fact that few of the original senior appointees in the national security apparatus felt any attachment to President Trump or his policy agenda (Woodward 2020, chapter 34; Rucker and Leonnig 2020; Taylor 2019).

Attitudes toward the target of an improper order appeared to matter far less than attitudes toward the incumbent. However, some officials might be more willing to push the boundaries if they were convinced an opponent of the Administration was corrupt or dangerous. For instance, one senior civil servant cited animosity toward Hillary Clinton within the FBI’s New York Field Office, whereas a former senior appointee during the George W. Bush administration referenced hostility toward candidate Donald Trump in the upper ranks of FBI headquarters.

Given these conflicting pressures, most officials preferred neither to execute improper orders nor to alienate superiors by confronting them directly. Rather, they expressed a desire to deflect requests, elide directives, and find ways to seek clarity about those that seemed (as one former official put it) “sketchy”. Whenever an order came from outside the normal chain of command, the first response of civil servants was to check with their immediate boss. Requesting an opinion from the general counsel’s office in their agency was another common reaction for both civil servants and political appointees. Civil servants also suggested that they might share misgivings about a problematic order with other officials, “call around” to get a better sense of what motivated it, or seek advice from mentors. Political appointees in both the White House and the bureaucracy were more disposed than civil servants to “give the Administration the benefit of the doubt” (in the words of one former senior official) when considering whether a request was lawful. This stance followed naturally from their partisan allegiances. However, it also reflected the reality that, in the American system, the precise limits of presidential authority are continuously being renegotiated. Of the three groups of officials, White House staff seemed most resigned to the notion that they might be drawn into something improper.

Officials expressed strong reluctance about availing themselves of whistleblower protections, talking to the Hill without superiors’ approval, or leaking to the press. In general, these were tools to be employed only when superiors repeatedly pushed them into a corner. Officials also regarded resignations and (for civil servants) requests for transfers as extreme measures, even where lucrative out-markets for their services existed. A number of political appointees also expressed the concern that their resignation could inadvertently lead to the advancement of less scrupulous individuals—something noted in published accounts by former Trump Administration officials (e.g., Taylor 2019).

Closet Brownshirts, True Believers, and Filtration Failures

In the American system, then, there is considerable resistance to political targeting, especially acts that are clearly illegal. However, interviews and conversations with officials also suggest which sorts of individuals might be most amenable to participation in such dubious activities.

One susceptible category comprises, in the words of one former senior political appointee in the homeland security enterprise, “closet Brownshirts”—career bureaucrats who strongly sympathize with the incumbent and simultaneously have little normative commitment to the rule of law. Two interviewees noted that such individuals exist in any organization, including major federal law enforcement agencies, with the “I’m 10-15” Facebook page scandal (Thompson 2019) being a signal example. The longer an administration lasts, the more the closet Brownshirts can be identified and maneuvered into place.

Among political appointees, the most likely allies of an autocratic-minded president are officials so committed to the leader that they would willingly participate in (or even spontaneously initiate) dubious activity. These often include younger officials dazzled by the office of the presidency, who may be more easily intimidated by the leader or insufficiently attuned to the long-term risks of committing an illegal act. The perfect political appointee for an authoritarian incumbent, then, is a young official who is also a true believer. One senior civil servant pointed to “MAGA child soldiers” in the homeland security community during the second half of the Trump administration as an exemplar of this group.

Because dangerous individuals would normally be detected by the selection process for high-level security posts in the bureaucracy, they are more likely to be found in the White House. However, just as closet Brownshirts are not necessarily screened out by the civil service system, true believers may slip through a system designed to filter them out of appointive security posts in the bureaucracy—especially when the Senate is controlled by the president’s party. Filtration could also break down if a senior official suffered some sort of psychological or cognitive deterioration after having passed through the selection process (e.g., see Kitfield 2016; McCullough 1996, 84; Gries 1996, 93). These “filtration failures” would effectively permit an administration to place loyalists in positions that could be used for political targeting. Again, the number of true believers and filtration failures who reach positions of power is likely to increase the longer an administration remains in office.

Conclusions

Whether agents of the state will participate in political targeting is central to the endurance of democratic institutions. Because officials are often confronted with ambiguous situations and can respond with varying degrees of reluctance or enthusiasm to improper orders, their beliefs and values matter.

When faced with an order to target a political opponent of the incumbent, officials frequently consider the professional consequences of their actions, normative obligations to the rule of law, and their attitudes toward the incumbent. The first of these factors is strongly influenced by the thoroughness of oversight and the specificity of rules governing internal bureaucratic procedures. The other two (constitutionalism and affinity for the incumbent) are largely a function of the process by which they are chosen for their position—from professional socialization to vetting at various stages of appointment.

Figure 3 notionally depicts these two dimensions. In the United States, the posts most likely to be involved in political targeting tend to be subject to considerable oversight (including internal procedural checks) and constraints on the president’s appointment powers. However, different types of officials—White House staff, political appointees in the bureaucracy, and civil servants—operate in different contexts.

Figure 3 Likely potential for political targeting in the United States

Where might other countries fall? Many parliamentary democracies are likely to reside near the bottom-right quadrant of figure 3. In the United Kingdom and Germany, for instance, civil servants occupy posts that would normally be filled by presidential appointment in the United States. In these same systems, however, parliamentary scrutiny of security agencies is notoriously cursory and toothless.

That said, where in figure 3 specific countries belong requires a much more fine-grained analysis of oversight and appointment processes than high-level variables like “presidentialism” or “parliamentarism” (inter alia Cheibub 2006; Adserá and Boix 2007) would provide. The relationships between politicians, political appointees, and civil servants differ across democracies. There may be more or fewer internal procedural checks on bureaucrats and more or less oversight from independent bodies to ensure that such checks actually operate. The absence of IGs (or some similar institution), whistleblower protections, legislative oversight, judicial review, and investigative journalism would remove some of the devices that officials have for resisting illegal orders in the United States.

The rules and practices governing appointments to positions that could be used in political targeting also differ from one democracy to another. For instance, informal deference to the chief executive on appointments (regardless of formal constitutional rules) could allow the incumbent to place true believers and closet Brownshirts in key security positions.

From Political Targeting to Erosion

In established democracies, political targeting is always improper and frequently illegal. Each individual signing off on or directly supervising the execution of an improper order is in a position to rebuff or expose it. Political targeting thus requires trust-based relationships within the government—knowing what sort of orders can be given to which individuals. In other words, “buyers” and “sellers” of improper orders must create a network for such transactions to take place (Kranton and Minehart 2001). For instance, a political appointee close to the incumbent might identify and cultivate allies in the civil service, who could in turn pass improper orders on to sympathetic lower-level officials.

For episodes of improper activity to aggregate into democratic erosion, many officials would need to be complicit. Therefore, consistent political targeting relies on identifying a number of allies within the civil service and promoting them, while at the same time maneuvering true believers into key appointed posts. This process of finding confederates and building the network necessary to carry out activity of increasing impropriety takes time (as the term “erosion” implies). One crucial bulwark against erosion, therefore, is a term limit for chief executives.

Given the potential for filtration failures, many established democracies could eventually be undermined by a long-serving leader who maneuvered loyalists into security roles. But in systems with term limits, constrained appointment powers, well-established legislative oversight, and formalized bureaucratic procedures, leaders will face a serious challenge in undermining the system. This analysis by no means suggests that American democracy—or democracy in any other country—is secure. It simply means that breakdown is less likely to occur as a result of the dynamics discussed here than as a result of other events, such as a contested presidential election in the context of political polarization (see, inter alia, Svolik 2019).

The United States is arguably unique in the extent to which checks designed to prevent political targeting have been built into the post-Watergate system. Procedures with respect to both appointments and bureaucratic operations are likely to be less clear in many other democracies. New democracies, as well as established democracies that have not undertaken the sort of reforms that Watergate prompted within the American system, often lack crucial controls. It is on these sorts of institutions and norms that political scientists concerned with democratic erosion may wish to focus their analyses.