Category Archives: State

Author Interview: Warlord Survival: The Delusion of State Building in Afghanistan, by Romain Malejacq (Radboud University)

Editor’s note:  Here CESS Blog is pleased to present the fourth installment of our author interview series for those shortlisted for the annual CESS book award in social sciences and humanities this year.  Niamatullah Ibrahimi (La Trobe University), author of both Afghanistan: Politics and Economics in a Globalising State (Routledge 2019, with William Maley), and The Hazaras and the Afghan State: Rebellion, Exclusion, and the Struggle for Recognition (Hurst Publishers 2017) interviews Romain Malejacq (Radboud University) about his book Warlord Survival: The Delusion of State Building in Afghanistan, published by Cornell University Press.

“How do warlords survive and even thrive in contexts that are explicitly set up to undermine them? How do they rise after each fall? Warlord Survival answers these questions. Drawing on hundreds of in-depth interviews in Afghanistan between 2007 and 2018, with ministers, governors, a former vice-president, warlords and their entourages, opposition leaders, diplomats, NGO workers, and local journalists and researchers, Romain Malejacq provides a full investigation of how warlords adapt and explains why weak states like Afghanistan allow it to happen.”  (from the press website)


Many congratulations on the publication of your excellent book, Warlord Survival: The Delusion of State Building in Afghanistan. Can we begin by reflecting on your personal journey that led to the writing of this book? How did you get interested in Afghanistan and the phenomenon of ‘warlordism’ in particular?

Initially, I actually had no intention of working on Afghanistan. In fact, I had absolutely no prior connection to Afghanistan. I spent a year in South Africa as an undergraduate student and that’s really when I became interested in peace and conflict studies. I had also travelled quite extensively throughout Africa before and I continued to do so during that year. It’s almost naturally that I decided to focus my research on African conflicts. I became interested in warlords long before I even thought about studying Afghanistan. I became acquainted with ideas and theories about warlordism while writing my Master’s thesis, which had nothing to do with Afghanistan, but focused exclusively on Liberia and Sierra Leone. I was specifically looking at theories that emerged to describe and explain the Mano River conflicts. Warlordism was an important topic there and I decided to continue on this topic for my PhD. I started my PhD thinking my dissertation would be a comparative study of warlords in a few African countries. It’s only then, the very first month of my doctoral studies, that I had the opportunity to work on and go to Afghanistan for a completely unrelated research project. I jumped on the occasion. I spent that month in Kabul and immediately became fascinated by the place. My PhD advisor then suggested to include Afghanistan as one of my cases, which I did. And I went back to Afghanistan, again and again, every time falling more in love with the place. The more I learned about Afghanistan, the more I realized how complicated it was, the more I realized how little I actually knew. Almost fifteen years later, I still feel this way sometimes, and that’s part of what captivates me and motivates me to go back, time after time. For me, Afghan politics are truly fascinating. Anyway, I quickly understood that there was no point in spreading myself thin and trying to understand multiple places and societies that I knew nothing, or very little, about. I quickly dropped the other cases, and focused on Afghanistan and Afghanistan only. And that’s the story of how I came to study Afghan warlords. I don’t regret a thing.

The author interviewing Ismail Khan in his palace, Herat (photograph by friend of the author)

You have conducted extensive field research over so many years to write this book. Did you encounter any challenges and major surprises during the field work?

Yes! Many challenges and at least as many surprises! First, because Afghan culture is so different from my own. It takes a while for one to get acquainted with people’s traditions and ways of life, and make sure to behave appropriately in all circumstances. I clearly made many faux pas the first times I visited Afghanistan. And I’m sure I still do. Hopefully, less and less… But that’s somehow inevitable. Second, of course, fieldwork is challenging because of the current security situation in Afghanistan. Conducting fieldwork in conflict settings is always challenging but doing research in Afghanistan is becoming increasingly difficult. Especially research on non-state armed actors, such as warlords, militias, or insurgent groups. Access, transportation, accommodation… everything becomes challenging. Last, but maybe most importantly, being confronted with civil war destruction and suffering is something Western researchers are not prepared for. It is of course incomparable with what Afghan people have to go through, and we, foreign researchers, have the ability to leave the conflict behind once we’re done with fieldwork, but experiencing civil war is always traumatizing in one way or another. It is something that needs to be better acknowledged and dealt with, along with all the ethical and moral issues, and sometimes the sense of guilt, that come with doing what we do.[1]

Now, as far surprises go, I actually had quite a few. The first one came almost as soon as I set foot on Afghan soil. I had never been to a conflict zone. I had ideas about what to expect but I vividly remember being quite in shock and impressed by the visuals of foreign intervention. The checkpoints, the sandbags, the barbed wires, the T-walls, I had never seen anything like that before. To add to this, the streets were almost empty, apart for pick-ups filled with young men in arms. After a night of travelling, it made quite an impression on me. It was really something I had never experienced before. What I only realized later is that it was not always like that. We had landed on September 9, on the anniversary of Ahmad Shah Massoud’s death, which was the reason for the empty streets and the men in arms. And then, a couple of days later, as my friend and I were driving back from a cultural evening at the French high school, I ended up with a green dot on my chest. Nothing happened, the driver slowly reversed the car and we turned around, but that was quite the first trip.

Portrait of Ahmad Shah Massoud with inscription “national hero” (graffiti adding “of Pakistan”), Kabul (photograph by author)

The word ‘warlord’ and its Dari translation jangsalar can have strong pejorative connotations. I remember, many in Afghanistan argued against using the term as they felt it was instrumentalised by some groups to delegitimise and even demonise their rivals. How do you think the practical use and abuse of the term may affect academic research on the topic?

I would like to think that it doesn’t, that the way the term ‘warlord’ is used in the media and by people in general does not affect the analysis, but it does. Academic work on warlords tends to be overly normative. The term is used to vilify a certain category of political actors which, it is assumed, the international community should get rid of. Scholars have compared warlords to viruses and parasites, feeding or encroaching themselves on the population and the state. To hyenas even! Many use the term to evoke these individuals’ violent and criminal behavior, but eclipse anything else that they do or represent. Often with a hint of Orientalism.

There are definitely valid arguments in favor of dropping the term altogether. I personally went back and forth for while. In the end though, I kept the term because I believe that it has academic value. Not to delegitimize political rivals or designate non-state armed actors in sensationalist ways, but as a social science concept. It is often used improperly, normatively, politically even, but it describes a category of political actors that operate according to a very distinctive logic.

In the book, I define warlords as “astute political entrepreneurs with a proven ability to organize violence and control territory, who exert and transform authority across different spheres (ideological, economic, military, social, and political) and at different levels of political affairs (local, national, and international)” (p.4). I make absolutely no normative assumption about them. At least I try not to. I just try to focus on what they do rather than on what is assumed of them. It won’t prevent people from making conjectures about what I say about warlords. It goes both ways anyhow. Some will say I am pro-warlords, others that I am anti-warlords. But in fact, I’m neither. I’m just trying to understand and explain a particular phenomenon.

So, going back to your question, yes I believe that the way the term has been used (and abused) does affect academic research, and in a problematic way. It’s even more the case in Afghanistan, where political actors themselves have used it to vilify others, but it does not make the concept less relevant. We just have to be very careful not to use it normatively, and to really distinguish between the social science concept and the popular use of the term.

Soviet tank remnant, Panjshir Valley (photograph by the author)

In your book, you explore how warlords accumulate and project power to survive in highly volatile environments. Can you please explain how power projection by a warlord is different from power politics more generally?

Power projection is the mechanism through which political actors convince others that they are legitimate, and have authority. That’s how warlords remain indispensable in the eyes of those who need what they can provide, whether it is votes, security, or job opportunities. This is not necessarily unique to warlords though. Populist leaders and dictators, for example, use propaganda and other means to boost their image. Most political actors develop sophisticated communication strategies. What is unique about warlords is that have the ability to use violence, which they combine with the capacity to reach all levels of politics.

Thanks to their ability to use violence, not only can warlords supply goods and services that most other political actors cannot, but they can also create disorder. In places where state capacities are weak, like Afghanistan, this makes warlords very distinct from other political actors. It provides them with a lot of new opportunities. And, since a warlord’s power is extremely difficult to assess, it also makes them very difficult to get rid of. Those who would like to see warlords wither away have to deal with them quite conservatively, as they fear their capacity to foster instability.

What I really try to explain in the book is that there is a circular logic to how warlords perpetuate their power. The more powerful they look, the more their wishes will be accommodated, and the more powerful they will become. And the more powerful they are, the more powerful they look. In fact, they remain powerful as long as long as others believe they are. This brings us to the warlords’ second distinctive feature. They have the ability to exert power and make themselves indispensable at all levels of political affairs, from local communities, to political elites in the capital, to foreign state representatives, which, combined with their ability to use violence, allows them to perpetuate their power.

Some have argued that the so-called warlords may find it in their interest to work through state institutions and thus potentially become state builders. You argue that this is a delusion. Can you please explain why you think this is a delusion? 

Well, I don’t necessarily argue that warlord cannot become state builders, or that they cannot be integrated into some sort of state-building process. On the contrary, I conclude the book by saying that state building can only be successful if it involves the incorporation and absorption of the warlords’ social capital and networks on terms that are useful to the state. The delusion is to think that external state building aimed at centralizing and monopolizing political authority will work. Warlords are integral to the way states like Afghanistan actually work and they won’t go anywhere. There is also no reason to believe that they will turn into benevolent civil servants and just abandon their other sources of authority. This is not to say that there can be no state and that warlords cannot play their part in building it. They will if it serves their interests indeed. But they will maintain their ability to harness different sources of power outside of the state. So, yes, warlords can help the state extent its reach, but their power will endure no matter what, whether we like it or not. In fact, they use the state’s authority to increase their own and use their own to increase the state’s. What I’m truly saying in the book is that external state building in Afghanistan and other similar environments is impossible without significant concessions to these warlords. They will not be eliminated and replaced by the state. So yes, external interventions that aim at building centralized, bureaucratic states are doomed to fail and, in these conditions, lead to what I call the delusion of state building.

[1] Malejacq, Romain and Dipali Mukhopadhyay. 2016. “The ‘Tribal Politics’ of Field Research: A Reflection on Power and Partiality in 21st-Century Warzones. Perspectives on Politics 14(4): 1011-1028.

More of The Same: Kazakhstan’s leadership change between ageing leadership and popular discontent, by Luca Anceschi (University of Glasgow)

If there is only one lesson to be learned from Kyrgyzstan’s recent presidential dispute—a chain of tumultuous events that led to the arrest and detention of erstwhile leader A.S. Atambayev[1]—is that post-transition relationships between Central Asia’s incumbents and its former presidents continue to represent one the most intriguing political mechanisms at play in the region. These relationships, it ought to be noted, seem to occur at very rare junctures: Central Asia’s leaders tend to remain in power for decades, reducing the transfer of power to élite-driven mechanisms that do normally set in motion only in the aftermath of a presidential death. The playbook for post-mortem tranzit vlasti was perfected through successfully orchestrated presidential successions in Turkmenistan (2006-2007) and Uzbekistan (2016); at the time of writing, there is no conclusive evidence to maintain that post-Rahmon Tajikistan will deviate significantly from this norm.

The option to observe newly elected (or appointed) leaders interacting with their predecessors is therefore only available in the Kyrgyz context, which continues to hold regular elections despite its continuously sliding democratic record and, since 19 March 2019, in Kazakhstan, where long-term leader N.A. Nazarbaev relinquished the presidency to facilitate the accession to power of K.K. Tokayev, an established regime insider, former foreign minister and, since 2011, the chairman of the Kazakhstani Senate.

Relatively free and fair elections do generally legitimise Kyrgyzstan’s elected presidents, who enjoy as a consequence a modicum of popular support throughout the single mandate allowed by the Kyrgyz Constitution.[2] A set of different dynamics came to the fore in post-Nazarbaev Kazakhstan, where presidential succession was sealed through an élite-dominated process that sought its popular legitimation via a regime-controlled election. Here, the Kazakhstani population attempted to throw a spanner in the works of such a carefully orchestrated tranzit vlasti mechanism: both prior and after the vote that formalised Tokayev’s accession to the presidency, protests and demonstrations erupted in the country’s principal urban centres, as anti-regime sentiments came to the define the local political debates in the spring of 2019. It is precisely to the contribution played by Kazakhstan’s politically active population to the establishment of a working incumbent/predecessor relationship in post-Nazarbaev Kazakhstan that this blog entry devotes its ultimate attention.

Kazakhstan’s much anticipated transition

A look at the overarching strategy put into place by the Kazakhstani regime between 2014 and 2019, but most emphatically in the period that followed the death of Islam Karimov (August 2016), indicates that Nazarbaev’s resignation did not represent an impromptu decision. There is sufficient evidence to maintain that, in the timeframe in question, the Kazakhstani regime had designed and implemented a comprehensive process to orchestrate a transition out of the Nazarbaev presidency. A series of government reshuffles and the frequent introduction of legislative adjustments represented the two key drivers sustaining Nazarbaev’s attempt to instigate a thorough mechanism of leadership rejuvenation without incurring in the risks inevitably associated with regime change dynamics.

Placing the spotlight on the recent career trajectory of K.Q. Massimov—the second most powerful regime member of the late Nazarbaev era—identifies with some precision the élite component of Kazakhstan’s leadership change mechanism. Appointed to the country’s prime ministership in the spring of 2014, Massimov was eventually moved to the chairmanship of the Kazakhstani KNB (September 2016). Rather than an apparent demotion, this move came to embody politically the gap existing between the inner location of Kazakhstan’s authoritarian power and its institutional representation. Massimov’s appointment shifted an unviable candidate for succession—his ethnic profile is regarded as generally incongruous with the wider process of Kazakh-ification of Kazakhstan’s political life that has been undergoing in the post-Soviet years—to a position of not visible, yet certainly not marginal, influence. Massimov emerged as one of the king-makers in the identification of a suitable post-Nazarbaev leadership, while his appointment at the helm of the KNB replicated the structural organisation of pre-transitional Turkmenistan and Uzbekistan, where indisputably central roles in leadership selection were assigned to prominent representatives of the power ministries, namely A.K. Rezhepov in Turkmenistan and R.R. Inoyatov in Uzbekistan.

At the time of his accession to the prime ministership, B.A. Sagintaev, who replaced Massimov as Kazakhstan’s PM, had apparently been groomed for a top political appointment for at least five years. The 2016 reshuffle seemed to have in this established a solid transitional diarchy to regulate an eventual presidential transition, while the elevation of Darigha Nazarbaeva to a Senate seat guaranteed Kazakhstan’s first family a further stronghold in the institutional configuration of the late Nazarbaev era.

The second constituent element of the authoritarian environment wherein to launch a presidential transition was represented by the establishment of an adequate legislative framework to safeguard, in case of a voluntary resignation from the presidency, the power position of Nazarbaev and of his immediate family, guaranteeing at the same time their business interests and immunity from crimes committed while in office. Carefully crafted and continuously amended between 2000 and 2010, the Law ‘On the First President of the Republic of Kazakhstan—the Leader of the Nation’ was basically meant to avoid the repetition, in post-Nazarbaev Kazakhstan, of the dynamics that led to the political marginalisation and expropriation of the wealth held by Uzbekistan’s first family, with particular reference to the arrest and protracted detention of Gulnara Karimova, the elder daughter of the late I.A. Karimov.

The premeditated nature of Nazarbaev’s resignation became evident in early 2019. On 4 February, the president himself addressed the Constitutional Council requesting detailed information about the powers he would retain in case of voluntary resignation. Nazarbaev’s eventual backpedalling did not silence those observers who regarded a change of guard in Ak Orda as an option not to be discarded a priori. A few weeks later (21 February), Nazarbaev demanded the resignation of the entire Kazakhstani government, appointing A.U. Mamin as the country’s interim prime minister. The exceptionality of this latter development did not reside in Nazarbaev’s very public reprimands of the government’s agenda—a common trait in Central Asian authoritarianism—nor did it relate to the president’s attempts to scapegoat Sagintaev for Kazakhstan’s poor economic performance. Indeed, Kazakhstan’s authoritarian practice has regularly presented the president’s direct intervention as a panacea for even the most severe economic crisis. Rather, the governmental reshuffle of late February 2019 is significant insofar as it realigned Kazakhstan’s transitional team to the patronage network of its principal king-maker. As a long-term associate of K.Q. Massimov, Askar Mamin was expected to rule in unison with the KNB chairman, completing a triumvirate of which the Nazarbaev family has to be seen as the third vertex.

The authoritarian milieu described above was undoubtedly established with a view to facilitate an intra-regime mechanism of presidential succession, that is excluding political outsiders or regime dissenters from the highest institution in the Republic of Kazakhstan. It is nevertheless difficult to identify the single factor that accelerated Nazarbaev’s decision to resign from the presidency. Nazarbaev’s self-perception of his age, his preoccupation with a rapidly eroding legacy, or even the ultimate retribution of previously sealed élite pacts may be some of the key factors behind the decision announced publicly on 19 March 2019. Speculations aside, the president’s announcement clearly noted that the succession process initiated by his resignation ought to follow the constitutional dictates, which stipulate the assignment to the interim presidency to the Chairman of the Senate, a post occupied at the time by K.K. Tokayev. The latter’s selection, yet again, did not represent an accidental development: on the one hand, Tokayev’s loyalty to the regime preservation agenda was (and continues to be) unquestionable; on the other, executing a presidential transition following the constitutional dictates set Kazakhstan aside from the regional praxis, where power transfers had hitherto led to the arrest (Turkmenistan) or voluntary resignation (Uzbekistan) of the legitimate presidential successors. For a regime that has traditionally put a premium of differentiating between the governance practices of post-Soviet Kazakhstan and those followed by its immediate neighbours, the execution of a constitutionally impeccable transition at the end of a long, and ultimately post-colonial, presidency did indeed constitute a significant achievement.

The final adjustment to Kazakhstan’s post-Nazarbaev institutional settings was represented by the selection of Darigha Nazarbaeva as Tokayev’s replacement on the Senate Chair. While this author has always approached with vocal scepticism any discussion on the prospects for dynastic succession in Central Asia, it is nevertheless true that this appointment places Nazarbaev’s daughter into the top succession position should Tokayev die or become incapacitated while in office. To my mind, Nazarbaeva’s rise to a top institutional post is the institutional facet of the legislative guarantees enshrined in the First President law, inasmuch as it protects the mid-term power position of the first family, while ensuring that, at least until Kazakhstan’s first president remains alive, the dynamics we saw at play in post-Karimov Uzbekistan are not to be replicated in Kazakhstan.

This orchestrated process was meant to be concluded smoothly, as the regime expected a rapid electoral validation through a tightly controlled vote scheduled for 9 June. Beyond the immediate shock for the departure of a long-serving, and generally respected, leader, some segments of the wider Kazakhstani population reacted with visible dissatisfaction to the post-Nazarbaev transition. As a consequence, Kazakhstani-watchers spent the early Tokayev era by looking at changing patterns of regime-population relations, rather than speculating on the development of a working collaboration between Nazarbaev and his hand-picked successor.

Mis-managed expectations: The population’s reaction to Kazakhstan’s transition

The Nazarbaev regime was not the only actor to anticipate Kazakhstan’s inevitable presidential succession. The people, or at least some segments of the Kazakhstani population, have been preparing for years to, or at least discussing the context leading to, leadership turnover in Ak Orda. For much of the 2010s, Kazakhstan lived through a Twilight Zone,[3] wherein flawed assessments of the country’s authoritarian stability and of the population’s political behaviour created an unstable political environment predicated upon Nazarbaev’s ageing leadership. The ambition to have an elected second president, rather than a merely appointed one, seemed to have been shared by a significant number of ordinary Kazakhs, and it underpinned the popular reaction to the orchestrated transition completed between March and June 2019.

The post-Nazarbaev era began with the brutal repression of popular demonstrations held across Kazakhstan to protests the perceived democratic deficit of the popular vote that sanctioned Tokayev’s election. The optics of video reports originating from Kazakhstan across June 2019 were quite dispiriting, establishing a direct parallel with the Zhanaozen events of 2011, in the sense that, yet again, we witnessed the public suppression of ordinary citizens manifesting their views while the country was meant to celebrate collectively an important landmark of its independent life.[4]

Images here from the Oyan, KZ facebook page, and of the reporting on the ‘You can’t run from the truth’ action in 22 April 2019.

The establishment of dissenting movements including Oyan, Qazaqstan [Wake up, Kazakhstan] and the popularisation of slogans challenging the nature of the presidential transition itself—Ot pravdy ne ubezhish’ [You can’t run from the truth]; U menya est’ vybor [I have a choice]—confirmed the views expressed by Dossym Satpayev[5] so far as the leadership’s misperception of the population’s passivity as an indicator of its fundamental loyalty to rules and norms imposed by the regime. It is not clear whether the public moment that pro-democracy activists have come to experience in 2019 can evolve into the institutionalisation of established opposition forces. However, the diffusion of anti-regime sentiments as the first president leaves Kazakhstan’s political limelight reveals a disconnect between ordinary Kazakhs and the post-Nazarbaev élites, a disconnect that, incidentally, has not emerged as visibly in the Uzbek context, where the population seems to be largely onboard with the political agenda introduced by Sh.M. Mirziyoyev.

Tokayev choose to relate to policies and practices established by his predecessor through a posture of unwavering continuity, misinterpreting the significant demands for change that the wider populations expressed more or less openly throughout the 2010s. Much of the regime’s future stability may be therefore predicated upon its willingness and ability to address this fundamental governance problem.


[1] Global coverage of these events includes:;; ;

[2] For an overview of contemporary politics and society in Kyrgyzstan, see for example Kyrgyzstan: Beyond ‘Democracy Island’ and ‘Failing State’: Social and Political Changes in a Post Soviet Society, edited by Marlene Laruelle and Johan Engvall (2015).

[3] D. Satpayev (ed.), Sumerechnaya Zona ili lovuuski perekhodnogo perioda (Almaty: Alyans Analiticheskikh Organizatsii, 2013).

[4] Protests and arrests were highlighted in international media coverage of the election, for example;; ;

[5] ‘Are Risks Increasing for Kazakhstan? An interview with Dossym Satpayev’, Voices of Central Asia, 18 April 2019. See also his recent commentary on these topics here:

Representing the Social Costs of Migration: Abandoned Wives or Nonchalant Women by Malika Bahovadinova, Oriental Institute, Academy of Sciences of the Czech Republic

The workshop

In the spring of 2013 a private workshop was organized by a major international donor for its Tajikistani state and NGO partner organizations in Dushanbe. The event was part of the reporting process related to a large labour migration program being implemented by three large international development agencies. I attended this event as a part of fieldwork on the bureaucracy of migration management I conducted between 2012-2014.

Continue reading Representing the Social Costs of Migration: Abandoned Wives or Nonchalant Women by Malika Bahovadinova, Oriental Institute, Academy of Sciences of the Czech Republic