ANALYTICAL ESSAY Fallacies of Democratic State-Building

This paper criticizes the epistemic foundations of democratic state-building, which are derived from a model of political transitions according to which liberal democratic institutions will transform a hitherto authori-tarian and troubled country into a more prosperous and stable society and, therefore, foreign interventions to establish these institutions are realistic and worthy investments, provided they are properly planned based on knowledge of what has worked elsewhere. This expectation is based upon two epistemological premises. The ﬁrst premise, linearity , is that social and institutional change exhibits identiﬁable input–output relations connecting socioeconomic conditions and outcomes. The second premise, ergod-icity , is that these relations, inferred from past samples, provide reliable probabilistic projections about future outcomes, which can guide the focus of policy interventions. Drawing from the study of complex systems, the paper indicates why these two premises offer a ﬂawed conception of political transitions and why radical and large-scale interventions, such as state-building, will tend to generate unintended consequences rather than the planned effect.


Introduction
Democratic state-building refers to strategic actions by a foreign power to establish a stable democratic system in another country by giving the national authority instructions , prescriptions , and support to create and consolidate liberal democratic institutions and by closely monitoring progress with various sanctioning or incentivizing instruments (cf.Hippler 2005 ;Patterson 2014 ).This is by far the most interventionist form of democracy promotion, a strategy that has been an important component of the foreign policy of the United States ( Ikenberry 2011Ikenberry , 2020 ; ;Jahn 2007b , 215).
In this paper, I discuss the following question.Is democratic state-building a realistic strategy?Is it plausible to expect that an exogenous intervention into a systemprompted, designed, and organized by a foreign power or coalition-can generate such a large-scale, system-level outcome?I explain why the answer is negative.I do so by scrutinizing the epistemological foundations of democratic state-building.First, it is the idea that there are linear input-output relationships connecting variables measuring socioeconomic conditions with outcomes.Second, these relationships can be inferred by statistical analysis of past samples and be interpreted as probabilistic statements about how other systems will behave too if they come to exhibit similar values in the relevant variables.Together, they have sustained a view of political transitions according to which there are qualities and properties supportive of a stable and prosperous democratic society, which are described in probabilistic statements connecting inputs and outputs, and the human environment is seen as a terrain that can be reorganized by planned interventions, guided by that type of linear and probabilistic analysis, to target and reconfigure these properties.
I start by locating the origins of this linear paradigm of political transitions in the modernization theory and new institutionalism.Modernization sustained the belief that there are socioeconomic variables conducive to democratization and economic development, meaning that countries that will develop these variables will most probably follow a similar path.New institutionalism singled out institutional reforms as the way by which political actors can speed up a country's pathway to democracy and prosperity, by incentivizing and directing collective behavior toward the attainment of these favorable conditions.A plausible policy goal is to establish "good" institutions that will exert a structuring effect on how a society evolves, pushing its members to behaviors supporting a prosperous and stable democratic society.Lamentably, experts and policy practitioners have witnessed that planned interventions generated serious unintended consequences whose type and scale were not predicted beforehand, as the developments in Iraq and Afghanistan demonstrate.
I then juxtapose the linear transitions paradigm with a complex transitions framework that understands social change as a nonlinear and non-probabilistic process.Insights from the study of complex systems challenge the epistemological foundations of democratic state-building and better explain the occurrence of serious unintended consequences from planned large-scale interventions.In human systems, change emerges from continuous interactions of social actors that generate novel and discontinuous behavioral patterns as well as differentiated trajectories across systems.Understanding this behavioral dimension of complexity sheds light on how human systems evolve, often intermittently and discontinuously, and regarding democratic state-building, explains why, rather than having the envisaged effect, foreign interventions are more likely to unsettle a local system without giving it a desired new direction, and why, even after new institutions have been established, this will not structure collective behavior predictably.Large-scale and radical interventions, such as military invasion and state-building, will tend to trigger varied and random sequences of behavioral responses, increasing uncertainty in a system rather than reducing it.

The Epistemic Dimension of Democratic State-Building: Is This Strategy Plausible?
Democracy promotion and democratic state-building are policies whose strategic merit and idealism have been thoroughly debated and problematized in the literature ( Cox, Ikenberry, and Inoguchi 2000 ;Ikenberry 2000 ;Monten 2005 ;Jahn 2007; Krasner and Risse 2014 ;Ikenberry 2018 ;Mearsheimer 2019 ;Krasner 2021 ;Lascurettes 2020 ).Couched in the language of liberal rights and democratic norms, democracy promotion seems to reflect the democratic ethos that has given the United States a higher moral ground against its authoritarian adversaries.Efforts to radically transform the political systems in countries such as Afghanistan, Iraq, Kosovo, Libya, and Liberia have been justified with reference to a strong ideological commitment to promoting democracy across the world (Jahn 2007).The United States, confident in its superior military capabilities, uses this power for regime change ( Seitz 2017 ), which involves intervening to restructure whole political systems by redesigning political and economic institutions to embed liberal democratic norms and values in a country (cf.Ikenberry 2020 ; Lawless 2020 ).On the other hand, democratic state-building has been criticized for serving as an instrument of the United States' hegemonic project ( Robinson 1997 ), and inconsistently so, if one considers the fact that the United States repeatedly and strategically supported allied authoritarian regimes and violent authoritarian guerrilla groups (cf.

Dufek and Mochtak 2019 ).
There has been a vibrant discussion regarding the challenges and problems of implementing democratic state-building plans in actual contexts.The literature has pointed to strategic and policy-related mistakes and has debated the value democracy promotion and state-building offers to the United States ( Finkel, Perez-Linan, and Seligson 2007 ;Bridoux 2013 ;Cox, Lynch, and Bouchet 2013 ;Krasner 2021 ).Scholars identified deficiencies in the design and implementation of these interventions, such as the neglect to give due consideration to domestic processes of regime transition, and problems such as lack of domestic democratic support ( Harberson 1998 ;Dalacoura 2005 ;Craner and Wollack 2009 ;Jamal 2012 ), the neglect to consider the effect of the broader international system ( Jahn 2012 ), misinterpretations of how capitalism and democracy work together in a normative model (Kurki 2014), the high fiscal cost and the logistical and time commitments of this strategy ( Etzioni 2004 ), the misdirection of efforts toward strengthening elections and executive power at the expense of other important aspects of a functioning democracy ( Carothers 2007( Carothers , 2009 ; Meyerrose 2020 ), and the failure to grasp how democracy, contestability, and state capacity develop or unravel together and, consequently, whether an intervention would strengthen human security or would increase the risk which societies face when they undergo such intervention ( Kurki 2010 ;Norris 2012 ).This debate has hosted views from different theoretical standpoints in the discipline of International Relations and provided diverse interpretations of how structural conditions and institutions affect collective behavior and how strategies aiming at building an international order can be properly formulated ( Haggard 1991 , 404;Wight 2006 , 91).
An important but rather overlooked question is why democratic state-building has come to be seen as a feasible project in the first place.Planning for democracy involves using "the widest range of actions that one state can take to influence the political development of another toward greater democratization" ( Bouchet 2015 , 4; see also Dursun 2012 ;Donno 2013 ).Democratic state-building is the most radical form of intervention in a political system.It aims at regime change and the consolidation of a new political order that will redirect the future behavior of the whole system toward political stability and economic prosperity.The ambition of radically reorganizing a political society is seen as a realistic project that a strong reformist power such as the United States can perform on a global scale.For Ikenberry: … liberal internationalism offered a vision of a reformed and managed western -and, eventually, global -order that would provide the organizational principles, institutions and capacities to negotiate the international contingencies and dislocations that threaten the domestic pursuit of liberal democracy ( Ikenberry 2018 , 13).
A similar aspiration for political reform has been considered in the context of peacebuilding efforts.Organizations, such as the United Nations (UN), foreign countries, and the expert community have been exploring ways to successfully combine peacekeeping operations with creating stable state institutions for lastingly peaceful and good-quality postwar governance ( Sambanis 2008 ).In the liberal peacebuilding model, peacekeeping is expected to benefit from the creation of liberal democratic governing systems and market-oriented economies ( Paris 2010 ; Call and Cousens 2008 ).However, as Krasner and Risse (2014 , 558) point out, the chances of success for these efforts depend not only on the legitimacy and the design of the intervention per se but also on the degree of complexity they will encounter, which is rather significant when a whole system is supposed to be redesigned and multiple actors will be affected.Lake and Farris noted that foreign interventions aiming at comprehensive state-building almost never succeed ( Lake and Farris 2014 ).
Why do international players maintain confidence in the capacity of such largescale interventions to successfully steer whole societies in a desired direction?Whether state-building is seen as a realistic ambition largely depends on the conceptual image policymakers have developed regarding what is feasible to achieve by means of political design and intervention.This also relates to the design of a grand strategy.As Brooks (2012) notes, a grand strategy is "the overarching concept that links ends, ways and means, the organizing principle that allows states to purposively plan and prioritize the use of 'all instruments of national power,' diplomatic, economic, cultural, and military."To that end, policymakers receive feedback from the epistemic community regarding the plausibility , not just the desirability , of a strategy.This feedback presents them with a linkage connecting policy instruments with desired political ends .
In that regard, democratic state-building looks like a plausible project if the development of stable democratic systems is associated with certain socioeconomic con-ditions, positive or negative, and there is the expectation that a considerable change of these conditions, now or in the future, will redirect the development of a political system in that direction accordingly.Previous statistical findings have shown a positive connection between institutional and societal conditions with prosperous and stable democratic systems.Once these associations are understood as generally relevant "inputs and outputs" relationships and are taken as probabilistic statements about how other social systems are likely to behave if they come to develop these conditions at similar measurements, it makes sense to consider creating a plan to interfere in a country to attain the same inputs, for instance, the right kind of institutions and norms, infrastructure, financial aid and technical assistance, open access to trade and investment, etc., as a way to initiate and accelerate a process of democratic transformation.
Broadly speaking, democratic state-building relies on implicit ontological and epistemological premises-regarding what the world is made up of and how we can come to learn about it.Ontological and epistemological premises form the basic epistemic foundations behind any analysis of natural and social phenomena.These premises merge into epistemic policy paradigms , namely frameworks of understanding that come to shape how we make sense of the world.As Wight put it, different epistemological differences and theoretical debates are embedded within, and dependent upon, prior ontological positions ( Wight 2006 , 6) and, consequently, '[p]olitics is the terrain of competing ontologies.Politics is about competing visions of how the world is and how it should be' ( Wight 2006 , 2).At times, a dominant epistemic paradigm emerges from the recurrent prominence of theories with set ontological and epistemological premises.This dominant paradigm is embedded in public policy and the routines of bureaucratic organizations ( Drezner 2001 , 60).It can be broad enough to host debates involving different theories and competing ideologies and may withstand criticisms launched by other "non-mainstream" approaches, for instance, critical and post-positivist theories, to the extent that it keeps providing the basic context for popular theories to develop and for the bulk of research to take place, both of which come to shape what political actors understand they can achieve.

Modernization and Institutionalism: The "Linear Transitions" Model
The epistemological foundations of democratic state-building can be traced in the development and popularity of modernization theory and new institutionalism, which were influential in shaping public policy perceptions of social and political change in the United States.Modernization studies sought to infer the socioeconomic conditions favorable for democratic transitions.The institutionalist theories that were later developed promoted the idea that institutions could play a catalytical role in fostering these conditions and that liberal democratic institutions could condition and accelerate a society's pathway to prosperity and stability.
Work within modernization theory relied on statistical (regression) analysis to identify preconditions for successful (consolidated) democratization and embarked on understanding how socioeconomic change and institutional change are related ( Neubauer 1967 ;Lenski and Lenski 1974 ;Bollen and Jackman 1985 ;Gonick and Rosh 1988 ;Sirowy and Inkeles 1990 ;Bresser-Pereira et al. 1993 ;Londregan and Poole 1996 ;Barro 1999 ;Inglehart and Welzel 2009 ;Miller 2012 ).The pivotal contributions by Lipset (1959Lipset ( , 1960 ) ), Lerner (1958) , and Pye and Verba (1965) were epitomized in the phrase "the more well-to-do a nation, the greater the chances that it will sustain democracy" ( Lipset 1959 , 75).This body of research has promoted a view of social change as structured by societal conditions treated as variables and connected in a dependent-independent relationship of inputs and outputs.Independent variables are interpreted as preconditions for social change.These correlations then ascertain how increases in various measures of quantifiable observations could foretell the emergence and consolidation of democracy.These findings are thus understood as indicators of future change-at least regarding what is likely to happen-and they provide clues to policymakers-at least regarding what they should expect and, probably, what they can accomplish.
Postwar studies of the modernization tradition dominated academic discussions of state development ( Krasner 2011 , 66) and were influential in the design of US foreign policy ( Latham 2000 ;Jahn 2012 ) supporting the promotion of democratic reforms and economic reforms across the world.They have also been backed by statistical analysis relevant for foreign policy that found that actual democracies and market economies generally had peaceful relationships with one another ( Rummel 1995 ;Russett and Ray 1997 ).As Latham noted: … the concept of modernization was much more than an academic model.It was also a means of understanding the process of global change and the identifying ways the United States could accelerate, channel, and direct it.( Latham 2000 , 1) At the same time, scholarly work examined the role that political agency plays in prompting or directing regime transitions and the consolidation of democratic institutions (cf.O'Donnell and Schmitter 1986 ;Pempel 1990 ;Rueschemeyer and Stephens 1991 ;Przeworski and Limongi 1997 ;Diamond 2002 ;Croissant 2004 ;Merkel 2004 ;Levitsky and Way 2010 ).Attention was paid to the role of institutions in structuring strategic behavior and to the contribution of political agency to institution-building.
In economics, new institutional economics (NIE) placed emphasis on the idea that "good" institutions, once established and routinized, promote economic growth and can sustain a path to democracy and prosperity (most notably, North 1981 ;Olson 1993 ;cf. Acemoglu andRobinson 2005 , 2012 developing Douglass North's theory).Institutions are presented as sets of constraints and payoffs that limit the choices of relevant actors and direct them into specific behavioral patterns.This position brought attention to the role of institutional design in fostering positive outcomes in the shorter and longer term.Research in political science sought to understand what political agency and strategy can achieve in processes of democratization ( O'Donnell and Schmitter 1986 ;Przeworski 1991 ;Burton, Gunther, and Higley 1992 ;McFaul 2002 , 225) and brought the role of institutional design center stage next to the established variables-based understanding of regularity and change in societies ( Schneider and Schmitter 2004 ;Levitsky and Murillo 2009 ).
The ensuing "institutionalist turn" in the study of development and democratization promoted the belief that reforming formal institutions and rules will have an anticipated positive impact on collective behavior elsewhere, directing policymakers to develop the right type of norms and values supportive of liberal democracy and conducive to economic prosperity.Technocratic experts have the task of enacting institutional reforms for countries to trigger convergence toward the optimal outcomes of democracy and market capitalism ( North 2005 , 158, 164).The belief is that, because institutions incentivize, condition, and constrain behavior, the redesign of political institutions and norms will steer collective behavior in a desired trajectory.On closer scrutiny, modernization and institutionalism create these expectations on the basis of two epistemological premises: 1. Linearity, meaning that the conditions and actions associated with certain observed outcomes are described as input-output relations.Socioeconomic conditions are represented in statistical analysis as measurable variables whose behavior is often studied by "holding all things constant" so that meaningful correlations between variables or even cause and effect relations are identified.In this linear view of political and social change, relationships between macro-level variables, such as economic growth and democracy, adequately explain a society's performance.In models and games, a system's dynamics depend on the structural and institutional properties of the system that define options and payoffs and affect how behavior is constrained and incentivized.In both statistical analysis and models, relations between the constituent parts of a system are supposed to adequately reveal how the whole is constituted.The assumption of linearity justifies efforts to identify universal laws of social behavior and build models to explain institutional convergence.2. Ergodicity, which refers to the premise that the future behavior of a system can be predicted with reference to corresponding properties and relationships observed in previous samples under study.An ergodic system is a system whose properties "can be fully described by a set of unchanging conditional probability distribution functions" ( Davidson 1996 , 479) or, in other words, "knowledge about the future essentially involves projecting statistical averages based on past and/or current realizations to forthcoming events" ( Davidson 1989 , 477-78).This assumption allows statistical inferences from a sample to be taken as probabilistic statements about the future ( North 2005 , 19).The inferred relationships allow us to ascertain which future behavior is probable under the presence of the same structural and institutional variables.Any future cases that will exhibit similar values in the pertinent variables are likely to demonstrate similar behavior.
Although real world systems will face a residual level of uncertainty from random events, the projected statistical probability is revealed by prior measurements and holds as a general expectation ( North 2005 , 22).These findings translated into reliable predictions about the future can support policy prescriptions (cf.North 2005 , 163).
Practically speaking, this means that countries with poor performance in terms of economic growth, democracy and political stability and overall prosperity are trapped in their suboptimal situation.Statistical findings from previous samples are interpreted as establishing a reliable probability of how these systems will continue to perform in the future unless something is done to change the relevant measurements associated with this poor performance.At the same time, any preconditions and behavioral models that research has derived from past samples are supposed to reveal the type of changes a plan of action should aim at.The premise of linearity suggests that transformations of whole systems can be accelerated by institutional reforms that will foster favorable "input" conditions (cf.March and Olsen 1998 ;Ruggie 1998a ).The assumption of ergodicity suggests that no matter how messy and noisy the local environment is, policymakers can utilize the above-mentioned knowledge and create plausible plans for its transformation.A plan for intervention will not control all uncertainties of implementation.Interventions must be properly adjusted to the particularities of each environment: 1. Local conditions are considered, but they are seen as "noise" that does not abstract from the generalized expectations regarding inputs and outputs that prior analysis has established.2. Unintended effects can be expected, but they are seen as equally tractable by the analysis of past observations and, therefore, preventable or at least mitigatable through careful design and better implementation that take this information into account.
This type of thinking builds a "linear transitions" paradigm.A view of world as structured by measurable input-output relations is coupled with a mechanical model of how change is propelled when these measurements change.This view nurtures an engineering approach to political change: understanding these relationships allows political actors to design and implement a comprehensive package of interventions that will add the relevant inputs or strengthen the structuring properties in a society to achieve or safeguard a desired outcome.This strengthens the belief that institution building is a plausible mission for policymaking, and that foreign intervention is both desirable and plausible because a country trapped in adverse conditions is less likely to change course by itself unless an exogenous force moves it from its suboptimal position (from a suboptimal equilibrium).

The Linear Transitions Paradigm in Democracy Promotion and State-Building
It is important to note that the linear transitions paradigm accommodates various competing theories and approaches (cf.March and Olsen 1998 ).From a constructivist or sociological perspective, social norms and cultural characteristics are seen as having a "structuring" or "conditioning" effect on how a society evolves ( Finnemore and Sikking 1998 ).Norms, values, and cultural traits are constitutive of structures and actors ( Wendt 1987 ;Ruggie 1998a , 871-77).Hence, under the premises of linearity and ergodicity, institution-building can be seen as an effort by political agency to reconfigure the socialization context , with interventions promoting norms and institutions to reshape beliefs and identities.Ultimately, this effort could alter collective behavior and could influence system-level performance.
On the other side, in rational choice models and game theory , behavioral models represent foreign policy actors as strategizing to maximize a defined utility expectation by establishing new sets of institutional and normative opportunities and constraints.The focus is on what incentivizes or constrains the behavior of state actors ( Snidal 1985 ). Game theory models, for instance, were used in analyses of conflict, such as the Cuban Missile Crisis ( Myerson 1991 ).Importantly, the very setup of a model suggests which parts of the model's specification are relevant as key elements of the structure.In that regard, if the model's parameters can be altered, the expected behavior of the model will be different, and this includes reforming institutions as key incentives and constraints on behavior.
In short, the linear transitions paradigm enables different theories and approaches to treat institutions and norms as features of regularity and consider them the closest point policymakers could get to bring some degree of predictability and stability in society ( North 2005 , 36, 52).Institution-building is expected to reconfigure the connections between ideologies, constitutive rules, values, and identities with patterns of human behavior ( March and Olsen 1998 ;Ruggie 1998b ).
Likewise, the linear transitions paradigm withstood diverse political debates and ideological differences in US foreign policy.For realists, democracy promotion and state-building can be seen as having an instrumental value for a great power such as the United States to influence how other states will behave in the future by placing them inside a new structure of constraints and norms (cf.Kitchen 2010 , 141-43).Establishing democratic institutions and norms makes sense if this will change the payoffs and options for other international players ( Schimmelfennig, Engert, and Knobel 2003 ).The challenge is for policymakers to choose the form of leverage that is more effective in each situation ( Levitsky and Way 2006 , 382) and to understand the broader context of international and domestic conditions that may facilitate or hinder this decision (cf.Mearsheimer 2019 ).On the other side, from a more idealistic perspective, norms, values, and new cultural habits and routines can reshape identities and preferences elsewhere, going far beyond what rational constraints do in the short term by building the normative foundations of a safer and more prosperous world.Associations between norms, institutions, and cultural and behavioral traits appear to be relevant for building a stable international system and have come to influence what policymakers thought they could achieve ( Yee 1996 , 69;Finnemore and Sikking 1998 , 891;Wertheim 2020 ).As Ruggie put it: …Kissinger now looks for salvation to the 'idealism' that he spent his entire career mocking but which is more properly described as the animating ideas and values that emerge out of America's own sense of self as a nation and which have always framed successful U.S. foreign policy… In summary, "making history" in the new era is a matter not merely of defending the national interest but of defining it, nor merely enacting stable preferences but constructing them.(Ruggie 1998, 878).
The collapse of the Soviet Union seemed to have vindicated the belief that "good" institutions matter for prosperity and stability.The power vacuum that this major event created in East Europe and Central Asia was seen as a unique opportunity for large-scale institutional engineering.Earlier cases of successful democratic state-building in postwar Western Germany and Japan, where the United States played a key role in drafting their constitutions and rearranging power relations, set an attractive precedent.There was renewed confidence in the belief that similar projects could equally succeed elsewhere ( Smith 1994 ).Emerging and reforming economies, such as Russia, China, and Vietnam, were invited to join global institutions, such as the World Trade Organization (WTO), and enter the dense Western network of economic relations with the expectation that their domestic politics will undergo positive change toward political and economic liberalization.
In US foreign policy, this conviction has nurtured the ambitious idea that, whenever the opportunity arises, if it is in the interests of the United States, the superpower can promote or even impose a carefully designed program of institutional reform to turn a hitherto troubled country into a trustworthy member of the liberal democratic alliance of nations it presides over.Culminating with the "end of history"' vision in the early 1990s, this type of thinking provided a foil for aggressive and ambitious military interventions with the goal of integrating countries such as Iraq, Afghanistan, and Libya to the broader family of rule-of-law democracies.Experts and think tanks were invited to prepare proposals for the proper content and sequence of reforms and they did so informed by prior observations of positive change ( Byman 2003 , 50, 70;Dawisha and Dawisha 2003 , 48;Lawson 2003 , 207).
The same type of thinking is discernible in other forms of institutional efforts to reform by way of foreign intervention, such as the establishment of rule-of-law judicial and legal systems or the promotion of economic reform packages ( Krause and Jütersonke 2005 , 451).Designing these interventions tends to be informed by what was observed to have worked well in other cases.For example, in the postcommunist transitions during the 1990s, it was a strong belief that economic liberalization fostered democratization ( Bunce 2001 ), including "shock therapy" reforms in that area and beyond.The case for economic reforms was then packaged into the discourse of democratization.As President George Bush Jr put it: America will encourage the advancement of democracy and economic openness in both nations, because these are the best foundations for domestic stability and international order.( Bush 2002 ) Finally, the United States will use this moment of opportunity to extend the benefits of freedom across the globe.We will actively work to bring the hope of democracy, development, free markets, and free trade to every corner of the world.( Bush 2002 ) Downloaded from https://academic.oup.com/isr/article/24/4/viac053/6765274 by guest on 26 October 2022 This linear transitions paradigm survived policy failures, mishaps, and successions in the US administration ( Jahn 2007a , 88).Foreign policy adaptations conformed to this organizing paradigm, such as the stance adopted by President Barak Obama: After World War II, America had the wisdom to shape institutions to keep the peace and support human progress -from NATO and the United Nations to the World Bank and IMF… Now, just as the world has changed, this architecture must change as well.At the height of the Cold War, President Kennedy spoke about the need for a peace based upon, "a gradual evolution in human institutions."And evolving these international institutions to meet the demands of today must be a critical part of American leadership.( Obama 2014 ) Following the linear transitions paradigm , designing, promoting and establishing democratic institutions have been seen as a noble, feasible but demanding task.Institution-building starts with the assessment of where a country is on the developmental and institutional continuum.There are objective measures and indexes linking policies or institutions to the actual properties and performances of the local system.Knowing the input-output relationships and the special circumstances of the case we are dealing with gives us the indications of what must be done.Second, to identify the best policy option, we match our knowledge about causes and outcomes with knowledge of the local circumstances.After the old regime is wiped out and resistance is contained, the field is open for a process of institutional reform to start.The next step requires careful planning for the consistent implementation of reforms, and this plan must be tailored to addressing any special local conditions.
It is important to mention that the premises of linearity and ergodicity do not suggest that policy interventions have deterministic and certain outcomes.Under the ergodic assumption, there are probabilistic statements about what can happen elsewhere.Knowing which conditions generally favor democratization is interpreted as a challenge for international agency to devise institutional reforms against specific and often adverse local conditions ( Gourevitch 1978 ;Pridham, Herring, and Sanford 1994 ;Carothers 1997Carothers , 2000 ; ;Burnell 2000 ;Crawford 2001 ;Ethier 2003 ;Gershman and Allen 2006 ).This requires due consideration of any unaccounted local factors but does not preclude the design of institutional reform guided by probabilistic expectations.In countries that have been trapped in chronic instability or tyranny, reforming political institutions is understood to be a delicate and difficult task but necessary for directing them toward growth and prosperity.The aim is to activate the institutional and socioeconomic variables found to support democratization and political stability elsewhere.The idea is that, if successful institutional programs are studied carefully, lessons can be learnt, and analogies can be discerned that will inform interventions in other locations ( Dobbins et al. 2003 ).Interventions must be well-designed and well-executed with good knowledge of the local circumstances ( Dobbins 2007 ;Mines 2020 ).Finally, institution building is expected to have a long-term and slower-burning effect too ( Pridham 1991 ).
In that regard, the ergodic and linear view of change offers a narrower understanding of how the local context matters .Local circumstances can obstruct the process of institution-building ( Barany and Moser 2009 ;Beisheim et al. 2014 ), but understanding their effect is still filtered through prevalent expectations about how key factors of change have worked in other cases ( Ciprut 2008 ).The main concern is about the right adjustment and "sequencing" of the interventions ( Mansfield and Snyder 2007 ).

Behavioral Complexity and the "Complex Transitions" Paradigm
The interventions in Iraq, Libya, and Afghanistan generated adverse and disrupting developments such as resurgence and the rise of ISIS in Northern Iraq spilling over in Syria, opportunities for Iran to meddle with the internal affairs of Iraq, tensions and conflict between Turkey and the Kurds, Russian interference and influence in the Middle East, a long-standing civil war in Libya, and the influx of refugees in Europe from the broader region with several destabilizing effects on European politics.Similarly, smaller-scale unintended consequences were also identified in various peace-keeping missions, some of which had policies of political and economic liberalization integrated in the management of the postconflict situation (Paris 2004(Paris , 2010 ; ;Call and Cousens 2008 ).
Recurrent failure raises a more profound question concerning how we evaluate prior knowledge in view of differentiated results and contexts ( Ungerer 2012 ) and come to hold expectations about future outcomes.Can we really reduce systemlevel dynamics to a few inputs observed in past statistical samples?Is it plausible to look for guiding evidence from statistical inferences as if they were "laws" regarding social phenomena (cf.Karl 1990 )? The problem with policy and theory conforming to the premises of linearity and ergodicity is that they overlook how, in each context, relationships and interactions among social actors tend to produce new properties in a system-novelty-and, consequently, a great degree of uncertainty over previously calculated probabilities (cf.Knight 1921 ).
Societies are complex systems in the sense that they have properties, such as norms, values, institutions, and patterns of behavior, that continuously emerge from interactions among social actors and their organizations within and across interconnected systems ( Root 2013 ).They exhibit intractable dynamics that tend to generate novel and unpredictable patterns of change.In the words of Peter Hall, the coevolution of actions, social relations, and institutions takes place in a "social ecology" ( Hall 2010 ).In a social ecology, human actors, far more than any other biological being, consciously readjust their behavior by reflecting on prior and incoming information that they receive through interactions with others.Human systems are in that sense complex civil societies ( Novak 2018 ).People learn from interaction with others and send their feedback through this interaction; they thereby revise their preferences and reorganize collective activities to change properties of their environment in line with revised preferences (cf.Dosi et al. 2003 ;North 2005 , viii).They do so while interconnected in networks across different systems ( Root 2013 , 14-15).Root called these systems complex adaptive systems ( Root 2013 ).Each network has a distinct structure shaping how information is channeled, who controls it, or who has greater influence on it, defining different structures of commands and communication ( Root 2020 ).In complex systems, any system-level outcomes are almost always not the direct result of intentional action but rather the outcome of interactions of myriads of actors whose actions and plans generate an ecology of behavioral adaptations across time.
To better illustrate how complexity captures processes of social change, we can consider how the cultural context evolves and intersects with social change.A strand of studies of democracy and democratization studied the role of civic culture, consisting of combinations of beliefs, values, and norms and diverse forms of civil society, in strengthening the prospect of democratization ( Almond and Verba 1963 ;Pye and Verba 1965 ;Weigle and Butterfield 1992 ;Putnam 1993 ).For Inglehart and Welzel (2009) , there is a set of values and norms relevant to the flourishing of a democracy, broadly defined as modernity and modernization, which can have a slower but significant effect on public attitudes.Yet, the idea that there are specific cultural environments hostile to democracy was disproven on many cases ( Kacowicz and Miller 2018 , 305).Moreover, a cultural environment is far from a static set of properties ( Krause and Jütersonke 2005 , 453).Cultural values are coevolving through interactions in each society, often unexpectedly, as was observed in evaluations of international aid programs ( Barnett and Bennett 2013 ).This casts doubt on how stable and reliable the structuring effect of cultural traits on society is ( Xenia 2005 ).Observing changes in attitudes and behavior in each setting suggests that the relationship between cultural factors and political institutions and norms is dynamic and complex, rather than linear and ergodic: cultural traits affect individual behavior, but they are also sustained or challenged by collective behavior (cf.Almond and Verba 1989 , 29).This is far from suggesting that exogenous interventions can reshape existing cultural traits at will.Patterns of behavior and sets of beliefs do indeed change, but not through clearly distinguishable linear inputoutput patterns that can be manipulated.
Here, interactions are a fundamental aspect of how patterns of collective behavior emerge.Our own preferences, identities, and, consequently, behavior are being shaped by existing norms and values as well as evolving feedback processes with others.Some of us may want to change a few circumstances or our broader environment, but the evolution of these circumstances depends on the interactions of multiple other actors whose aspirations and plans may clash, assist, or disrupt our own plans and aspirations.
This means that what we see as a social context at any given time is not the product of additive inputs of the parts of a system but, instead, the whole is constituted and reshaped by how multiple factors, by interacting with one another, reshape the properties of the system, and will continue to do so indefinitely, often unconsciously or with unintended consequences (cf.Lewis 2010 ).These interactions will differ from context to context leading to diverse and often divergent trajectories of change across societies.Hence, future developments cannot be predicted by the statistical analysis of other samples from the past taken as probability statements.What is more, both domestic and international events include several systems with multilevel connectivity ( Root 2020 ).These systems interact at multiple levels (i.e., markets, institutions, organizations, domestic groups, etc.), and actions at one scale impact on behavior at another, adding complexity and uncertainty to the system.Hence, complex processes of change in human societies cannot be adequately understood through the scheme of linear cause-and-effect relations and cannot be interpreted under the premise of ergodicity as probabilities (cf.North 2005 , 22).
Furthermore, time is a fundamental dimension in terms of how societal interactions evolve.The socializing, informational, creative, and entrepreneurial aspects of human agency continuously generate changes in social systems.Social change is occurring with differences in its "duration," "tempo," "acceleration," and "timing" ( Grzymala- Busse 2011 ).Change can occur as "gradual transformations" or can take the form of "breakdown and replacement" ( Streeck and Thelen 2005 , 9;Thelen 1999 ;Pierson 2004 ;Hall 2010 ).This often makes the boundaries between institutional continuity and change porous.Pierson (2000) describes path dependency as the self-reinforcing process by which institutions and embedded norms promote stability in systems of political and social organization but also discerns slow-moving processes that can be "big" in terms of the impact they have ( Pierson 2004 ).On the one hand, a degree of institutional stability can be attributed to patterns of "reproduction by adaptation," which may also host gradual and stealthy transformations described as "layering," "conversion," and "drift" ( Thelen and Steinmo 1992 ;Thelen and Streeck 2005 ).On the other hand, behavioral adaptations can suddenly escalate to critical junctures that open the possibility for radical institutional change ( Capoccia and Kelemen 2007 ).Sudden transitions can be triggered even by what initially seems to be an insignificant event, as was the case with the revolutions in the Arab world in the early 2010s and the fall of communism in Eastern Europe in 1989.In that regard, any regularities observed in a social system at any given time are not planned; they are temporary and always incomplete.
Behavioral complexity challenges the idea that collective behavior is structured by institutions with a good degree of regularity, because it questions the very idea of institutional order acting as a relatively stable and reliable container of societal change.Institutional change, often taking the form of discontinuities in the interpretation of the same norms and drifts in the application of these rules, is, instead, contingent on societal interactions in which evolving preferences, ideas, strategies, and actions matter.This illustrates the limitations of thinking in terms of inputoutput expectations.In complex systems, the effect of a planned intervention is far from linear and predictable.It is impossible to predict the full range of responses to the intervention and how this will unfold in the future.This also explains why a forceful exogenous intervention in a system will trigger behavioral responses that will most likely move that system in an unexpected direction A foreign intervention, even if followed by a strong military presence, will upset processes by which a modus vivendi has been so far established.It will even unsettle endogenous processes that could have unleashed a political transition at some point in the near future.By abolishing or severely disrupting previous institutional structures, practices, and routines while initiating a process of institutional redesign, it will create new decision situations that most members of this society will find unfamiliar uncertain.New patterns of collective behavior will emerge in a dynamic and unpredictable fashion.Significant events, such as a failure of infrastructure, administrative delays, and local episodes of protest, can trigger large-scale dissatisfaction, protestations, or even violence.The initial sequence of responses will create new networked connections among social actors, further propelling a cascade of behavioral changes.These interactions can accelerate shifts in both the preferences and the strategies of social and political actors, especially as people respond to new information.The new environment may also trigger opportunistic responses by key political figures.Diverse adaptations and organizational formations will unleash new and unpredictable changes in this environment.With a domino of preference changes and behavioral readjustments, the direction of change is difficult to foretell.It may and go as far as challenging and subverting the new status quo.
It is important to note that, after an intervention, the previous properties of the institutional and normative status quo do not disappear (cf.Pierson 2000Pierson , 2004 ).Yet again, how preexisting power networks will behave cannot be predicted.Public officials can be keen to collaborate with the new political force or may decide to join the opposition or engage with resurgence forces.The latter type of response will be received as a message by other people disappointed or outraged with the consequences of the intervention itself.For instance, the removal of a dictator creates new opportunities for opportunistic behavior that can take the form of new alliances, radicalization, collective action, paramilitary action, and incendiary rhetoric whose effects cannot be predicted in advance and cannot be easily contained afterwards.Religious organizations may provide the ideological repertoire and the material infrastructure for insurgency.Other foreign governments or organizations may find a chance to interfere.Developments in one place may spill over to neighboring areas and countries, as was the case of ISIS in North Iraq and Syria, destabilizing larger regions and creating new and more urgent problems for the foreign occupying force.
Complexity suggests that the evolution of interactions and adaptations after radical large-scale interventions can be not only very consequential for a system but also unpredictable.This includes the anticipation of side effects too, those developments that severely derail the direction of a system compared to the original plan of intervention.Because behavioral adaptations are diverse and continuously generate novel properties in their environment, planning for social and political change on such a large scale confronts a higher degree of indeterminacy.An exogenous intervention will trigger behavioral adaptations and new interactions that will add uncertainty.These behavioral responses can manifest themselves as unintended and unforeseen consequences.The nature and scale of unintended consequences cannot be predicted ex ante so that preemptive and mitigating measures can be taken early enough.In Libya, for example, regime change had a surprisingly disruptive effect on intercommunity linkages within that society, unleashing opportunistic behavior from Libyan leaders and foreign powers that escalated into years of conflict.The case of Iraq is another example of an intervention that pushed several political and social actors to behavioral choices diametrically different from the one planned and envisaged-domestic terrorism, insurgency and civil war, and unwelcomed interferences by Iran and Turkey.
Understanding the behavioral dimension of complexity is an epistemological take on reality justifying moderation regarding how we read research findings from the study of social systems and interpret them.It becomes evident that we cannot obtain an analytically tractable picture of possible patterns of behavioral reactions and counterresponses that will be consequential for a given system.This curbs the aspirations of foreign policy and grand strategies.Local knowledge, no matter how comprehensive it might be, cannot help us attain the desired outcome of the intervention or predict the type and scale of unintended consequences that we need to preempt or mitigate.In democratic state-building, complexity suggests that simply adding or replacing a few key components of the system-as the result of an organized and well-executed external intervention-will not have a predictable "structuring" impact on how the targeted environment would readapt to this intervention.
This raises another question.Can interventions achieve policy objectives with a lower degree of uncertainty?What complexity indicates is that the larger and more radical the intervention is, the greater the uncertainty it faces.Task complexity matters ( Krasner and Risse 2014 ).Targeted interventions for emergency assistance, infrastructure building, and basic state capacity building may have higher chances of success, because the scale of the intervention is significantly smaller compared to the much more wide-ranging task of state building.In that regard, a grand strategy can envision a sequence of smaller, less far-reaching steps, even if it wants to promote broader system-level changes in the long term.For instance, instead of interventions to steer the behavior of large units such as states in a desired direction in relatively small periods of time, there can be efforts to approach and co-opt leaders and key political players without disrupting the system they inhabit.Jordan is an example of a country whose foreign policy has been transformed under King Abdullah after a political rapprochement with the United States.However, if there are strong indications that elites and the vast majority in a society wish to follow a democratization path as in the case of East Timor ( Krasner 2011 , 73-74), this path can be facilitated by foreign assistance and guidance.A more comprehensive plan must enjoy a higher degree of public legitimacy in that society ( Lake and Farris 2014 ) and will essentially work as a helpful and affirmative nudge rather than a violent transgression.Finally, the consistent promotion of liberal norms, ideas, and values across human societies can indeed influence identities and preferences in other societies in the longer term.The broader idea of democracy promotion does not have to be abandoned.Understanding complexity warns against radical, largescale, and abrupt interventions in processes of social change and, at the same time, points to endogenous dynamics that a step-by-step strategy can influence.

Conclusion
Democratic state-building is a controversial aspect of democracy promotion, even more so following developments in Iraq and Afghanistan.It is an effort to steer the behavior of a society in a desired direction.Such an ambitious idea will only make sense if it is regarded as a realistic project of action.This paper has criticized the epistemological assumptions that make this strategy look plausible.Democratic nation-building is based on the idea that liberal democratic institutions tend to make a political system prosperous and stable and the expectation that an intervention to establish institutions in a foreign environment will help recreate these conditions, provided it is properly planned and executed.This idea relies on a dominant "transition paradigm" built on two flawed premises regarding change in human societies: first, that processes of social and institutional change have identifiable input-output relations connecting socioeconomic conditions that can be measured as statistical variables (linearity) and, second, that observations from past samples offer a reliable probabilistic basis for planning new interventions and predicting their effect elsewhere (ergodicity).Under the premises of linearity and ergodicity, if the same variables appear in another place in similar measurements and no previously unaccounted factor comes up whose effect was not present in the sample and could not be calculated at the time of that analysis, the probabilistic projection of input and output is expected to hold.Through this lens, it appears realistic to expect that, if we know what has made democratic transitions successfully consolidated elsewhere, we can prompt and accelerate a similar process in another system, using the same knowledge to change the relevant institutions and behavioral norms as indicated by previous observations while making a few careful adjustments to local conditions.
Using insights from the study of complex systems, I explain why these assumptions are misleading and why radical foreign interventions are bound to generate large-scale unintended consequences.In human systems, responses to exogenous interventions involve sequences of multiple and intractable behavioral reactions developing through continuous feedback processes and generating novel and diverse social and political trajectories.The scale and nature of these responses in each case cannot be forecast and anticipated beforehand.These behavioral adjustments may manifest themselves as cascades of events that completely derail the original plan in very adverse directions.A political system or the whole region may experience system-level unintended and unforeseen turmoil.Although there are cases in which institution-building seems to have got it right, these success stories must not be understood as vindicating a general expectation that these cases offer clear indications of what should be done or what must be avoided to achieve a similar result elsewhere.Complexity suggests that, even if the formal institutional order is modified, the behavior of the whole system will most likely not be reconfigured as planned, particularly if the exogenous intervention is radical and abrupt.