Zombies Ahead: Explaining the Rise of Low-Quality Election Monitoring

The international election monitoring regime has become considerably more complex in the 21st century. Although the number of organizations engaged in high-quality election monitoring has plateaued, the number of low-quality monitors—commonly known as zombie monitors—has continued to grow. Low-quality election monitors threaten democracy because they validate flawed elections and undermine the legitimacy of the international election monitoring regime. This article argues that international politics have played a crucial role in the diffusion of low-quality election monitors. It hypothesizes that ties with autocratic powers that promote low-quality observers and membership in authoritarian regional organizations significantly increase the likelihood that a country will host low-quality monitors at its elections. To test the hypotheses, the article draws on original data on international election observation between 2000 and 2020 that identifies the most comprehensive set of groups of election monitors to date. A statistical analysis of the dataset supports the argument.


Introduction
Over the past two decades, a troubling trend of democratic backsliding has unfolded across the world.States that have long been thought of as solid, well-established democracies have witnessed an erosion of democratic norms and an accompanying rise of domestic corruption and centralization of executive power.While the democratic structures may remain intact from a technical standpoint, in practice, the ability of the judiciary and legislative branches to check the leader has been drastically curtailed.Examples of this abound from Hungary and Poland in Europe to Brazil in South America.Even the United States, whose leaders often boast about the state being a "beacon of democracy," has fallen victim to backsliding in recent years.
Not surprisingly, there is a growing literature that seeks to understand the causes of democratic erosion.While this is certainly an important question, it is equally vital to examine the consequences of democratic reversal given the pivotal role democracy plays in the world today.In the international relations literature, for example, democracy is often argued to be the solution to many of the world's ills.Scholars find that democracies engage in less conflict with each other, experience fewer instances of civil wars, have greater respect for human rights, and have robust economic growth. 1  The decline of democracy, therefore, carries significant implications for global peace and prosperity.In this paper, we examine one part of this decline-the effects of democratic backsliding on international conflict behavior.There are at least three reasons to expect democratic backsliding to affect a leader's international behavior in a negative way.Importantly, all three of these reasons center on executives' goals of further solidifying their domestic power.First, acting as a provocateur against other states reminds the support base of the leader's willingness to take dramatic action to serve their interests.The more bombastic, antagonistic, or outlandish the behavior, the stronger the signal resonates domestically and the more they will give their support.Second, drawing attention to the international realm reinforces the executive's importance relative to the legislature in the eyes of the public as foreign policy is typically the executive's purview.Third, many conflictual interactions between states, especially those that involve actual physical violence, have the potential to create a crisis atmosphere domestically.In addition to opening up more opportunities for executives to tighten their grips (e.g., invoking martial law or using other war powers), it also creates the potential for a rally effect, consolidating their domestic support.
The phenomenon of low-quality election monitoring has recently attracted considerable interest-and alarm. 1 Low-quality monitors, often called shadow or zombie groups, are international observers that are neither willing nor able to report on election fraud when it occurs. 2Christopher Walker and Alexander Cooley describe them as "fake monitoring groups that praise obviously flawed elections in an effort to drown out more critical assessments by established monitoring organizations" (Walker and Cooley, 2013).
Low-quality monitors pose two threats to democracy according to analysts.First, they support the survival of autocrats.By validating flawed elections, low-quality monitors en-courage the public to view those elections as credible and disrupt the opposition's attempts to coordinate around criticism from more reputable international groups.In this way, low-quality monitors short-circuit the process of self-enforcing democracy that high-quality mon-itors (or groups that are both willing and able to report on election fraud) support (Hyde and Marinov, 2014).Second, low-quality monitors harm the international election monitoring regime.The spread of international election monitoring to authoritarian regional organiza-tions (ROs) such as the Shanghai Cooperation Organization (SCO) and Commonwealth of Independent States (CIS) is feared by many practitioners because it calls into question the credibility of election monitors writ large (Merloe, 2015, 86).
Consider, for example, the 2020 parliamentary election in Azerbaijan.Ten election monitoring organizations were present, including both domestic and international monitors as well as both high-and low-quality monitors.The Organization for Security and Cooperation in Europe (OSCE) was one of the high-quality groups.Its 42-page final report came to a negative conclusion, reaffirming the OSCE's preliminary report that stated, "The restrictive legislation and political environment prevented genuine competition in the 9 February 2020early parliamentary elections in Azerbaijan, despite a high number of candidates." 3  Conversely, the 10-page final report of the CIS judged the election to be ". . .competitive, open, free and consistent with the principles of democratic elections." 4 Another low-quality observer group, the Organization for Democracy and Economic Development (abbreviated GUAM), took direct aim at the credibility of the OSCE monitors.Not only did it con-tradict the OSCE's findings by praising the election, but it stated that the election was ". . .organized and conducted in compliance with international obligations and standards of democratic elections, including the obligations and the standards of the OSCE and the Council of Europe." 5   Despite the rise in low-quality monitoring and the challenges it brings to democracy, a full understanding of the phenomenon has been limited by a lack of systematic data.In particular, it is key to study where and why such observers are present at elections in order to identify low-quality observers' effects on democracy and the election monitoring regime.We thus collected new data (described in more detail later) on international observer presence at national elections between 2000 and 2020.We gathered information on what we believe is the most comprehensive set of groups yet: 141 unique election monitoring entities.We identify which groups are low-quality and document a dramatic increase in their presence at the world's elections, shown in Figure 1.Low-quality monitors observed around 23 percent of elections in our dataset in 2000 and 39 percent in 2020.By contrast, the number of elections observed by highquality monitors has remained relatively stable.By the later years in Figure 1, high-and low-quality election monitors were present at around the same number of elections.4 Translated from the original Russian.Available at https://cis.minsk.by/news/13329/zajavlenie_missii_nabljudatelej_ot_sng_po_rezultatam_nabljudenija_za_podgotovkoj_i_provedeniem_vneocherednyh_v yborov_v_milli_medzhlis_azerbajdzhanskoj_respubliki (accessed February 9, 2024).

5
Available at https://guam-organization.org/en/statement-of-the-group-of-observers-of-the-guam-parliamentaryassembly-on-the-extraordinary-elections-of-the-milli-majlis-of-the-republic-of-azerbaijan/ (accessed February 9, 2024).We use this data to shed light on which states host low-quality election monitors at their elections and why.The conventional wisdom is that incumbents invite low-quality observers to increase the likelihood of their political survival.The logic is that lowquality election observers can convince some members of the public, especially lowinformation voters, that an election was clean and the government was legitimately elected, even when incumbents rig elections in their favor.We find some evidence that incumbents in less democratic countries are more likely to host low-quality monitors, but it is fairly limited.For example, this conventional wisdom fails to explain why both South Africa (a democracy) and Zimbabwe (an autocracy) have invited low-quality monitors from the Southern African Development Community (SADC) to their elections. 6Similarly, both Ukraine, a partial democracy, and Uzbekistan, a consolidated authoritarian country, have hosted low-quality observers from the CIS at their elections.
To understand this phenomenon, we argue it is important to consider international dynamics.Two international-level factors play a crucial role in encouraging states to invite low-quality monitors to their elections.First, we hypothesize that ties with powerful autoc-racies that support low-quality election monitors predict the presence of lowquality moni-tors at countries' elections, even in some democratic countries.Powerful authoritarian "en-trepreneurs" increase the supply of low-quality observers available to countries where they have stronger ties.Second, we argue that obligations within authoritarian ROs encourage states to host low-quality monitors at their elections.

6
SADC trains its observers and developed a code of conduct for election observation (see https://www.sadc.int/faqs/what-does-code-conduct-observers-entail,accessed April 24, 2024).However, SADC has not signed the Declaration of Principles for International Election Observers (DOP), which as discussed later is a key document, and is not able or willing to report on fraud due to political sensitivities; SADC member states do not allow SADC to release election observer missions' final reports to the public.

Low−quality
Our analysis supports this international argument.As we will detail, the main country that backs low-quality election monitors is Russia, which has made delegitimizing Western election observers and countering democracy promotion important parts of its foreign policy (Cooley and Nexon, 2020, Ch. 4).In line with our expectations, countries are more likely to host low-quality monitors when they have closer ties to Russia.Membership in authoritarian ROs, which is to say membership in ROs comprised mostly of authoritarian member states, is also associated with hosting low-quality observers.Although Russia is a leader in some authoritarian ROs, there are also groups elsewhere such as the Economic Community of Central African States, which has been an active low-quality election monitor in its region (Cottiero and Haggard, 2023, 12).These findings' primary contribution is to the literature on international election monitoring, where they provide a better understanding of the relatively new phenomenon of low-quality observers.We discuss how our research builds on this literature and deepens our understanding of international election monitoring in the section that follows.Our findings also contribute to several other literatures.The first is the literature on autocra-tization, which is concerned with the strategies autocrats and would-be autocrats use to reduce the public's role in selecting their leaders (Tomini, 2021, 4).Low-quality observers help autocratizing regimes undermine political contestation.We show the significant role that international forces play in the diffusion of this strategy.
In addition, the findings advance scholarly debates about international order.Henry Farrell and Abraham Newman, for example, have shown how illiberal states, including Rus-sia, have taken advantage of contemporary international politics' open information envi-ronment, "convert[ing] openness into a vector of attack" (Farrell and Newman, 2021, 334).Promoting low-quality monitors is part of this wider phenomenon, with revisionist states and authoritarian-led organizations using tactics like information flooding and disinformation to undermine the liberal order (Goddard, 2018, 793;Adler and Drieschova, 2021).
Finally, the new data introduced in this article represents an important contribution in its own right.It will allow researchers to answer questions such as the following: when does inviting low-quality monitors replace or combine with other strategies of election manipulation; when and why did some ROs begin monitoring elections; and what are the effects of low-quality monitors, including on high-quality monitors, incumbent turnover, and domestic monitors.We elaborate on these points in the conclusion.

The Complex Regime of Election Monitoring
Numerous scholars have noted the growing complexity of the international election monitoring regime (e.g., Kelley, 2009b;Pratt, 2018).To understand the phenomenon, we draw on both foundational work on election monitoring, as well as more recent work on zombie monitoring, autocracy promotion, and authoritarian international organizations.
International relations scholars initially analyzed the emergence of a norm of inviting international monitors to countries' elections. 7As Susan Hyde and Judith Kelley showed, international election monitoring increased dramatically during the 1980s and 1990s (Hyde, 2011;Kelley, 2012b).International monitors in the late 20th century often came from Western-led democratic ROs, such as the European Union (EU) and the OSCE, and Western-based non-governmental organizations (NGOs) such as the Carter Center and National Democratic Institute (NDI).This trend was part of a larger growth in democracy promotion (Bush, 2015).States' decisions to invite international monitors at this time are generally understood as reflecting their desire to demonstrate a commitment to democratic norms in the post-Cold War era, given the international benefits associated with democracy.
Most of the groups that contributed to the growth of international election monitoring around the end of the Cold War were high-quality monitors.Groups like those from the EU, OSCE, Carter Center, and NDI were both willing and able to report on election fraud.In other words, they shared a commitment to supporting democracy and, stemming from that, could detect fraud using various methods they honed over time.Publications including the OSCE/Office for Democratic Institutions and Human Rights (ODIHR) Election Observation Handbook (first published in 1996) and the Handbook for European Union Election Observation (first published in 2002) have codified their standards (Hyde, 2011, 169).
Today, high-quality monitors share many characteristics.They send teams that can visit many polling stations and observe the complete electoral cycle, from voter registration and campaigning through the vote count.They rely on statistical techniques such as parallel vote tabulation to check the official results.To some extent, they hold each other accountable; for example, in 2005, numerous groups signed an agreement on norms of election integrity and the role of election observers called the Declaration of Principles for International Election Observers that contains a peer accountability mechanism (Merloe, 2015, 81).
Although even high-quality monitors have limitations and biases (Bjornlund, 2001;Kelley, 2009a;Kavakli and Kuhn, 2020), thanks in part to these methods, international monitors during the late 20th and early 21st centuries were found to support democracy on net.The presence of high-quality international monitors was associated with higher-quality elections and increased turnover (Kelley, 2012b, Ch. 7;Roussias and Ruiz-Rufino, 2018), as well as lower vote shares for incumbents at monitored polling stations (Hyde, 2007).By the 2000s, election observers were also linked to violence prevention by the United Nations (UN) and other peacebuilding organizations (Matanock, 2018;Garber, 2020).
Yet the international election monitoring regime became more complex in the 21st century.There was considerable growth in election monitors that fall short of the standards of high-quality monitors.While we are most interested in the growth of lowquality monitors, there has also been growth in middle-quality election monitors, which are groups that are (1) willing but generally not able to report on election fraud, as is the case with one-time or infrequent election monitors that do not develop professional standards; or (2) able but generally not willing to report on fraud, as may be the case with an RO such as the Economic Community of West African States, which does sometimes criticize elections but usually is publicly lenient toward non-democratic member states. 8Since the groups that fall into this middle-quality category are quite heterogeneous, we do not advance theoretical predictions about the determinants of their presence at elections.Nevertheless, our dataset also allows us to include them in some of our analyses.
The growth of low-quality election monitoring as a distinct phenomenon occurred in the early 2000s.After the Color Revolutions, fear of democratic diffusion motivated autocrats to intensify their opposition to Western-led election observation missions.Sponsoring and inviting low-quality international observers were some of the tactics they used.As Figure 1 suggests, the number of ROs and international NGOs engaged in high-quality election monitoring plateaued, while the number of lower-quality monitors grew substantially.8 8For a graph that replicates Figure 1 but includes middle-quality monitors, see Supporting Information (SI) §1.The SI and replication data are available on the Review of International Organization website.
As defined earlier, low-quality monitors are neither willing nor able to report on election fraud when it occurs.They are therefore the inverse of high-quality monitors.Their lack of willingness and capabilities is often tied to their support of or tolerance for authoritarianism.In contrast to high-quality monitors' large teams and sophisticated methods, low-quality monitors may send missions of just a few individuals who are present in a country only on election day.They may not visit many (or any) polling stations and do not attempt to conduct exit polls or parallel vote tabulations to assess the veracity of election results.And crucially, they issue positive statements about deeply fraudulent elections-sometimes even before polls have closed.Low-quality observers typically either do not produce final reports or do not release those reports to the public.
Two studies to our knowledge have explicitly addressed the question of under what conditions low-quality observers are present at elections. 9First, Judith Kelley used her Dataset on International Election Monitoring (DIEM) to identify less-critical monitors, which she defined as groups that criticize "highly problematic" elections less than half the time.In a bivariate analysis, she found that countries with better political rights according to Free-dom House were less likely to invite only less-critical groups to their elections between 1980 and 2004 (Kelley, 2012b, 55).Second, in a study that primarily sought to understand the conditions under which incumbents invite a combination of high-and low-quality observers, Ursula Daxecker and Gerald Schneider (2014, 86) similarly find that when countries are more democratic, they are less likely to invite one low-quality observer group to their elections.This finding is based on a multivariate statistical analysis and also uses DIEM.Although DIEM included several prominent lowquality organizations such as the CIS, the number of low-quality elec-tion observers has increased dramatically since the end of the data in 2004 (see Figure 1).Thus, it is time to revisit-both theoretically and empirically-the factors that make countries more likely to host low-quality observers.9 The broader literature on low-quality election monitors, on which we draw in the next section to develop our hypotheses, makes related contributions, such as introducing the concept of zombie monitors (Walker and Cooley, 2013) and developing a theory about how autocracies use low-quality monitors to legitimate themselves (Debre and Morgenbesser, 2017, 329).

Why States Invite Low-Quality International Observers
Hyde argues that states are more likely to invite international monitors when they are uncertain in their commitment to democracy, allow opposition parties to compete, and can access international benefits for being perceived as democratic (Hyde, 2011).This argument makes sense for international observers that are genuinely committed to democracy, as in her study.However, for low-quality monitors, there is a different logic at work.In this section, we focus first on domestic mechanisms (the conventional wisdom) and then on international mechanisms (our novel theoretical contribution).
Although our article explores why states invite low-quality international observers, we note that low-quality domestic observers also exist (Grömping, 2017, 417-18).A different framework is likely required to explain their presence at elections.Although low-quality domestic observers could potentially validate authoritarian elections in the way some have argued low-quality international observers do, there are reasons to think they may be less effective than international monitors at this goal.Research from Tunisia and Zambia suggests that citizens view domestic observers as less capable and neutral than some international groups (Bush and Prather, 2018, 681;Macdonald and Molony, 2023).Moreover, while we outline international factors that encourage the presence of low-quality international observers, it is unclear why they would encourage the presence of low-quality domestic observers in a similar manner.Studying why states host low-quality domestic observers and the role of low-quality domestic observers in autocratization therefore remain important questions for future research.

Low-Quality Observers as Election Validators
In democracies, politicians seek to gain or stay in office through free and fair competition.In autocracies, elections may also be held.But autocrats use elections as mechanisms to stay in power, including through cooptation, patronage, and policy concessions (Gandhi, 2008;Svolik, 2012).To ensure their political survival, authoritarian incumbents undermine election integrity through strategies like making it more difficult for opposition parties to compete freely, controlling the media, and even manipulating the tabulation of votes.Yet, winning an election through fraud that is widely recognized can create problems.It may undermine the government's legitimacy at home and abroad, leading to withdrawal of support that is crucial for regime maintenance.In some cases, stolen elections have led to widespread protests and revolutions (Beissinger, 2007;Tucker, 2007;Bunce and Wolchik, 2010;Beaulieu, 2014;Daxecker et al., 2019). 1010 Meanwhile, publics are more accepting of authoritarian governments and less likely to protest when they perceive elections as fair (Williamson, 2021).
Inviting low-quality monitors is a potential strategy for incumbents who plan to commit electoral malpractice but want to reassure the public about their commitment to democratic norms.This strategy is why zombie monitors are commonly characterized in the literature as election validators (e.g., Debre and Morgenbesser, 2017).As Walker and Cooley (2013) write, low-quality monitors' "methods are designed to warp public understanding." This argument about low-quality monitors and election validation has two steps, although they are sometimes more implied than directly theorized in previous work (see also Bush and Prather, 2022, Ch. 2).First, citizens usually have some uncertainty about election quality.Uncertainty exists because many forms of election fraud are difficult to observe; for exam-ple, manipulation may be done in secret (e.g., ballot box stuffing) or involve subtly altering electoral laws in ways that require specialized knowledge to understand (e.g., changing cam-paign finance regulations).Moreover, politicians have incentives to mislead the public about election quality.Election winners want to claim their victories are legitimate, whereas losers may want to discredit the results.These conditions are relevant in a range of regime types, but the absence of a free press in autocracies can exacerbate citizens' lack of information concerning fraud there.
Second, given the public's uncertainty about election quality, international monitors may be a source of information.Sarah Bush and Lauren Prather argue that this information comes via several mechanisms (Bush and Prather, 2022, 50-3).For example, election observers' activities can make cheating more costly by increasing the likelihood that malpractice is detected and publicized.Positive reports from international monitors can further reassure the public that cheating did not occur, whereas negative reports can do the opposite.Several empirical studies of observers' effects on public trust are consistent with this argument (Brancati, 2014;Robertson, 2017;Bush and Prather, 2022).
As Maria Debre and Lee Morgenbesser (2017) argue, even low-quality monitors may have these informational effects, despite their flaws.One reason is that domestic audiences may not realize that some observers are low quality.Low-quality observers often have legitimate-sounding names, invite individuals of high status (e.g., prominent former politicians), and mimic other features of high-quality observers.They are also, as discussed earlier, a relatively new phenomenon.Thus, citizens in the observed country may (erroneously) view low-quality observers as credible sources of information about election quality.Although it is usually straightforward for rulers to determine monitor quality in advance of an election-whether through direct observation of groups or insider conversations-this information is more difficult for the public to obtain, especially if the government seeks to conceal it.
Another reason the public may trust some low-quality observers relates to their countries of origin.Some audiences in the Global South are concerned about neoimperialist overreach by the United States through election observation (Nganje and Nganje, 2019).Many monitors that we code as low quality come from non-Western states, ROs, and NGOs that domestic audiences may view as more legitimate actors than Western ones.Supporting this logic is a study that found that observers from the Arab League-an RO with authoritarian member states-enhanced perceptions of election credibility in Tunisia more than higher-quality observers from the EU and United States (Bush and Prather, 2018).Although Arab League observers are not high-quality observers (Boubakri, 2012), Tunisians viewed them as relatively capable and unbiased.It is plausible that the public is more likely to look at identity cues to determine an observer's quality than rulers (or researchers), who are more likely to seek out information on an organization's track record of criticizing elections.This discussion suggests that low-quality monitors may reassure an uncertain public with their presence at and positive evaluations of elections.Incumbents who intend to cheat, such as leaders of authoritarian countries, may view this reassurance as helpful for their political survival.Even if low-quality monitors are not completely convincing, they can still cast doubt upon the more-critical assessments of high-quality monitors.Indeed, elsewhere in the special issue, Kelley Morrison et al. (2024) find that when international observers reach different verdicts at the same election, post-election mobilization and contention are less likely than when observers are more uniformly critical (see also Daxecker and Schneider, 2014).To maintain their grip on power, less democratic incumbents may therefore choose to invite low-quality monitors.This logic leads to our first hypothesis: Hypothesis 1: Authoritarian countries will be more likely to have low-quality mon-itors at their elections.

International Reasons to Invite Low-Quality Observers
Steven Levitsky and Lucan Way have described how linkage with and leverage from powerful democratic states have been important democratizing forces during the post-Cold War period in Central Europe and Latin America.Democratic linkage refers to the density of political, economic, and organizational ties with powerful democratic countries, whereas leverage refers to governments' vulnerability to external pressure to democratize (Levitsky and Way, 2006, 379).Linkage with and leverage from powerful autocratic states can similarly encourage authoritarian practices, including the invitation of low-quality election observers.We focus on two related dimensions of autocratic linkage and leverage: ties with states that support low-quality election observers and membership in authoritarian ROs.

Low-Quality Observer Entrepreneurship
Powerful states can use many tools to influence the regime types of other countries, including conditional aid, trade, investment, and membership in the ROs in which they play a leading role.They can also lead by example and through socialization.The evidence in favor of these dynamics is clearer with democracy promotion than autocracy promotion (Way, 2016), but there are growing signs of the latter in the cases of powerful autocracies, including China, Russia, and the Gulf monarchies (Hyde, 2020(Hyde, , 1193-994)-994).11Supporting low-quality election observers is a form of autocracy promotion.
Powerful states can serve as low-quality election observer entrepreneurs in two ways.The first mechanism is autocratic leverage.In much the same way that states are thought to invite high-quality election monitors to access Western foreign aid, RO membership, and other benefits (Hyde, 2011), states can invite low-quality election monitors to signal their commitment to a non-Western international order and access associated benefits, whether material or social.Close relations with powerful autocracies can bring a government benefits that are useful for staying in power such as investment, foreign electoral interference, cyber-attacks, military assistance, and access to surveillance technology (Way and Casey, 2018;Levin, 2020).Powerful autocracies such as Russia have also provided material benefits to specific political parties and politicians, including in democratic countries (Orenstein and Keleman, 2017;Lipps and Voeten, 2023).These actors pressure their governments to take friendly positions toward their autocratic patrons, which is a dynamic that may extend to inviting lowquality election monitors that are tied to those patrons.
The second mechanism is autocratic emulation and learning and thus is more closely related to autocratic linkage.Having ties with a state that is a low-quality election observer entrepreneur makes it more likely a country will be familiar with low-quality monitors and want to invite them.Authoritarian incumbents want strategies for holding elections without jeopardizing (and, ideally, strengthening) their rule.They can learn these strategies from other countries.States with ties to entrepreneurs of low-quality monitoring will have better access to a supply of low-quality monitors and will be more likely to turn to them.For both reasons, then, we expect countries with economic and political ties to a low-quality election observer entrepreneur will be more likely to host low-quality election monitors.We cannot empirically disentangle leverage and linkage, as high linkage gives powerful autocrats more leverage over their partners.Linkage (dense ties) may also be a prerequisite for leverage.Hypothesis 2 summarizes our expectation: Hypothesis 2: Countries with ties to a low-quality observer entrepreneur will be more likely to have low-quality observers at their elections.
Although both China and Russia (as well as some other autocracies, including Iran, the Gulf monarchies, and Venezuela) are emphasized in the literature on autocracy promotion, Russia is the most significant supporter of low-quality observers.Subverting liberalism abroad has costs and benefits for the actors involved (Wohlforth, 2020), and they have aligned in favor of promoting low-quality election monitors in Russia.Russia and other post-Soviet countries routinely invited international observers to their elections starting in the 1990s as part of their membership obligations in ROs such as the OSCE.These high-quality monitors often criticized post-Soviet elections, including Russia's, posing a threat to continued au-thoritarian rule there and becoming a source of political tension (Fawn, 2006;Marchesano, 2015).The threat was seemingly realized with the Color Revolutions, in which popular protests occurred after flawed elections that were condemned by OSCE/ODIHR observers (Beissinger, 2007, 264).
Russia led the response in the early 2000s.It attempted to delegitimize Western election observers such as the OSCE/ODIHR by arguing that they used "double standards" to criti-cize post-Soviet elections (Fawn, 2006(Fawn, , 1141)).Russia invited lowquality election observers to its elections, as did other states such as Azerbaijan.The groups came from both existing and new organizations.For example, the CIS adopted a framework for monitoring elections in the post-Soviet region in 2002.Russia also promoted the creation of new low-quality NGOs; for example, in 2002, the Commonwealth of Independent States-Election Monitoring Organization (CIS-EMO) was founded in Nizhny Novgorod, Russia.CIS-EMO is technically an NGO; despite its nominal independence, the group issues reports that are "virtually identical" to those of the CIS (Hyde, 2011, 160).More recently, in 2018, Russian-government-affiliated individuals helped create the Association for Free Research and International Cooperation (AFRIC) NGO to monitor elections in the Democratic Republic of Congo, Mozambique, and Zimbabwe, among others (Shekhovtsov, 2020, 14).Thus, countries with stronger ties to Russia have gained special access to Russia-backed low-quality observers.Although Russia's rationale for inviting and promoting low-quality monitors may have ini-tially been rooted in domestic considerations, these groups became part of the government's larger strategy of attempting to prevent the spread of democracy in its region, undermine the liberal international order, and promote counter-norms to democracy (Cooley, 2010;Tol-strup, 2015;Bettiza and Lewis, 2020;Cooley and Nexon, 2020, Ch. 4).By contrast, China has played a limited role in promoting low-quality election observers.China expressed disap-proval of international election observation at the UN throughout the 1990s and early 2000s, describing the practice as external interference in countries' internal affairs.12Yet China does not hold national elections, unlike Russia, and accordingly does not have obligations to open itself to criticism from international observers in the way that Russia does.Thus, China has less domestic interest in supporting low-quality observers that could promote counter-narratives to high-quality observers about its regime type.Although ROs in which China plays an important role like the SCO now engage in low-quality election monitoring, promoting low-quality observers globally has been a lower priority for China than Russia (Cooley, 2010).Thus, our tests of Hypothesis 2 focus on the importance of ties to Russia.In the SI, however, we also examine whether a country's ties to China are correlated with hosting low-quality observers and find they are not (SI §2).
Future research might also explore the role of other autocratic powers in spreading low-quality monitors.For example, the literature on low-quality election observers has high-lighted Venezuela as an active host of low-quality monitors at its own elections (Merloe, 2015, 90;Debre and Morgenbesser, 2017, 333).It is possible that Venezuela or other states could emerge as low-quality election entrepreneurs, especially in their regions, although we have not identified significant evidence that this dynamic has occurred to date.

Membership in Authoritarian ROs
The second international dynamic we theorize is that countries will be more likely to invite low-quality observers when they are members of authoritarian ROs.Authoritarian ROs are regional intergovernmental organizations with predominantly autocratic members.They develop strategies to legitimate (or de-stigmatize) the practices of their authoritarian members (Obydenkova and Libman, 2019;Debre, 2021Debre, , 2022;;Cottiero and Haggard, 2023).Providing positive reports about members' elections is one such strategy.We expect that membership in authoritarian ROs may encourage states to host low-quality monitors through two mechanisms that relate roughly to leverage and linkage respectively.
First, some authoritarian ROs that engage in election monitoring formally require or encourage member states to invite observers from the organization to their elections.For example, the 2002 Convention on the Standards of Democratic Elections, Electoral Rights and Freedoms in the Member States of the CIS explains that members "proceed from the as-sumption that the presence of international observers promotes openness and transparency of elections and ensures the observance of international commitments of the states." 13Mean-while, the Common Market for Eastern and Southern Africa (COMESA) states that hosting its monitors is "a sign of solidarity of the people of the COMESA Region with their brothers and sisters in a Member State as they reaffirm their commitment to democratic system of governance." 14In this way, authoritarian ROs are similar to more democratic ones like the Organization of American States (OAS) and OSCE, which have encouraged if not required member states to invite their monitors (Donno, 2013).Uganda illustrates the utility of membership in organizations such as COMESA for non-democratic incumbent governments.Uganda has hosted observers from COMESA at several of its elections, including the 2016 presidential election won by President Yoweri Museveni (in power since 1986).Although more credible sources of information about the election were highly critical, 15 COMESA monitors praised the election.Second, authoritarian ROs are sites for authoritarian learning and socialization.Meeting at their summits, authoritarian leaders share strategies for managing domestic political threats (Kneuer et al., 2019;Obydenkova and Libman, 2019;Debre, 2021Debre, , 2022;;Cottiero and Haggard, 2023).Learning about the utility of hosting low-quality monitors for reducing the likelihood of protest (Morrison et al., 2024) and observing co-members hosting these monitors socialize authoritarian leaders to the practice.Low-quality election observation is characterized by considerable emulation across countries and organizations; groups such as the Union of South American Nations and other monitors in Latin America and Africa often closely mimic the language in each other's standards and reports (Debre and Morgenbesser, 2017, 332-33).Through participation in authoritarian ROs, autocracies learn about and copy the practice of low-quality monitoring in much the same way that membership and participation in democratic ROs socializes states to democratic practices (Pevehouse, 2005).We note that insofar as Russia is a leader in some authoritarian ROs such as the CIS and SCO, this argument overlaps with Hypothesis 2. However, there are also many authoritarian ROs in which Russia plays less or no role (Debre, 2021;Cottiero and Haggard, 2023).For example, the list of authoritarian ROs active in election monitoring also includes the Community of Sahel-Saharan States, Intergovernmental Authority on Development (comprised of states in the Horn of Africa and Nile Valley), and the Organization of Turkic States.SI §3 lists countries that appear in the bottom quartile of RO democracy scores for at least one election year; as it shows, the majority are in Africa and do not appear in the top quartile of observations for export dependence on Russia.Thus, membership in authoritarian Ros may independently increase a country's likelihood of hosting lowquality observers.This discussion leads to our final hypothesis: Hypothesis 3: Countries with memberships in more autocratic ROs will be more likely to have low-quality monitors at their elections.

New Data on Observer Groups
To better understand complexity in the international election monitoring regime, we collected original data on observer presence at national elections.

The Need for New Data
Although there are several excellent datasets on international election observation, none were ideally suited to answering our research question.On the one hand, some datasets cover years prior to the current heyday of low-quality monitoring.DIEM, for example, includes information on 18 international election observation groups, including several that are not high-quality, but ends in 2004 (Kelley, 2009a, 773).The Enforcement of Democratic Electoral Norms dataset by Donno also includes information on multiple monitors (9 groups), but ends in 2008 (Donno, 2013, 54).On the other hand, some datasets are more contemporary but lack information we need.Most prominent is the National Elections Across Democracy and Autocracy (NELDA) dataset (Hyde and Marinov, 2012), which runs until 2020.Yet NELDA does not track the specific observers at elections beyond a variable designating the presence of "Western" monitors.We build on these prior data collection efforts by focusing on more recent years and a wider set of monitors.
Our dataset includes information on 1,067 national elections held between 2000 and 2020.We include all countries except longstanding advanced industrial democracies, following other studies in the literature (Hyde, 2011), because dynamics of international election observation operate differently there. 17The dataset covers legislative and presidential elections, including run-offs.Each round of multi-round elections enters the data separately.
Rather than limiting our data collection to a pre-selected list of monitors, we sought to identify every international monitoring organization present at national elections in our sample.Although it is almost certain that we are still missing some organizations, we believe our dataset represents the most comprehensive list for the countries and years it covers.For each election, we started with reports from the Carter Center and EU (if they were present), because they often list other organizations that monitored an election.
We also searched for evidence of other organizations using RO websites, government websites, and news sources.After an organization sent monitors to a country, we also searched for signs of that organization's presence at the country's subsequent elections.From there, we collected the reports of monitoring organizations present.Our team of research assistants included individuals fluent in Arabic, French, Russian, and Spanish so as to access non-English language sources and reports.

Coding Monitor Quality
A key measurement decision was how to code monitor quality.We conceptualize lowquality observers as groups that are neither willing nor able to report on malpractice.No compre-hensive list of low-quality observers exists to the best of our knowledge.Therefore, another approach was required.
Recognizing the complexity of monitor quality, we use two approaches.First, we code organizations as low-quality if they have not signed the Declaration of Principles for International Election Observers. 18This non-binding declaration has been signed by 54 groups.
Its signatories include all 12 groups classified as high-quality in a study by Alberto Simpser and Daniela Donno that categorizes groups as such using criteria that closely match our con-ceptualization (i.e., monitors that are willing and able to report on election fraud) (Simpser and Donno, 2012, 505-6). 19The high-quality monitor category in their study includes well-known NGOs such as the Carter Center and NDI as well as democratic ROs such as the European Parliament and OSCE/ODIHR. 20DOP signatories also include other groups that are not considered high-quality by Simpser and Donno but are not categorized as low-quality monitors for reasons we will discuss.

18
For the full list of organizations we code as high and low quality, see SI §4.The list of DOP signatories is available at https://www.ndi.org/declaration_endorsing_orgs(accessed April 24, 2024).
19 High-quality organizations have played an especially active role in the DOP.The Carter Center, NDI, and the UN initiated it, while those organizations and others including the International Republican Institute, the OAS, and the OSCE/ODIHR have convened annual DOP implementation meetings that are attended by less than half of the DOP signatories.For example, see "10th Implementation Meeting of the Declaration of Principles for International Election Observers, 21-22 October 2015, New York."Available at https://electionstandards.cartercenter.org/wpcontent/uploads/2017/01/DoP-10th-Implementation-Mtg-Notes-final.pdf(accessed April 26, 2023).20 Kelley (2011Kelley ( , 1546) also provides a list of high-quality observers, which is closely overlapping with but slightly more selective than that of Simpser and Donno (2012).
Although the DOP has some limited enforcement through its peer accountability mechanism, it is possible (and in fact, common) for organizations to sign it and fall short of its standards.Thus, we use signing it as a proxy for observer quality mainly because of what it reveals about monitors' willingness to publicly commit to a detailed set of standards related to election integrity and monitoring.In practice, organizations that do not sign the DOP typically have little or no commitment to democratic principles (i.e., they are unwilling to detect fraud) and do not want to develop the capabilities necessary to report on fraud (Debre and Morgenbesser, 2017, 330).The organizations classified as low-quality observers accord-ing to this approach include both authoritarian ROs associated with pro-incumbent election monitors, such as the CIS, and also truly "zombie" NGOs (in the sense of being corpse-like) like the Central European Group for Political Monitoring, which lack physical headquarters and other evidence of organizational vitality (Gray, 2018).
This approach risks both under-and over-counting low-quality observers.On the one hand, some organizations that are limited in terms of their capabilities or commitment to revealing fraud are not classified as low-quality because they have signed the DOP.One ex-ample is the SADC Parliamentary Forum, which is an international organization representing parliaments from SADC member states, a plurality of which are authoritarian. 21The SADC Parliamentary Forum's observers' preliminary evaluations of elections often avoid mentioning significant irregularities, and many of the organization's final reports are not available online.Yet by signing the DOP, we can understand them as being committed (at least rhetorically) to minimum standards in terms of democratic principles and to conducting missions with some professionalism and rigor.And indeed the SADC Parliamentary Forum adheres to regional standards for election observation, its observers undergo training, and it sometimes criticizes member states for mismanaging their elections as in Zimbabwe in 2002. 22Thus, the SADC Parliamentary Forum and other non-high-quality DOP signatories fit into a residual category of middle-quality monitors that follow neither best nor worst practices.
On the other hand, we may mistakenly classify some organizations that are willing and able to report on fraud as low-quality because they have not signed the DOP.Organizations in our dataset like the World Peace Council that engage in election monitoring as a one-time or rare event may not be aware of the DOP or view it as necessary to sign it.A lack of awareness of the DOP potentially indicates a lack of knowledge about monitoring standards and thus perhaps credulity when it comes to detecting and reporting on certain methods of electoral malpractice (Morgenbesser,21 While the SADC Parliamentary Forum is listed as a signatory of the DOP, SADC itself is not.2023).Nevertheless, some monitors that have not signed the DOP may be willing and able to criticize flawed elections.
Our second approach to coding monitor quality is therefore based on whether organizations conform with widely accepted standards for international election observers related to transparent reporting.The DOP and other codes of conduct explain that international observers ought to publicly release reports. 23If the general public and other election stake-holders cannot read what election observers have written about the conduct of their election, they cannot use observer reports as a basis for reform or tool for holding their leaders ac-countable.Organizations ranging from the African Union and OAS to the Carter Center and NDI typically release preliminary reports shortly after an election and post final reports from their observer missions online within weeks of an election's conclusion. 24  Yet many organizations routinely (or always) fail to issue reports publicly.Election pro-fessionals acknowledge that organizations that keep their reports private are often responding to political pressures to conceal their evaluations.For example, the Economic Community of West African States, an RO with a mixture of authoritarian and democratic member states, claims to produce detailed final reports, but its member states typically do not allow it to release them. 25In other cases, monitors may not release public reports because they never wrote reports in the first place.Given these dynamics, we code organizations as low-quality if they seldom release the final reports of their observer missions to the public.Specifically, if more than half of an organization's final reports are missing, we code it as a low-quality group. 26 We further code organizations as low quality if more than half of their reports are very short; we use three pages as the length cut-off for this analysis. 27The DOP and other standard-setting documents for international observers explain that monitors should observe many facets of elections.The African Union began releasing its final reports to the public in 2012.

25
Author's interview with senior elections management and monitoring expert in Addis Ababa, Ethiopia, on August 10, 2023.

26
"Missing" reports is a category that includes both reports that are entirely missing or only short statements on websites but not realized as stand-alone documents.

27
As a robustness check, we use a threshold of having more than two-thirds of the reports missing or shorter than three pages.The results are similar (SI §5.1).
Very short reports, even if publicly released, indicate that the organization lacks the willingness and capacity to report accurately on an election's integrity at a level of detail that matches this expectation.The superficial analysis found in short re-ports typically accompanies election validation.By contrast, the average report page length for highquality observers in our data is 43 pages.
Following the classification approach laid out above, we consider 107 out of 141 organiza-tions in our dataset that have not endorsed the DOP to be low quality by our first measure.Based on the second measure, 113 organizations (including 14 DOP endorsers) with more than half of their final reports missing or shorter than three pages are considered to be low quality.The 20 most-frequent election observers in our dataset are listed with their quality classification in Table 1.As Table 1 makes clear, high-quality monitors are, on average, present at many more elections than other groups.Several well-resourced, high-quality organizations observe many elections in every region.By contrast, some low-quality groups monitor only one or a handful of elections after their creation.
Turning to the election level, Figure 2 allows us to identify which countries' elections were observed by low-quality monitors between 2000 and 2020.Moving along the Xaxis, points further to the right indicate a country is more democratic using the Varieties of Democracy (V-Dem) polyarchy index (Teorell et al., 2016), which ranges from 0 to 1.The polyarchy index is an appropriate measure for our purposes because of its focus on what Robert Dahl defined as the electoral components of democracy: elected officials, free and fair elections, freedom of expression, associational autonomy, and inclusive citizenship.Since high-quality election observers assess countries' electoral performance, it is countries that fall short on electoral democracy-versus broader notions of liberal or egalitarian democracy that might be captured in other measuresthat are most plausibly seeking to validate their elections through the invitation of lowquality observers (Dahl, 1971;Teorell et al., 2016).Electoral democracies are conventionally defined as countries with polyarchy scores of 0.5 or greater.Higher Yaxis values indicate that more low-quality organizations were present at an election.
Table 1 20 Most-Frequent Observers of National Elections, 2000-2020.Low-quality observers are identified in this table as non-DOP signatories.High-quality observers are identified using the list in Simpser and Donno (2012, 505-06).Middle-quality observers are ones that are neither low-or high-quality according to these definitions.Figure 2 reveals significant cross-regional variation in low-quality monitoring.Countries in the former Soviet sphere are especially active in hosting low-quality monitors, including some that are relatively democratic (e.g., Georgia and Moldova), which is in line with our expectations regarding the importance of ties with Russia.Most of the competitive authoritarian African countries host low-quality observers, as well.Guinea-Bissau, Mada-gascar, Seychelles, and Zimbabwe are particularly active, while democratic Botswana also frequently hosts low-quality observers.These countries lend plausibility to our hypothesis about the role of authoritarian ROs in promulgating low-quality monitors.For example, Zimbabwe, Madagascar, Botswana, and Seychelles are members in several predominantly competitive authoritarian ROs that monitor elections, including SADC and the Electoral Commissions Forum of SADC Countries (ECF-SADC). 2929 ECF-SADC is formally independent from SADC but every SADC member state's electoral management body is represented.

Organization
These ROs' observers are classi-fied as low quality because they have not endorsed the DOP and do not publicly release final reports.Zimbabwe and Madagascar also have significant historical ties to Russia and have hosted monitors sponsored by Russia-most notoriously AFRIC (Shekhovtsov, 2020, 14).Hosting low-quality monitors is less common in Europe and Latin America, the regions in our dataset that are most democratic, although low-quality monitors are present there, as well.Finally, we observe that despite growth in low-quality election observers, some authoritarian states do not host any of these groups, underlining why regime type alone may be an insufficient explanation for the growth of low-quality observers.

Empirical Strategy
To test our hypotheses, we statistically examine the correlates of low-quality observers' presence at 1,067 national elections held between 2000 and 2020. 30Our dependent variable is a binary measure that takes the value of 1 if at least one low-quality election observer is present and 0 otherwise.This operationalization is in line with our hypotheses, which delineate the domestic and international reasons why countries have low-quality observers present without advancing specific predictions about the number of low-quality monitors or relative balance between low-and high-quality observers.It is also consistent with the empirical approach of some prior studies of when states host international observers, which distinguish between monitored and non-monitored elections without theorizing the total number of monitoring groups (e.g., Hyde, 2011, Ch. 2;Kelley, 2012b, Ch. 7).That said, we use a count of the number of low-quality monitors present at an election as an alternative dependent variable in robust-ness checks.The same factors that lead countries to invite any low-quality election monitors also lead them to invite more lowquality election monitors (SI §5.2).
Our first hypothesis is that authoritarian countries will be more likely to have lowquality monitors than democratic countries.We operationalize regime type using a measure of the country's lagged level of electoral democracy (polyarchy) from V-Dem (Coppedge et al., 2011).In robustness checks (SI §5.3), we use an alternative indicator of democracy from Polity, an indicator of pre-election concerns about fraud in NELDA, and a dichotomous indicator of regime type from Carles Boix et al. (2013).
Our second hypothesis is that countries with ties to Russia will be more likely to have at least one low-quality monitor at their elections than countries with weak or no ties.We operationalize this concept in two ways.First, we create a variable that measures the relative importance of trade with Russia annually as the value of a country's exports to Russia divided by the value of the country's total exports.Trade ties are an important feature of Russia's linkage with other countries (Cameron and Orenstein, 2012).Dyadic trade data covering 2000 to 2020 is assembled from the UN Conference on Trade and Development by Michael Lischka and Fabian Besche-Truthe (2022).
Next, we draw on the Authoritarian Regional Organizations Dataset (Cottiero, 2022;Cottiero and Haggard, 2023) to create a variable that measures the number of a country's co-memberships with Russia.This annual count reflects how many ROsincluding general-purpose, security, economic, and politically-focused organizations-a country belongs to in which Russia is also a member.We expect that Russia drives ROs of which it is a member to become involved in low-quality election observation.Once these organizations begin sending observers, member states are more likely to invite them to elections.In addition, belonging to more ROs alongside Russia likely reflects greater exposure to Russia's influence and higher likelihood of adopting the illiberal norms Russia promotes.
Our third hypothesis is that countries with more memberships in autocratic ROs will be more likely to have at least one low-quality monitor than countries with fewer memberships in these organizations.We draw again on the Authoritarian Regional Organizations Dataset to capture the V-Dem polyarchy scores of each country's RO co-members, averaged annually across all ROs of which the country is a member. 31 As expected, a country's polyarchy score and its membership in authoritarian ROs are negatively correlated (ρ = −0.62).
In addition to including these variables designed to test our hypotheses, we also include some additional control variables to address potential confounders and temporal dependence in the data.First, following Daxecker and Schneider (2014), we use the NELDA indicators for countries experiencing their first elections after a suspension and first multiparty elections.Hyde (2011) and Kelley (2012b) find that such countries are more likely to invite international monitors, and this pattern may extend to low-quality monitors.Second, we control for the year-to-year change in a country's democracy score, lagged by one year, to reflect whether a country was backsliding or democratizing.Though there is little prior re-search on whether countries invite more or fewer monitors while backsliding, countries with improving democracy scores invite more (high-quality) observers and join more democratic ROs (Pevehouse, 2005;Poast and Urpelainen, 2018).Third, we control for countries' lagged gross domestic product per capita (GDPpc) because higher income countries may face less pressure to invite election observers in order to secure support from partners. 3231 Cottiero (2022) calculates these scores for co-members (excluding the state under consideration) in all ROs that are "general-purpose" (e.g., integration organizations covering many policy domains) or focused on political, economic (trade or finance), and security cooperation.32 Data on GDPpc, using current U.S. dollars, comes from the World Bank.
Finally, we control for whether a state has hosted low-quality monitors at a previous election, which addresses temporal dependence in our data.As we will show, once countries host low-quality observers, low-quality observers are more likely to be present at subsequent elections in that country.
The unit of analysis is the election.We use linear probability models with region fixed effects and robust standard errors clustered by country.Linear probability models are a more appropriate estimation method than logistic regression, which assumes the independence of observations.The results from logistic regression are, however, similar (SI §5.4).

Analysis and Results
Table 2 presents our results with low-quality monitors identified as organizations that have not signed the DOP.Models 1 and 2 in Table 2 are our baseline models, which include the three variables implied by our hypotheses: level of democracy, ties with Russia (measured by trade ties or co-memberships in ROs), and authoritarian RO memberships.In Models 3-4, we introduce the control for having previously hosted low-quality observers.Our fully saturated models, Models 5-6, introduce measures for changes in democracy, first multiparty elections, and GDPpc.
In line with Hypothesis 1, more democratic countries are less likely to host low-quality monitors, and the coefficient estimate is statistically significant in Models 1-4.When we add controls for changes in democracy (lagged), first multiparty election, and GDPpc (lagged), however, the size of the coefficient shrinks and no longer has significance at traditional levels.
Next, we find consistent evidence in Table 2 that ties to Russia are positively related to hosting low-quality election observers.When operationalized based on the importance of exports to Russia in the previous year, ties to Russia are significantly related to the likelihood a country hosts low-quality election observers.The left panel of Figure 3 illustrates this relationship by graphing the predicted probability of hosting low-quality observers at different levels of trade dependence.Holding covariates at their mean values, the predicted probability of hosting low-quality observers for a country that does not export to Russia is 42 percent.For a country where exports to Russia make up 25 percent of total exports, the predicted probability of having a low-quality monitoring group rises to 60 percent. 33We interpret this roughly 50 percent increase in the likelihood of hosting low-quality observers as a substantively large effect.

33
Few countries have export dependence to Russia greater than 30 percent.We find some support for Hypothesis 3, as well.Membership in democratic ROs (lagged) generally has a negative relationship with the presence of low-quality election monitors, as expected.The right panel of Figure 3 illustrates this relationship by graphing the predicted probability of hosting low-quality observers at different levels of RO democracy scores.Holding covariates at their mean values, the predicted probability of hosting low-quality observers for a country that belongs to deeply authoritarian ROs, with an average co-member polyarchy score of 0.2, is approximately 58 percent.For a country that belongs to democratic ROs, with an average co-member polyarchy score of 0.6, the predicted probability of having a low-quality monitoring group falls to 37 percent.However, the relationship is not statistically significant in the models that also control for ROs with Russia.This finding suggests to us that it is countries' memberships in Russia-led authoritarian ROs that are particularly important for their likelihood of having low-quality monitors at their elections.
Finally, in terms of the control variables, we note that improvement (or deterioration) in democracy in the last year as captured by our lagged change in democracy variable is not significantly correlated with hosting low-quality monitors.As a country becomes more democratic, we do not observe that they are less likely to have low-quality monitors at elections.On the other hand, we also do not find clear evidence that a country holding its first election is more likely to host a low-quality election observer.Since holding multiparty elections for the first time is an important predictor of hosting (high-quality) international observers, this finding suggests a clear way that (unsurprisingly) states' decisions to host high-and low-quality monitors differ.GDPpc also does not have a clear relationship with the presence of low-quality monitors.Lastly, previously having hosted low-quality observers is a clear correlate of doing so again in our data. 3434 Since this variable is missing for most countries' first elections due to a lack of information about which groups, if any, monitored these elections prior to the year the country enters our dataset, we verify that excluding these elections altogether does not significantly alter the results of our analyses in SI §5.5.
In Table 3, we turn to the results using our alternative indicator for low-quality monitors.This indicator codes an observer as low quality if more than half of its final reports are missing or shorter than three pages long.One difference in results in Table 3 when compared to Table 2 is that the lagged democracy score is no longer significantly related to hosting low-quality monitors in any models.It is unclear why this difference occurs.One possibility is that democracies deliberately avoid issuing invitations to organizations that are not members in the DOP "club" given their support for this institution and the international election monitoring regime, but that they are less concerned about hosting observers that are inconsistent in their reports. 35At the same time, turning to the control variables, we note an additional difference.Improvement (or deterioration) in democracy in the last year is significantly correlated with hosting low-quality monitors in these models.As a country becomes more democratic, they are less likely to have low-quality monitors at elections, using our alternative indicator for low-quality monitors.Alternatively, the written evaluations of democratic elections by observers could be quite short if they are truly wellrun elections with little room for improvement.However, there is no correlation between democracy score and average number of pages across observer groups present.
More broadly, the lack of a consistently significant relationship between democracy and hosting low-quality observers across our models suggests this variable is a less important consideration than the conventional wisdom on zombie monitors has suggested.One explanation as to why is related to how Russia has supported politicians and political parties in many countries, including some democracies.In such settings, domestic politicians may invite low-quality monitors to their countries' elections for reasons other than election validation.Another possibility, which is consistent with earlier research by Daxecker and Schneider (2014), is that the relationship between democracy and the presence of low-quality observers is non-linear.Supporting that explanation, we find that partial democracies are more likely to host low-quality observers than are the most autocratic and democratic countries in our data (SI §6.1).
We also find that the association between democracy and the presence of low-quality observers changes over time (SI §6.2).Low-quality observers are more likely to be present at elections in more democratic countries in the early 2000s and less likely to be present in more democratic countries in later years.This pattern could reflect growing awareness of the threats posed by low-quality observers in democracies.
Meanwhile, we continue to see strong support for our expectations related to international influence.We find support for the hypothesis that ties to Russia encourage the presence of low-quality observers.In all models in Table 3, exports to Russia are strongly and positively associated with hosting low-quality monitors.We also see that both the autocratic RO variable as well as the ROs with Russia variable are correlated with the presence of low-quality observers at elections, though the significance of the coefficients varies across models.We view these results as additional support for Hypotheses 2 and 3.In fact, regardless of how we code low-quality observers, the variables we use as indicators of both Russian influence and authoritarian RO membership are more consistently correlated with the presence of low-quality observers at elections than a country's level of democracy.
In the appendix (SI §7) we present results with another alternative measure of our dependent variable that combines low-and middle-quality observers into one "nonhigh-quality" observer category.We find, again, that democracy is not significantly associated with hosting low-quality observers using this broader categorization.We still see that Russian influence is a significant predictor of hosting low-quality observers using this broadened definition.RO democracy scores are also still associated with the likelihood of hosting low-quality observers, though this relationship is attenuated.

Conclusion
In this article, we presented new data on the complex regime of international election monitoring.The data revealed important insights about which states have low-quality monitors at their elections.We advanced an argument about how international factors-specifically ties to Russia and learning within authoritarian ROs-have led to the diffusion of these groups, which our data supports.This study therefore provides an important addition to the previous literature on zombie election monitoring, which has focused more on how these groups can reassure the public about election credibility and thus help autocrats hold onto power (Walker and Cooley, 2013;Debre and Morgenbesser, 2017;Morrison et al., 2024).
Beyond its contribution to our understanding of international election monitoring, this study sheds light on broader trends related to authoritarian resurgence and autocratization.In particular, it contributes to our understanding of the geopolitical dimensions of these phenomena (see also Cooley and Nexon, 2020;Hyde, 2020;Samuels, 2023).An authoritarian power-Russia-has played a key role in the diffusion of low-quality election monitoring, which is a practice that undermines democracy and global democratic norms.Geopolitics may play a similar role in diffusing other anti-democratic practices, such as laws restricting the activities and funding of civil society organizations (Chaudhry, 2022).Although the rise of China and other powerful autocracies may be important in some anti-democratic trends, it is countries in Russia's sphere of influence that have been especially likely to adopt new global counter-norms related to low-quality election monitoring.
This study also suggests at least three promising directions for further research.First, it is widely known that authoritarian governments have a "menu" of options for manipulating the electoral playing field (Schedler, 2002;Morgenbesser, 2020).They have proven adept at moving away from outright fraud in favor of longer-term strategies like manipulating the media and stacking courts.Yet despite the availability of such tactics, low-quality monitors are an increasingly common feature of elections.Future studies might explore whether hosting low-quality monitors complements or replaces other strategies that would also help them stay in power.
Second, our data shows that dozens of organizations embraced the practice of lowquality election monitoring in the 21st century.What explains variation in which ROs have begun to monitor elections?Elsewhere in this special issue, researchers have demonstrated the importance of member states' regime types for good governance mandates (Hafner-Burton et al., 2024) and human rights shaming (Meyerrose and Nooruddin, 2023).RO membership may similarly explain which organizations begin to monitor elections and with what standards.
Finally, the effects of low-quality monitors can be studied.Their growth raises questions about how high-quality monitors have responded.In the past, competition among highquality international observers-which often monitor the same election-has prompted a "race to the top" in which these groups have professionalized and improved their methods (Hyde, 2012, 59).When low-quality monitors are also present at the same election as high-quality monitors, do the latter respond strategically to maintain their authority, perhaps by being more critical in their reports?Are high-quality monitors' efforts to distinguish them-selves from low-quality monitors through initiatives such as the DOP successful?Answering these questions has the potential to shed light not only on the organizations active in election monitoring but also contribute to broader debates about how organizations establish and maintain authority within complex international regimes (Tallberg and Zürn, 2019).
The rise of low-quality election monitoring also raises important questions about whether inviting these groups helps autocrats who seek to remain in power through elections.In another contribution to the special issue, Morrison et al. (2024) show that when monitors-including high-and low-quality groups-issue competing verdicts about flawed elections, the public is less likely to engage in elected-related protests and violence than when monitors reach a consensus about election quality that is critical.A next step in the research agenda is to explore whether the likelihood of incumbent turnover is lower when low-quality observers validate elections.Morrison et al. (2024)'s data on international media coverage of monitors, drawn from Donno and Gray (2023), might also be combined with ours to understand which types of monitors get covered in the press.Indeed, future research should take up the question of how international audiences respond to the activities of low-quality observers.The end of the Cold War increased the international (Western) benefits associated with democracy and thus with inviting high-quality election observers (Hyde, 2020(Hyde, , 1194)); by extension, inviting lowquality election observers may have been associated with international costs.But the recent decline of such benefits and the rise of authoritarian great powers suggests that there is less to lose and more to gain internationally by hosting low-quality monitors, and these shifts may contribute to authoritarian regimes' survival.International reactions to low-quality monitors should therefore also be investigated going forward.

Figure 1 .
Figure 1.Count of Monitored Elections by Organization Type, 2000-2020.We discuss variable operationalization and measurement in more detail later.

Figure 2 .
Figure 2. Presence of Low-Quality Observers by Country Democracy, 2000-2020.Graph plots the total number of low-quality observer missions across all elections, 2000-2020.The X-axis indicates the country's quality of electoral democracy as measured by V-Dem(Coppedge, 2019) during the election year.Low-quality observers are identified in this table as non-DOP signatories.

Figure 3 .
Figure 3. Predicted Probabilities of Hosting Low-Quality Observers.Calculated based on Model 5 in Table 3 using 95% confidence intervals.Hatch marks along the X-axis indicate the distribution of the explanatory variable.

Table 3 .
Correlates of Hosting Low-Quality Monitors (missing reports or short statements), 2000-2020.Estimates are based on linear regressions with region fixed effects.Robust standard errors clustered by country are in parentheses.*** p < 0.01, ** p < 0.05, * p < 0.1.35