Political Uncertainty E-Book

Gergana Dimova

• Herausgeber: ibidem
• Kategorie: Geisteswissenschaft
• Serie: Soviet and Post-Soviet Politics and Society
• Sprache: Englisch

Seitenzahl: 317

Veröffentlichungsjahr: 2022

Beschreibung

This timely book provides a comprehensive, multi-dimensional and comparative analysis of political uncertainty. It is innovative in introducing the notions of inter-institutional, verbally induced, and historical uncertainty. It argues for an inclusive approach which considers multiple aspects of uncertainty, even when they are of a different nature. Combining aggregate statistical analysis and qualitative case studies, it compares political uncertainty in established and non-consolidated democracies. Overall, this book furnishes important insights into uncertainty in political life and how the discipline of political science is coming to terms with it.

Leseprobe

ibidem Press, Stuttgart

To M & M

Contents

1 Uncertainty as a Multi-Dimensional Concept A Brief Summary of the Argument

2 Uncertainty A Critical Overview

3 Uncertainty in Non-Democracies

4 Uncertainty in Democracies

5 Inter-Institutional Uncertainty Do Institutions Compete or Cooperate with Each Other?

6 Democratic Uncertainty Fundamental or Processual?

7 Historically Induced Uncertainty

8 Verbally Induced Uncertainty

9 Future Research on Political Uncertainty A Plea for a More Integrated Approach

1 Uncertainty as a Multi-Dimensional ConceptA Brief Summary of the Argument1

As I write this manuscript in the early stages of a coronavirus pandemic, I stare at the face of uncertainty. I am very uncertain as to whether I should make an international trip. According to one classical definition of uncertainty, I experience the “inability to assign probabilities to the likelihood of future events” (Stevens 2014, p. 432). I am uncertain about the probability of negative consequences, if I decide to take the trip. At that point, the probability of catching the virus constitutes the ultimate “state of uncertainty as existing for an event when no numerical probability of the event occurring can be assigned" (Knight 1920 cited in Cyert and DeGroot 1987, p. 3). The very existence of two very stark, mutually opposed scenarios, to travel or not to travel in the midst of a ranging pandemic, and the inability to assign precise probabilities to either of them, creates a high degree of uncertainty. Having a choice is very unsettling, but having an ill-defined choice is even more unnerving. This relatively simple scenario multiplied itself endlessly during the pandemic as people asked themselves whether they should cancel their vacation plans, whether they should go to the shops and whether they should send their children to school. Suddenly, the whole planet was overwhelmed by uncertainty—and that is by no means an uncertain statement.

The pandemic brought with itself the concept of uncertainty defined as “heightened unpredictability” not only in terms of personal choices. Uncertainty was also unleashed as the belief in the power of science to predict the rise and spread of the pandemic was shattered. Before the eyes of the public, epidemiologists argued, disagreed, sometimes even contradicted their own past statements about the deadliness of the virus, its infectiousness, whether wearing face masks was beneficial and how long the virus survives on metal surfaces. The idea that science is not an unambiguously reliable source of guidance augmented the sense of uncertainty. As doubt was cast on the precision of scientific knowledge, it became clear that the utilisation of scientific knowledge in politics and public policy was, to a degree, subjective and political, and therefore it was rife with uncertainties as well. A rift between science as a monolithic body and scientists as a heterogeneous community was laid bare. It became clear that politicians handpicked the scientists, whose advice they wanted to follow, thus undermining the belief that “science-based” policy is unequivocal. This realisation did not decrease uncertainty either.

While 2020 was the year when the notion of uncertainty reasserted its powerful grip, it was by far not the only instance when uncertainty reigned supreme. The pandemic heightened the importance of studying uncertainty, but it did not create uncertainty per se. It created the urgent necessity to talk about uncertainty. But the questions have been long overdue, long before the pandemic made its point. Here are some of them: is uncertainty borne out of people’s inability to imagine such a powerful exogenous shock to the world? Does uncertainty stem from newness, in the sense that scientists by default do not have enough data to gauge the nature and the extent of a new phenomenon and a new threat? Or is uncertainty related to the lack of sophisticated methodology to assess the already available data? Alternatively, is uncertainty so endemic, systemic and pivotal that no matter how much data is available, there will still be a large, unfathomable margin of error in any prediction? Finally, could uncertainty be fueled by the diverse and often conflicting reporting of an event in the media? These are only a small number of the questions that the present book raises.

The book provides a comprehensive and methodological understanding of uncertainty in politics, but it relates it to uncertainty in many disciplinary fields. This introductory chapter builds upon the accumulated inter-disciplinary knowledge to create a framework for studying political uncertainty in particular. To do so, it performs a comprehensive overview, which critically analyses the sources, types, definitions and the measurements of uncertainty. Looking at the available scholarship, it is fair to conclude that there is a great degree of variation on understanding “uncertainty” as a concept.

Uncertainty: Two Main Claims of the Book

Before positing what the book claims, it is important to state what it does not claim. The main argument is not that uncertainty is invariably dangerous, insidious, destructive or that it is entirely knowable. On the contrary, it is fully possible that uncertainty breeds many positive aspects of life. John Keane (2022) makes a convincing argument to this effect. The book argues, however, that to the extent that political uncertainty should and could be conquered, it should be subject to a comprehensive and systematic analysis. Thus, it makes two main claims: (1) that we need to view uncertainty as a multidimensional, integrated concept and (2) that we need to consider new types of uncertainty.

The first claim—conceptualizing uncertainty as a multidimensional concept—is borne out of the understanding that the multiplicity of existing types of uncertainty should not be viewed individually but collectively. It is also based on the contention that most political processes should not be conceived as a one-stage event but as a chain of events. Even when these events are very minor, they still need to be differentiated because they embody different types of logic and a different type of uncertainty. Viewing the pandemic through the lens of only one type of uncertainty would be misguided. Many of the existing types of uncertainty could be applied to it: (1) endogenous vs. exogenous uncertainty, because the virus came from outside of the countries and was an external shock, but it was resolved endogenously within each country; (2) procedural vs. substantive (or output) uncertainty, because sticking to the procedures of dealing with the virus implied a different kind of uncertainty from the uncertainty underlining the outcomes of implementing this procedure; (3) known vs. unknown uncertainty, because the uncertainty related to how people reacted physically to the virus was less known than the uncertainty related to how much people were afraid of the virus and were willing to follow government advice; (4) state vs. effect vs. response uncertainty is also a useful concept because it captures the difference between the inability to assign probabilities to the spread of the pandemic, to the effects of the pandemic, and to the various response options respectively.

Of all these types of uncertainty, the book focuses on two already established types of uncertainty-institutional and substantive; and it adds three more types: inter-institutional, verbally-induced and historically-induced types of uncertainty. The reason why the book singles out institutional and substantive types of uncertainty is that they have been instrumental in gauging uncertainty in new democracies, but also because the study of the uncertainty-reducing effects of institutions has been fundamental in the fields of international relations, public policy and public administration: “To cope with uncertainty, institutions align incentives for information revelation; to handle difficult problems, institutions create incentives for diverse problem-solving approaches; and to harness complexity, institutions adjust selection criteria, rates of variation, and the level of connectedness” (Page 2008, p. 115). Whether institutions are perceived as a procedure—and this is mainly the procedure of elections—or whether institutions are taken at face value—as government departments, NATO, European Unions, etc.—they all have one ‘magic’ quality, which is usually put at the centre of studying uncertainty—they equalise expectations about how things should be done.

Institutions reduce the unpredictability of coordinating multiple people, who undoubtedly have diverse and often incompatible ways of doing things. While institutional uncertainty is about how things are done, substantive uncertainty is what has actually been done. It would be one-sided to study only institutional uncertainty without studying substantive uncertainty because that would presume that outcomes always follow institutional rules, which will be a severely reductionist argument as it will automatically exclude a myriad of other causes. Thus, not only viewing institutional and substantive uncertainty together, but viewing the relation between the two of them is paramount. The book takes a step in that direction, although more research correlating the different types of uncertainty in a systematic way is needed.

Apart from building on scholarly wisdom about institutional and procedural uncertainty, the present analysis underscores the astuteness of existing actor-based accounts of uncertainty. In the context of the book, actor-based uncertainty is presented in a new light. A decision does not have to be made collectively—such as an electoral vote, for example—to influence a collection of people. Imagine, for example, the disastrous effect that a pilot’s error of judgment can have on the fate of a flight. This view will chime in with the conclusions of the scholarship on the impact of individual leaders. Even if the counterargument holds that leaders are a product of structural circumstances and express a collective identity, the argument about the uncertainty inducing power of individuals still holds because the individuals triggering havoc need not be high-placed or elected. This person could be a nameless driver who causes a small incident on the highway, thus changing the plans of possibly hundreds of people in the cars behind him. It could be an anonymous terrorist who blows up a marketplace, thus changing not only the lives of its victims but those of their families as well. Actor-related uncertainty is important to recognise and to study.

The media age is the main reason why actor-related uncertainty is more important nowadays than ever. Media platforms can augment individual claims, so it does not matter whether a claim is made by one person or a majority of the population. Through the media, this claim can reach a sizeable share of the population. In this case, the number of makers or triggers of the event is irrelevant because the event becomes known to many people. Before the advent of the media age, people had to congregate physically or sign petitions to have the ear of the government or society. Alternatively, they had to have access to the relatively closed circle of newspaper, TV and radio networks. With the advancement of telecommunication technologies, these impediments are almost obsolete. People can make claims in easily accessible platforms, and sometimes these claims “travel.” They can be “picked” by more popular outlets. Alternatively, the makers of the claims can email or send electronically compromising material. Such information “dumps” are not uncommon. They can range from Wikileaks to an arguably authentic video taken in the residence of the former Bulgarian prime-minister Boyko Borisov, which showed a drawer in his nightstand packed with 500 Euro notes, gold bars and a pistol. It takes only one person to blow the whistle, and technology has diminished the importance of how powerful that person is. That is why individually induced uncertainty matters, especially in the Media Age.

If we already have such a long list of types of uncertainty—fifteen and counting—why should we include three more types? The ones introduced in the book are: verbally-induced uncertainty, inter-institutional uncertainty and historically-induced uncertainty. The first type of uncertainty is verbal uncertainty, or verbally-induced uncertainty. It denotes the uncertainty not only whether someone is telling the truth or not, and fake news has made the possibility of this uncertainty more probable. Verbal uncertainty is also the uncertainty induced by various blame avoidance strategies, which seek to interpret and re-interpret events. If we exclude verbal uncertainty, we exclude the possibility that words uttered in the media can change the public perception of the world.

This neglect of blame avoidance strategies as a source of uncertainty is puzzling but easy to explain. Blame avoidance strategies exists as a separate and self-sufficient sub-field in communication and the public policy fields and integrating it in the study of uncertainty would constitute not only a disciplinary bridge but a disciplinary breach. To understand the link between uncertainty and blame avoidance, we need to imagine certainty before and after blame avoidance utterances. Image, for example, that the pictures of the Bulgarian Prime Minister’s cash-packed drawer were published and the prime minister had not said anything or he had confessed to possessing a huge amount of cash, some gold bars and a pistol. The level of uncertainty would have been smaller as most people would have most likely accepted this as a fact.

Now imagine that the prime minister said, as he actually did, that this is a set-up, and that it makes no sense for him to keep cash in his nightstand, when he had a safe in the corridor. Furthermore, he said that the prolific quantities of cash did not tie in with the “ascetic” way of life portrayed in the video overall. And finally, he added, the person who took the video knew exactly which was the drawer with the cash, and opened it, when there were several identical drawers. All these facts, the prime minister concluded, proved that the maker of the video put the cash and the gold in his drawer to compromise him and to destabilise the government (DarikNews, 2020).

From this point on, the number of people who “bought” Borisov’s version of events shot up. It does not particularly matter who these people were. What counts, from the point of view of uncertainty, is that an alternative version of events came into existence, and it sparked an avalanche of supportive comments under the online articles reporting the prime-minister’s version of events. What this means in terms of uncertainty is that the prime minister’s utterance has sown the seeds of doubt. It has moved the discussion from the sphere of facts as observed through pictures into the sphere of opinion as manifested through verbal statements. Where there is doubt, there is uncertainty.

The second—and most consequential—type of uncertainty that the book introduces is inter-institutional uncertainty. Inter-institutional uncertainty is the uncertainty arising from the interaction between institutions. While institutional uncertainty tends to focus on institutions in general or on very specific institutions, such as elections, it neglects to understand how the dynamics between the institutions unfold. This is a consequential omission because institutions neither exist in isolation, nor do they all embody the same rules. In fact, most of the accountable institutions tend to check and balance each other, which means that they are expected to compete rather than cooperate. This element of competition between institutions raises the level of uncertainty. But even if institutions cooperate, it is unclear what will come out of that cooperation. The very choice whether institutions will cooperate or compete further complicates uncertainty, which I call inter-institutional uncertainty.

The third type of uncertainty introduced in this book is historically-induced uncertainty. The purpose of creating this category of uncertainty is to ensure that decision makers and decision takers portrayed in political science do not suffer from amnesia. It is unrealistic to expect that future predictions should be based on all sorts of factors anchored in the present, while omitting past behaviour. This does not mean that power-holders are particularly wise in “learning from history,” as it has been proven that they rarely learn from history (Wilsford 1994; Greenhalgh et al 2011). Neither does historically-induced uncertainty imply a path dependent or a deterministic argument. Rather it is a rationalization purporting that people may look back to pick clues for present and future behaviour. This is especially applicable to transitional and formerly authoritarian countries because they lack a clear system of signalling popular discontent, and thus occasional uprisings in the past offer important information about public support in the present.

Having briefly described the three types of uncertainty the book introduces, it is important to go back to my original claim that uncertainty needs to be perceived as a multi-stage and a multi-dimensional concept. No single type of uncertainty can “capture” all uncertainty. This is so because life unfolds as a process rather than as a single event. Even a single event—such as a pandemic—could be dissected into a process consisting of a series of sub-events. This is not splitting hairs. It is a methodological exercise which acknowledges that the beginning and the end of an “event” is ultimately the result of a subjective decision of a researcher. It is essential to differentiate between these smaller sub-events because each of them is likely to express a different type of logic and uncertainty. This view acknowledges the distinctiveness of each stage of the process and the inter-connectedness between all stages in the process. To demonstrate the explanatory utility of the approach of perceiving uncertainty in processes as multi-stage chain events, I briefly apply it to the above-mentioned case of the nightstand of the former Bulgarian Prime Minister Borisov.

Figure 1.1: Uncertainty as a Chain Process: The Cash in the PM’s Nightstand Story

As figure 1.1. suggests, political life in Bulgaria went up in flames with the explosive photos and videos showing pictures from the former prime minister’s bedroom. The material was compromising in that it insinuated that the then prime minister was much richer than he had admitted. The secrecy and the abundance of the cash hinted at its unsavoury origin. This accusation testifies to the power of a single person or a small number of people with the access to compromising material to cast a big shadow over the government. This is the uncertainty injected into political life by political and other actors in the first stage of this process.

The second stage in this illustrative case is the strategy for spinning the story. Mr. Borisov came up with a few arguments why the possessions could not have been his but were inserted there by ill-wishers. In this stage, it is the verbal framing of the pictures that introduces another level of uncertainty. In the next stage, institutional uncertainty arose because it was unclear how the accusations would play out within the Bulgarian parliament, the European Commission, and the anti-corruption protests that erupted. But the point I want to make is that inter-institutional uncertainty was equally as important, as it reflected the dynamics not within but among the Bulgarian parliament, the European Commission, and the anti-corruption protests. It could be argued that the institutional responses were linked: because the photos were not condemned by the EU and the EU gave Bulgaria huge loans just a few days after the pictures emerged, the Bulgarian parliament turned down a vote of no confidence in Mr. Borisov, and the protests lost steam. This move by the European Union reduced the political uncertainty unfurled by the photos. In the end stage—which comes fourth in figure 1.1.- there were only two substantive outcomes, a resignation of the government or a lack thereof. The government did not resign. This case illustrates that a fundamental difficulty of conceptualizing uncertainty is to find the balance or the sum between the different types of uncertainty involved in multi-stage and multi-dimensional processes.

Why Should We Study Uncertainty as a Process?

The book examines uncertainty as a political process. This approach presents several considerable advantages in comparison to existing approaches. One current approach to uncertainty is to pick one or a few types of uncertainty (for example, institutional, perceptual or endogenous) and see how they apply to economic or political processes, or how they are generated by them. The reverse approach is to focus on a very narrow interaction, which is usually a single shot game and to analyse the uncertainty inherent in it. Both of these methodological approaches—studying uncertainty as type or as a single shot game—carry two dangers: that some types of uncertainty are omitted and that the relationship between the different types of uncertainty is underexplored. For example, if one sets out to study institutional uncertainty, they may miss out the uncertainty inherent in verbal interactions. If they study uncertainty as a single-shot game, they may miss how uncertainty unfolds over a number of games that do not presuppose fixed structures underlying all games. Analysing uncertainty in a process helps avoid both dangers. It enables the researcher to include a variety of types of uncertainty, which emerge in the consecutive stages of the process. At the same time, it enables the scholar to observe and analyse the parameters, determinants and causality between the different types of uncertainty arising in different situations.

Studying uncertainty as a political process (rather than as a single type or a single shot game) is achieved in this book by approaching uncertainty through the lens of government accountability. The book studies the types, sources and constellations of the types of uncertainty, when the government is publicly alleged in the media. The huge importance and diversity of uncertainty was revealed to me while I was writing my book Democracy beyond Elections: Government Accountability in the Media Age (2019). In a way, the present study is a logical continuation of this book. I was surprised how uncertain it was whether the government would be held accountable for media allegations or not. Above all, I was struck by how decentralized a process accountability was. While studying which groups make the accusations, which groups pursue the accusations, which forums investigate the accusations, the enormous repercussions of this decentralization dawned on me. The idea of studying uncertainty gathered speed while I was describing the various groups that are interested in pursuing various types of allegations, on the one hand, and the various forums, which enable them to do so, on the other hand. The degree of uncertainty was finally revealed to me when I realized how uncertain it was how these accountability seekers will pair up with the accountability forums. It is this particular “pairing up” that was the best argument to pay greater attention to uncertainty through the lens of accountability. In retrospect, it becomes evident that it is particularly suitable to study uncertainty through the process of accountability for the three reasons described below.

First, approaching uncertainty through the lens of accountability is both productive and illuminating because accountability is a chain process consisting of well delineated stages, or sub-events: an accusation, verbal response, investigations and sanctions. The benefit is that it is easier to “capture” different “pure” types of accountability in the separate sub-stages. Thus, one can isolate the importance of actors in the stage of accusation-making, the importance of verbal uncertainty in the stage of blame shifting, to identify inter-institutional types of uncertainty, to tease out the uncertainty of matching accusers to forums in the investigative stage, and to think of sanctions in terms of substantive uncertainty in the sanctioning stage. This approach also has the advantage of being a linear process in the sense that the accusation cannot come before the investigations, for example.

The second reason is that viewing uncertainty as a sequence embedded in the accountability process makes methodological sense. Deconstructing a process into sub-stages and then deducing the types of uncertainty embedded in each sub-stage is diametrically opposed to settling on one type of uncertainty first and then trying to figure out how this particular type of uncertainty is manifested in various processes. The methodology used in this book is easier to track down as it unfolds in a relatively easily observable sequence of stages, rather than as a compilation or an aggregation of events or spheres, which are difficult to disentangle.

From a methodological point of view, it is wiser to study uncertainty as a chain process rather than uncertainty as a nested phenomenon. From a pragmatic point of view, however, it may be hard to tell the difference. For example, in The New Despotism, Keane (2020) suggests that institutional uncertainty—as manifested in the pretence to follow democratic rules while undermining them—is heavily connected to ambiguities in many other spheres of life, such as the economy and the media. This beckons the question whether a high degree of institutional uncertainty can only co-exist with high degrees of other types of uncertainties. Relatedly, another interesting and pressing inquiry is whether institutional uncertainty can exist independently in a single sub-stage of a process and whether it is only functional insofar other uncertainties play a supporting role. The graphical representation of the two approaches to uncertainty is depicted in figure 1.2.

Figure 1.2: Uncertainty in a Chain Process and Uncertainty as a Nested Phenomenon

Future research should study the precise degree of causal connection between the various types of uncertainties. Can we view uncertainty as a nested phenomenon or as a chain process? Are uncertainties in different spheres of life, uncertainties caused by different factors and uncertainties measuring different dimensions linked? When does one type of uncertainty imply another type of uncertainty? For example, one could argue that the fact that the Russian president exerts a lot of power over the prosecutorial office could be interpreted both in terms of institutional and inter-institutional uncertainty. Institutional uncertainty is prevalent because there is so much informal influence that it is very uncertain whether the prosecutor will follow the formal rules of the office. However, this finding, which is gauged in terms of institutional uncertainty, can also be conceived as inter-institutional uncertainty, which measures the unpredictability arising from the interaction of the presidential institution and the office of the prosecutor general.

The third advantage of studying uncertainty in connection to analysing government accountability is that it is possible to embed uncertainty within substantive debates in political science. Too often uncertainty is either studied in the abstract or is applied exclusively to economic phenomena. Isolating the notion of uncertainty from thorny questions in political science is counter-productive. By contrast, all theories in political science could be interpreted to gauge the degree of uncertainty in the world. The studies employing quantitative methodology report on the likelihood that a particular variable has an impact on an outcome they are interested in. In a sense, they ask: what is the likelihood that A causes B, and how certain are we about this causal connection? But uncertainty features only implicitly in such analyses. Explicitly, uncertainty is rarely embedded in substantive political science debates. Some notable exceptions are studies of uncertainty as it relates to political parties, political violence, legislations, Supreme Court decisions, deterrence, elections and regime change (e.g., Mueller and Rauh 2017; Combs 1980; Stephenson 2005; Lupu and Riedl 2013; Schedler 2013; O’Donnell and Schmitter 2013; Kilgour and Zagare 1991).

In this book, uncertainty is linked to three political science questions: the EU accountability deficit, the process of presidentialisation of the accountability process and the phenomenon of the multiple accountability disorder. Chapter three highlights the lack of uncertainty guiding the relations between the president and other institutions, when investigating the government in Russia. By showing how the president has monopolized the investigative and sanctioning process, it presents a novel perspective on the debate over presidentialisation. Chapter four throws light on the EU democratic deficit from the point of view of the EU investigating the German government. Learning whether the EU has a disproportionate power over holding the incumbents to account provides a novel point of view on democratic accountability. Chapter five links uncertainty with the thorny issue of the interaction between institutions and how they impact democracy.

Why Should We Study Uncertainty in Democracies and Non-Democracies?

The book takes a comparative approach to the study of uncertainty. It is the first study that presents an explicit and integrated analysis of uncertainty in democracies and non-democracies. It is generally assumed that uncertainty “behaves” differently in democracies and non-democracies, both in terms of its intensity and its type. Interestingly, Schedler (2013) suggests that both in consolidated autocracies and in non-consolidated democracies, institutional certainty is high. Figure 1.3., which is reproduced from Schedler’s seminal work on political uncertainty, posits that high uncertainty is prevalent both in authoritarian regime crises and in democratic regime crises. In general, it is argued that “strong institutions create deep certainties, weak institutions much less so” (Schedler 2013, p. 23). So measured in terms of intensity, uncertainty is the same in diametrically opposed regimes. The logical question that arises then is whether this high degree of uncertainty is of the same type in democracies and autocracies. Is a high degree of institutional uncertainty manifested similarly in dictatorships and non-dictatorships? For example, Schedler’s analysis explains convincingly that institutional uncertainty is high in electoral authoritarianism because of the threat of rivals, the threat of rebels, the threat of retaliation and ignorance (Schedler 2013, p. 23). Given the lack of a similarly thorough comparative analysis of the high degree of institutional uncertainty in competitive regimes, should we assume that the above factors that lead to institutional uncertainty in authoritarian regimes are simply nullified in competitive regimes, or do they still apply to a degree?

Figure 1.3: Continuum of Institutional Strength and Uncertainty2

It is not necessarily so problematic that uncertainty in non-democracies is studied more extensively than uncertainty in democracies. The problem is that uncertainty is often studied using different explanatory paradigms, and it is consequently not clear how they translate from one regime type to another. The study of uncertainty is somewhat one-sided but this one-sidedness could be avoided by juxtaposing and analysing the regime types in conjunction. For example, it is argued that regime uncertainty is high in authoritarian countries and that it results in “political actors, who must make strategic decisions while assigning positive probability to the breakdown of democratic institutions without being able to foresee the subsequent institutional arrangements” (Lupu and Riedl 2013, p. 1,346). In the absence of a comparative analysis of regime uncertainty, we are led to believe that by implication it is non-existent in consolidated democracies. However, the recent literature on democratic backsliding (Cianetti, Dawson and Hanley 2018) and the demise of democracy (Runciman 2018) seem to suggest that regime uncertainty should not be automatically written off in democracies. For example, Mechkova, Lührmann and Lindberg (2017, p. 163) suggest that “in 2013 alone, five countries went from autocracy to democracy, and nine went the other way. This volatility suggests a fair amount of uncertainty as to how robust the democratic gains of the last four decades or so actually are.” Furthermore, there is the argument that developing democracies are set to suffer from economic uncertainty, because their small economies are particularly viable to exogenous shocks and to the market volatility of the global economy (Lupu and Riedl 2013, p. 1,346). But developed and well globalized economies are also very exposed to exogenous shocks. Does it mean that they suffer from uncertainty as well? By focusing on uncertainty separately and individually in different regime types, we fail to trace whether factors that contribute to uncertainty in one type of regime feature in another type of regime, without causing the same uncertainty. We also fail to see how the variations both in the types and the degrees of uncertainty co-vary across established and non-established democracies.

Figure 1.4. makes some tentative suggestions about the variation in the intensity and the nature of uncertainty. It conjectures that institutional uncertainty—that is the likelihood that actors will comply with institutional rules—is high in non-democracies, but actor-based, inter-institutional and verbal uncertainty is low. Conversely, institutional uncertainty in democracies is low, but actor-based, inter-institutional and verbal uncertainty is high. This logic is empirically tested in chapters three and four. But this is just the beginning of thinking about uncertainty as a multi-dimensional process. We need a systematic analysis of the causal links between them.

Figure 1.4: Hypothesized Types and Levels of Uncertainty in Democracies and Non-Democracies

Uncertainty and Political Science as a Discipline

Uncertainty is intrinsically linked to the larger question about whether political scientists can predict phenomena or simply explain them retrospectively. Political science does not have an outstanding track record of predicting things. It failed to predict the end of communism, the outcome of the Brexit referendum (Hay 2017) and most scientists failed to predict the rise of President Trump (Blakely 2016), for example. Trump’s victory was so surprising that some pundits asked whether it did not “destroy political prognostication forever” (Cillizza 2020). Without a doubt, prediction is more attractive than explanation. “Harry Truman once joked that he wanted to hear from a one-armed economist because he was sick of hearing “on the one hand … on the other …” (Tetlock, Mellers and Rohrbaugh 2014, p. 138). Even if explanations can furnish a sense of order in an inherently chaotic world, predictions provide a sense of control in an unsettlingly variable future.

But there is more to the soothing sense of control that predictions can contribute: predictions can alter our behaviours so we can avoid negative outcomes: “we explain in order to give ourselves a more informed perspective on how we might intervene to produce better outcomes in the future (our interest is prospective)”(Hay 2017, p. 184). Predictions can have many practical implications. Certainly, predicting the onset of political violence is important (Mueller and Rauh 2018). So is predicting Supreme Court decisions (Ruger et al 2004). The practical significance of political science has been highlighted by the rise of the political risk industry. Firms, such as Eurasia Group, The Economist Intelligence Unit and Control Risks, are employed by businesses and governments to predict the risk associated with certain investments and policies.

Given the attractiveness of predictions, and political science’s unconvincing record of making predictions, the question arises “is political science obsolete”? If the discipline can offer explanations only after the phenomena in question have transpired, is this “retrospective explainability” (Taleb 2007) or “retrospective intelligibility” (Fernando, Smith and Perez 2021) good enough? Certainly, if meteorologists could not predict a thunderous storm that could affect the path of a plane, we would be less likely to respect meteorology as a discipline. Similarly, if medicine is powerless to predict the progression of a disease, but it can only say how the disease developed after it has spread, that would cast a big cloud over its value added. Shall we judge political science by means different from those employed in the natural sciences? And if we do so, and political science fails the test, does it mean that it is an obsolete discipline in this respect and an obsolete subject to study at universities?

There is no easy answer to this question, but I will nevertheless provide four arguments that could be used to salvage political science as a discipline. One way out of this conundrum is to conjecture that the natural world, just like the social world, is indeterminate and unpredictable. Heisenberg’s uncertainty principle, which applies to quantum physics, roughly states that “one cannot assign exact simultaneous values to the position and momentum of a physical system. Rather, these quantities can only be determined with some characteristic “uncertainties” that cannot become arbitrarily small simultaneously” (Stanford Encyclopedia of Philosophy, 2001). If probabilities and predictions are impossible or approximate in physics, it would be similarly acceptable to acknowledge the same uncertainty in regard to political science. However, if the challenge is not to be entirely predictive but to be at least similarly predictive to the degree that natural sciences are, however approximate their predictions, we will need other lines of defence.

Thus, we turn to the second argument: political science has become a lot more forthcoming in terms of its relationship with uncertainty. This openness and acceptance of uncertainty has been manifested in its purported effort to make political science more “sciency” by incorporating statistical methods, analysis of large scale data and game theoretical methodology, among others. By doing so, political science has avoided being an all or nothing science, in the sense that events either happen or not or that causes either impact outcomes or not. The discipline has made real progress in reporting uncertainty both in terms of error terms, coefficients, confidence intervals, causal weights and R2. “R-squared (R2) is a useful measure of uncertainty because it represents the proportion of the variance for a dependent variable that's explained by an independent variable or variables in a regression model… So, if the R2 of a model is 0.50, then approximately half of the observed variation can be explained by the model's inputs” (Fernando, Smith and Perez 2021). Knowledge of what portion of the outcome is explained by a model constitutes a huge reduction of uncertainty. The regression coefficient measures the strength of the correlation between two variables. It reduces uncertainty because it shows to what extent the outcome is dependent on a particular cause. It helps the research avoid the perfidy of all or nothing causal relations.

As mentioned, political science has also become more disciplined and open about uncertainty through resorting to statistical analysis, which has enabled it to take advantage of past trends through collecting and analysing data, thus determining the probability of future occurrences. The error term in regression equations reduces uncertainty because it shows what portion of the theoretical model of the phenomenon we try to explain differs from the actual empirical phenomenon. Additionally, by being able to assign probabilities to various outcomes, political science has tamed uncertainty, or at least has shown that it has made an effort to control it. By doing so, it should have won the public confidence as a value neutral discipline, whose value added is to present the world in probabilistic terms, and leave the ultimate judgments to politicians who will make decisions in explicitly and openly value-laden terms. This effort has somewhat backfired, as I conjecture later, but it nevertheless has proved that political science has become more systematic about reporting and assessing uncertainty.

The third line of defence of political science is to discount the value and feasibility of prediction. Prediction could be even conceived as an illegitimate goal: “The notion of prediction (which might be neutral in the natural sciences) here generates the wrong—indeed, an illegitimate—expectation” (Hay 2017, p. 184). A “perfect prediction” is something we know to be impossible in a probabilistic world (or even in a deterministic world where we know we shall never have knowledge and measures of every possible influence) (King 1991, p. 1,048). What would be the grounds for not expecting political science to predict phenomena in the same way as natural sciences do? Why would an equivalence between the natural and human world be false? Chapter two discusses several reasons to treat predictions related to humans as infeasible: imagination, the lack of imagination (Black Swans), social coordination and doubt in other people’s sympathy.