The political ecology of authoritarian learning

By Steven Heydemann, Janet W. Ketcham Chair, Middle East Studies, Smith College

*This memo was prepared for the workshop, “Transnational Diffusion, Cooperation and Learning in the Middle East and North Africa,” held June 8-9, 2016.

Since mass protests first broke out in the Middle East in late 2010, authoritarian regimes in the Arab world have demonstrated significant resilience. Not all have survived. Some are engulfed in violent conflict. Yet in a majority of cases, the uprisings of 2011 failed to bring about either the breakdown of authoritarian regimes or transitions to some form of post-authoritarian governance. Among scholars of Arab politics – even those who view it as too soon to declare that the uprisings have failed – there is a widely shared consensus that among their many effects the “resurgence of the security state” and the reconsolidation of authoritarian governance rank among the most important and consequential for the future of the region (Lynch 2016; POMEPS 2015).

Despite this consensus, the resurgence of authoritarian governance in the Arab world has done little to resolve debates about the causes of authoritarian resilience. What enabled most authoritarian regimes in the region to survive sustained, large-scale anti-regime uprisings? Explanations for the capacity of regimes to contain, suppress, or prevent mass protests have varied widely. They have focused on the effects of regime type, the redistributive capacity of regimes, civil-military relations, the presence of cross-cutting coalitions, levels of sectarian diversity, and the extent of institutionalization among security sectors.

In addition, and of particular interest in assessing regime resilience, scholars have shown growing interest in the effects of authoritarian learning and transnational dissemination of authoritarian practices on the survival of Arab regimes (Bank and Edel 2015; Patel, Bunce, and Wolchik 2014; Heydemann and Leenders 2014). Extending the insights of literatures on the adaptive or recombinant qualities of authoritarian regimes (Heydemann and Leenders 2013), and building on comparative lessons drawn from authoritarian responses to the “color revolutions” and regime efforts to counter Western-backed democracy promotion, research programs on authoritarian learning and dissemination have begun to explore the forms, content, and mechanisms through which authoritarian regimes in the Arab Middle East upgrade their governance practices in the face of new challenges.

Such processes have been highly visible among Arab regimes, where we see longstanding evidence of convergence in the tactics and strategies they have adopted to sustain themselves (Heydemann 2007). They have been widely characterized as central to the governance repertoires that regimes developed to mitigate the effects of the 2011 uprisings. Yet even as this literature has grown, and even as the causal effects of learning and dissemination are becoming increasingly apparent, the processes through which they take place, and the causal relationships between dissemination, learning, and resilience, remain understudied and under-theorized.[1]

In particular, important efforts to identify and trace the causal pathways through which international factors reshape regime practices have outpaced attempts to untangle how imported ideas become integrated into the standard operating procedures of authoritarian institutions. Researchers have worked to identify the mechanisms that facilitate dissemination – including emulation, appropriation, socialization, and inter-elite cooperation among authoritarian epistemic networks (Heydemann 2009; Xiaoyu 2012; Levitsky 2005) – and to pinpoint distinctive domains in which learning leads to policy change (Bank and Edel 2015). But how these mechanisms generate learning—that is, lead to durable and meaningful changes in the behavior of actors and institutions responsible for the maintenance of authoritarian regimes—remains murky.[2]

We have established a number of plausible mechanisms of dissemination. We see evidence of their impact on governance practices. But how they actually work is less clear (Erdmann et al 2013, 7).[3] The links between dissemination and the institutionalization of learning are too often underdeveloped.[4]

This gap is evident in how dissemination and learning have been approached in research on authoritarianism in the Arab Middle East. Claims about the role of these processes in the maintenance of authoritarian regimes has relied heavily on observed shifts in regime behavior during periods of stress (Bank and Edelman 2015). Relevant shifts seem to include those that meet one or more of three criteria:

(1) they appear to express conscious and intentional imitation in governance practices (Bunce and Wolchik 2006);

(2) they reflect behavioral changes among key regime actors that are triggered by personal experiences (Bermeo 1992; Levy 1994 in Bank and Edel 2015);

(3) they increase convergence in governance practices among authoritarian regimes.

These criteria are useful. They take into account that dissemination is likely to increase the extent to which authoritarian regimes adopt similar governance repertoires to address similar threats and challenges. They also acknowledge that learning can lead to innovations, hybrid practices, or contextually-specific adaptations. However, these criteria do not address how newly-learned practices become institutionalized within authoritarian systems of rule to become standard operating procedures.

This working memo, using the related concepts of recombinant authoritarianism and the political ecology of authoritarian learning, is a preliminary and partial effort to address this missing link. It relies heavily but not exclusively on experiences of the Syrian regime of Bashar al-Assad in the period since the start of the Syrian uprising in March 2011. It uses these concepts to highlight the importance of principal-agent dynamics in authoritarian learning. It goes on to suggest that further research on principal-agent relations – a surprisingly neglected aspect of the learning literature – is needed to understand patterns and variation in authoritarian dissemination and learning. Through a focus on principle-agent relations, it sheds tentative light on why the top-down consolidation or institutionalization of new governance practices seems to occur more easily in some organizational domains than others, and varies across cases as well.

Dissemination and the Institutionalization of Learned Practices

For some scholars of authoritarianism (not only in the Arab Middle East), the processes through which governance practices become institutionalized are obvious and uninteresting. In authoritarian regimes decision authority is tightly held by a small number of individuals who rule by fiat, ruthlessly ensure compliance, and cultivate risk averse organizational cultures. Under authoritarianism, in other words, principal-agent problems are minimized. The introduction of new and expanded coercive practices in Syria’s security sector since 2011 serves to illustrate this view of authoritarian learning.

Coercive strategies of compliance as a means to overcome resistance to the adoption of controversial modes of repression and violence are evident in two recent accounts of the practices adopted by the Assad regime to respond to an emergent uprising in the city of Deir al-Zour located near Syria’s eastern border with Iraq (Borger 2015; Taub 2016). Regime documents acquired by the Commission for International Justice and Accountability (CIJA) provide unprecedented insight into processes of decision making and the downward dissemination of orders concerning the methods to be used in repressing protests. According to these documents, local agents of the regime’s security apparatus fed information concerning protest activities upward to a high-level security committee, the Central Crisis Management Cell (CCMC), constituted of the very top tier of trusted regime officials. The CCMC, in turn, oversaw the work of the National Security Bureau (NSB), an agency responsible for coordinating four internal security organizations as well as the Ba`th Party security apparatus, including their regional and local security committees.

As protests escalated in Deir al-Zour during the spring of 2011, the CCMC instructed local security forces to increase their use of torture, expand the range and intensity of torture techniques used against detainees, and increase the numbers of detainees by lowering the threshold of behavior warranting detention. In at least one documented case, a local security official criticized these procedural changes, expressing concern about the brutality with which new measures were being implemented. Such criticism was quickly suppressed, however, and local security forces shortly integrated and routinized new coercive procedures. Emphasizing the effects of an authoritarian bureaucratic culture on learning, the founder of the CIJA describes the Syrian security apparatus as an organizational context favoring the rapid dissemination and integration of new practices:

“It [the internal security apparatus] is highly bureaucratized . . . It generates an awful lot of paper, because it is a culture in which decision-making by subordinates is implicitly discouraged, so people are forever reporting upwards, trying to get others to take responsibility for decision making, and covering their ass. . .” (Borger 2015).

A similar account of coercive compliance in the regime’s security sector has been reported within the Syrian air force (SyAAF). According to a military analyst, the Assad regime issued orders to helicopter pilots to target civilians:

“Through 2012, reports began to circulate that the regime in Damascus had ordered all SyAAF squadron commanders to bomb civilians in insurgent-controlled areas. Damascus instructed all commanding officers at first, and then all officers in each operational unit, to acknowledge the order with their signatures.

Although a majority of SyAAF pilots at that time were Alawite  –  there were by then very few Christians, Druze and Sunnis left with the service  – the order met with strong dissent. Pilots who refused to obey the order disappeared. A few reappeared after a week or two in prison, where torture was not uncommon. Others were never seen again.”[5]

In both instances, the regime’s leadership used repression to overcome non-compliance with new rules within its security sector. Resistance to the adoption of new techniques was simply crushed. These cases highlight the absence of barriers to dissemination and the institutionalization of new practices in specific organizational contexts and under exceptional circumstances. When the institutionalization of new practices requires internal adaptations of the state bureaucracy, when it occurs within the most hierarchic of organizations – the security sector – when it concerns governance practices that are seen as necessary to ensure the survival of a regime, and when the costs of non-compliance by agents are seen by rulers as unacceptably high, the effectiveness of dissemination is likely and predictable. Adherence to new procedures may not be complete. Some agents will continue to find ways to dissent – passive resistance exists within even the most hierarchic and coercive institutions – but overall, coercive learning can, in some cases, eliminate principal-agent problems.

In other instances, however, when the consolidation of new rules or procedures is not only internal to a state institution but also requires the compliance of external actors, resolving principal-agent problems is more difficult and less predictable. Regime efforts to adapt the rules of economic governance and mitigate the impact of war on the Syrian economy have had decidedly mixed results. Regulations restricting the export of capital have not prevented capital flight. In 2011, the imposition of new rules restricting imports was resisted by business actors and quickly rescinded. More recently, subsidy cuts have led to protests in areas under regime control.

These examples illustrate the variation in the consolidation of learning among different organization or sectoral domains. Where the institutionalization of new practices requires the compliance of a large and diverse set of agents, where incentives for non-compliance are high and the costs less severe, and where coercive means for securing compliance themselves involve high costs, even an authoritarian regime in the midst of economic collapse may find it challenging to resolve principal-agent problems and secure the institutionalization of new governance practices.

Institutionalized Resistance to Learning

A second, contrasting view of authoritarian regimes in the Arab Middle East leads to very different conclusions about the ability of political leaders to disseminate and institutionalize new governance practices. From this perspective, possibilities for authoritarian learning are highly constrained. Regimes in the Arab Middle East are viewed as sclerotic, resistant to change, and locked into established practices by corrupt, clientalist organizational cultures that impede innovation and undermine the capacity of rulers to ensure the compliance of subordinates. This view posits that authoritarianism, with its lack of transparency and accountability, exacerbates principal-agent problems and works against processes of dissemination and learning.

The downward causal links needed to institutionalize and routinize new practices are present, but weak and fragmented. This perspective is perhaps most widely expressed in analyses of elite tensions that marked Egypt’s uprising from 2011-2013. When accounting for the struggles and turmoil that led to the overthrow of President Morsi in July 2013, assessments have highlighted the institutional incoherence of the Egyptian state, the diffusion of authority and control, and the ominous role of “the deep state” to explain the difficulties elected officials encountered in their efforts to implement Islamist-oriented projects of political reform, respond to popular demands for political change, and restore economic stability. As a prominent Jordanian economist observed:

“[Salafist] FJP party leaders contend that deep state networks are an obstacle to their sovereignty, while other political camps argue that the Brotherhood’s influence is becoming more pronounced in state institutions. Some groups also claim that there is no impetus to improve the performance of public institutions, to render them more responsive to the demands of the citizenry, and to address the problems the system faces . . .

Once again, the key sources of weakness in the Egyptian economy must be pinpointed. Does weakness stem from institutions so fragile that they are incapable of improvement, irrespective of who is in power? In a political framework that gives control of critical institutions to the ruling party or the covert resistance to this control? Egypt lacks institutions that may facilitate dialogue on this question, while the rival groups continue to hold each other responsible for the current situation” (Saif 2013).

What Saif describes as a rigid state with fragile, ineffective institutions – meaning, institutions in which agents can prevent principals from disseminating and consolidating new governance practices – is consistent with the view of authoritarianism in the Arab Middle East as sclerotic and resistant to change, with leaders who can rule but not govern. Following his coup against Morsi, however, General and then President Sisi succeeded in expanding and institutionalizing a broad repertoire of repressive measures that build on and significantly enhance prior practices. In responding to the threat mass politics posed to the regime, Sisi was able to achieve what Morsi could not: to mobilize the repressive, regulatory, and legislative institutions of the state in a large-scale process of authoritarian adaptation. This was accomplished, in part, through the use of coercive means to ensure compliance among agents within the state apparatus. It also mattered that Sisi was able to frame the consolidation of new practices as necessary for the survival of both state and society, and as an expression of the state’s role as guarantor of security and stability. More broadly, however, the contrast between Sisi and Morsi underscores the difficulty of generalizing about processes of learning and dissemination within authoritarian regimes in the Arab Middle East.

Both approaches – perspectives that view authoritarianism as either a source of principal-agent problems or as their solution – offer important insights into learning dynamics within authoritarian regimes. Both capture significant aspects of the organizational cultures that prevail in authoritarian regimes in the Arab Middle East. Yet neither is wholly complete or entirely satisfactory. Authoritarian regimes in the Middle East vary along any number of dimensions – regime type, resource base, demographic diversity, and state capacity. Despite these differences, however, Arab regimes cannot easily be classified as either highly centralized with tightly-coupled institutional structures that facilitate dissemination and learning or highly decentralized with loosely-coupled structures that impede diffusion and learning.

Both sets of attributes are present in every Arab regime. In fact, the “learning profiles” of Arab regimes may exhibit higher levels of “within regime” variation than of “cross regime” variation. Thus, even as documents surface revealing how norms of deference and compliance in Syria’s security sector supported the rapid dissemination of shifts in coercive practices, we find numerous accounts of the erosion of regime authority and the rise of autonomous economic and military actors in regime-held areas of the country. Even as Egypt’s deep state frustrates the ambitions of the Muslim Brotherhood, we find the rapid dissemination and adoption of governance practices that have intensified internal repression of the media, civil society, and Islamist opposition movements since 2013 – tactics evident to differing degrees in virtually every authoritarian regime that survived the 2011 uprisings.

Authoritarian Ecologies and the Challenge of Learning

If authoritarian regimes are inconsistent and uneven in their capacity to institutionalize new ideas and practices, how can we account for the impact of dissemination and learning on governance or on authoritarian resilience? Two related concepts might prove useful in attempts to theorize processes of authoritarian dissemination and learning under such conditions. One is the notion of recombinant authoritarianism, which defines authoritarian regimes as “systems of rule that possess the capacity to reorder and reconfigure existing instruments and strategies of governance, to reshape and recombine existing institutional, discursive, and regulatory arrangements to create recognizable but nonetheless distinctive solutions to shifting configurations of challenges” (Heydemann and Leenders 2013, 7). The second, intended to provide the empirical foundations needed to test claims concerning the recombinant capacity of authoritarian regimes, is based on mapping the political ecologies of regimes to identify (1) when, within what domains, and under what conditions they exhibit the attributes of either tightly-coupled systems in which principal-agent problems are minor, or loosely-coupled systems in which principal-agent problems obstruct dissemination and learning; (2) whether and under what conditions rulers are able to change the balance of attributes within distinct domains of governance, moving along a spectrum from tightly-coupled to loosely-coupled; and (3) the conditions under which dissemination and learning encounter resistance, are ineffective, or fail to become institutionalized and routinized within regimes.

The notion of a political ecology is important in this respect in its implication that regimes are not fixed at one point on this spectrum (and, indeed, may occupy more than one point at any given time). Yet it should also be used with caution to avoid the inference that dissemination and learning can somehow be seen as comparable to natural evolution. With this caution in mind, these two concepts offer a framework for both theorizing dissemination and learning and testing empirically claims concerning their effects on authoritarian resilience. As this memo has suggested, the path forward in the development of such a framework lies in mapping the political ecologies of authoritarian governance in the Arab Middle East: disaggregating processes of learning and dissemination to unpack how and under what conditions governance practices in distinct domains are imported, disseminated internally, and institutionalized. Such a framework, moreover, requires a conception of authoritarianism as a recombinant system of rule, in which the potential for adaptation is universally present but unequally distributed, and where the capacity of regimes to manage principal-agent relations stands out as a key indicator of how effectively they institutionalize new governance practices in specific organizational contexts. Further research along these lines will, I believe, help bridge the empirical and theoretical gaps that now limit our understanding of how dissemination influences processes of authoritarian learning in the Arab Middle East.


[1] In this working memo, I distinguish between learning and dissemination, on one hand, and authoritarian cooperation, on the other hand (Erdmann, Bank, and Hoffmann 2013). Not all cooperation involves learning and not all learning requires cooperation. Cooperation may facilitate learning and dissemination, but is not necessary for them to occur. See also footnote 3.

[2] Literatures on democratic learning exhibit a similar gap. In Bermeo’s (1992) work, for example, learning is presented as a process that operates at the level of individual actors. Experience affects the attitudes and beliefs of influential actors, changing their views about the desirability of democracy. In response, actors behave differently and, through their actions, affect outcomes. For durable systemic change to occur, however, individual preferences have to be institutionalized, routinized, and integrated into both bureaucratic processes and the attitudes and beliefs of publics at large.

[3] This concern does not apply to all instances of dissemination and learning. In some cases, the causal mechanisms linking dissemination to outcomes are explicit and visible. These include the dissemination of authoritarian practices through direct cooperation among governments, where we can identify the agents who transfer knowledge across borders in defined contexts for defined purposes. Iran’s involvement in upgrading the skills and competence of Syrian loyalist militias and other arms of the Assad regime’s security apparatus, and in restructuring security forces to undertake roles for which they were not previously equipped – notably urban warfare – is a prominent example of visible and traceable modes of authoritarian dissemination and learning though formal, inter-regime cooperation.

[4] One of the effects of this gap is that we have fewer tools with which to explain why dissemination and learning do not occur in some instances when authoritarian regimes face challenges for which tested and effective responses exist in global repertoires of authoritarian governance, or why learning sometimes fails (Bank and Edel 2015).

[5] Tom Cooper, “The Hind Gunship Is One of Syria’s Worst Terror Weapons: Aging Helicopters Attack Indiscriminately.”

Selected References

André Bank and Mirjam Edel (2015). “Authoritarian Regime Learning: Comparative Insights from the Arab Uprisings.” GIGA Working Papers, No. 274 (June).

Bermeo, Nancy (1992). “Democracy and the Lessons of Dictatorship.” Comparative Politics. Apr 1: 273-291.

Julian Borger (2015). “Syria’s Truth Smugglers.” The Guardian (12 May).

Valerie Bunce and Sharon Wolchik (2006). “International Diffusion and Postcommunist Electoral Revolutions.” Communist and Postcommunist Studies, 39, no. 3 (September 2006), 283-304.

Sheena Chestnut (2010). “Regime Security and Counter-Diffusion: Sources of Authoritarian Learning and Adaptation.” Democracy & Society 7.2: 5-8.

Larry Diamond (2005). “Authoritarian Learning: Lessons from the Colored Revolutions.” Brown Journal of World Affairs 12.

Gero Erdmann, André Bank, Bert Hoffman, and Thomas Richter (2013). “International Cooperation of Authoritarian Regimes: Toward a Conceptual Framework.” GIGA Working Papers, No. 229.

Evgney Finkel and Yitzhak M. Brudny (2012). “No more colour! Authoritarian regimes and colour revolutions in Eurasia.” Democratization 19.1: 1-14.

Steven Heydemann (2007). Upgrading Authoritarianism in the Arab World (Brookings Institution).

Steven Heydemann, (2009). “Authoritarian Learning and Current Trends in Authoritarian Governance,” in Oil, Globalization, and Political Reform, ed. Shibley Telhami (Brookings Institution).

Steven Heydemann and Reinoud Leenders, (2014). “Authoritarian Learning and Counterrevolution,” in The Arab Uprisings Explained, ed. Marc Lynch (Columbia U Press).

Steven Heydemann and Reinoud Leenders (2013). Middle East Authoritarianisms: Governance, Contestation, and Regime Resilience in Syria and Iran (Stanford U Press).

John G. Ikenberry (1990). “The International Spread of Privatization Policies: Inducements, Learning and Policy Bandwagoning.” The Political Economy of Public Sector Reform and Privatization 99 (1990).

Steven Levitsky (2005). “International Linkage and Democratization,” Journal of Democracy 16:3 (July), pp. 20-34.

Marc Lynch (2016). The New Arab Wars: Uprisings and Anarchy in the Middle East (Public Affairs).

David Patel, Valerie Bunce, Sharon Wolchik,(2014). “Diffusion and Demonstration,” in The Arab Uprisings Explained, ed. Marc Lynch (Columbia U Press).

POMEPS (2015). The Arab Thermidor: The Resurgence of the Security State. POMEPS Studies 11.

Maria D. Popova (2013). “Authoritarian Learning and the Politicization of Justice: The Tymoshenko Case in Context.” Available at SSRN 2274168

Ibrahim Saif (2013). “Uncertainty and Economic Crisis in Egypt.” Carnegie Endowment for International Peace (March 13).

Ben Taub (2016). “The Assad Files.” The New Yorker (April 18).

Pu Xiaoyu (2012). “Socialisation as a two-way process: Emerging powers and the diffusion of international norms.” The Chinese Journal of International Politics 5.4: 341-367.