VoxEU Column Economic history Politics and economics

Failed states and the paradox of civilisation: New lessons from history

While cases of state failure have risen in the last decade, most notably in the Middle East and sub-Saharan Africa, they are not a new phenomenon. Historical evidence from the early modern period, and even the Bronze Age, shows that the majority of formed states have failed rather than thrived. This column introduces the ‘paradox of civilisation’ to characterise the obstacles settlements face in establishing civilisations. The paradox defines the success of a civilisation as a trade-off between the ability to produce economic surplus and to protect it. It is therefore important to correctly balance military and economic support when providing aid.

‘Failed states’ are the most pressing international policy concern in the post-Cold War era (Fund for Peace 2015). Failed states have been the source of major humanitarian catastrophes, including mass famines (Chad, Ethiopia, and Burkina Faso), devastating civil wars (Libya, Mali, and Niger), or both (Sudan, Somalia, and DR Congo). The interest of the international community in failed states is not only altruistic. Security concerns also have a decisive weight, as failed states are often breeding grounds for international terrorist organisations, especially radical Islamism, in Afghanistan, Pakistan, Iraq, Syria, and Yemen.

Two questionable lessons

Policy responses to failed states have been disappointing. The effectiveness of the current strain of foreign interventions, inaugurated by George W Bush’s “War on Terror”, is still being hotly debated. Since the invasion of Iraq in 2003, cases of state failure have grown steadily, in terms of numbers and the scale of damage. It is far from clear what a counterfactual Middle East or a counterfactual sub-Saharan Africa would look like today had foreign powers abstained from intervening. One big message from recent experience is undisputable: removing a bad government is much easier than parachuting in a good one.

Two more specific lessons about ‘reconstruction’ have forged a new policy consensus, but both need re-examination. The first lesson has become a mantra, that ‘institutions matter’. Reconstruction involves much more than economic aid, either humanitarian or infrastructural. It also requires the creation of political rules and social habits for conflict management and government legitimacy (Fukuyama 2006). Second, successful reconstruction requires a specific ‘policy sequence’. Since state failure is a multidimensional problem – including political violence, ethnic conflict, economic collapse, and the breakdown of basic health and transportation services – reconstruction involves building different components of civilised life. The order in which components are put in place will affect the final outcome (Herbert 2014).

A broader historical perspective

‘State failure’ is a new phrase, but the dramatic facts of political and economic collapse are as old as states themselves. There is no reason to restrict analysis to developments in the 21st century. Looking at a larger set of historical cases actually helps to rectify the lessons about institutions and sequence derived only from contemporary experience. History shows that, first, force is more important than institutions, and second, that the balance between economic prosperity and political order is more relevant than the sequence.

Failed states seem a novelty only in relation to the bipolar world of the Cold War, where the US and USSR made of the prevention of collapse in client countries an imperial priority. But failure has a long lineage. According to Charles Tilly, in early modern Europe, the very birthplace of the (Weberian) state, “the substantial majority of the units which got so far as to acquire a cognisable existence as states [from 1500 to 1850] still disappeared” (Tilly 1975).1 Failure is even more prominent in pre-modern times. While the rise of the homo sapiens occurred 200,000 years ago, civilisations emerged as recently as 6,000 years ago, and only in half a dozen selected world regions. Before the rise of Mesopotamia and Egypt, economic stagnation and conflict had been endemic. Moreover, civilisations were far from irreversible outcomes. By the end of the Bronze Age, major Eastern Mediterranean civilisations had collapsed under the pressure of invasions by less developed societies, the ‘Sea Peoples’. The re-emergence of a civilised order in Mediterranean Europe had to wait for another 500 years. The Roman Empire, the peak expression of the new order, also fell prey to invasions from the Goths, Huns, Vandals and other barbarian tribes.

A broader temporal focus thus reveals that failed states are not the exception but the norm in human history.

The paradox of civilisation

The common denominator of successful societies, from early civilisations to modern states, is the dual ability to produce surplus (prosperity) and to protect surplus (security). Contemporary cases of state failure are just instances of the large class of societies that failed to produce surplus and protect it. The class is so large that it actually accounts for about 98% of the human timeline, and covers no less than a fifth of the contemporary world. 

The perennial nature of state failure, as well as the exceptional character of state formation, is rooted in a simple but powerful paradox. Every society in the process of development faces a fundamental trade-off between prosperity and security. The efforts of a society to create wealth will undermine its own sovereignty if the new prosperity attracts predatory attacks from rival groups (either inside or outside the social territory). In a recent paper, we introduce the ‘paradox of civilisation’ to characterise the dilemma shared by thousands of early agrarian settlements prior to the rise of pristine civilisations in Sumer and Egypt, hundreds of ports, cities, and villages in Medieval Europe and post-colonial Latin America, and current attempts at reconstruction in the Middle East and sub-Saharan Africa (Dal Bó et al. 2015).

Most of these societies, during most of their existence, were trapped between two bleak alternatives. On the one hand, the dangerous option of ‘self-defeating prosperity’, i.e. investment efforts that would induce predatory attacks, and on the other hand the safer but stagnant option of ‘backwardness by design’, which would prevent predation at the cost of keeping economic activity close to subsistence levels, hence shutting down the path toward civilisation. The reader will have no problem in finding cases of self-defeating prosperity in history books, which are replete with examples of productive polities that, at different stages of development, fell prey to the voracity of economically simple but militarily aggressive societies. By contrast, cases of backwardness by design are ‘dogs that didn’t bark’. Their aborted development implies scarcity of historical traces. However, as recently as the eve of American independence, Anglican minister Woodmason provided a perfect example in South Carolina, in that “the finest lands in the province [remain] unoccupied and rich men [are] afraid to clear them” because they do not want to “become a Prey to the Banditti” (Sayre 1994).2

Pre-institutional sources

To address the paradox of civilisation we developed a formal model in which a society with the potential to generate economic surplus (the incumbent) faces a hostile group that can capture its output (the challenger). The incumbent has to decide how to allocate resources between surplus production and surplus protection in anticipation of predatory raids by the challenger, whose incentives to attack rise (fall) with the level of prosperity (defence) achieved by the incumbent. Informed by anthropological and historical literature, we emphasise pre-institutional forces, especially the physical aspects of the geographical environment, which define the initial resources of the incumbent (parameter v in our paper), and its ability to transform initial resources into future income (growth capability, ρ) as well as into defence (defence capability, κ).

The model produces a number of implications, some of them quite contrary to conventional wisdom on the origins of political order and economic prosperity. The implications also encourage a re-interpretation of the rise and fall of Sumer and Egypt—the first societies to jointly achieve prosperity and security—and help to rectify the lessons from contemporary failed states.3

The first contribution of the model is the identification of the conditions—combinations of different values for the parameters v, ρ, and κ­­—under which societies achieve different levels of economic prosperity and security. Four general outcomes are relevant.

  • Failed states

Societies remain trapped in a context of economic stagnation and military vulnerability when defence capabilities κ are low, and growth capabilities ρ are high relative to κ. Especially if its initial income (v) is high, the incumbent society will be a magnet for hostile groups, but its low κ will impede effective deterrence. In anticipation, the incumbent society will make no efforts at growing the economy.

  • Order without prosperity

When ρ is low and κ is high relative to ρ, societies may achieve security by forgoing opportunities to grow the economy. Keeping the economy poor keeps predatory groups away.

  • Prosperity without order

Societies will grow their economy despite remaining highly vulnerable to predatory attacks when ρ is disproportionately high in relation to κ. This is a truly anti-Hobbesian outcome but one that captures many situations in which societies continued investing even though predatory attacks were very likely, like the Anglo-Saxons and Chinese in the face of imminent Viking and Mongol invasions (10th and 13th centuries) respectively.

  • Civilisation

Societies will jointly achieve security and prosperity (civilisation) when κ and ρ are not only relatively high but also balanced. Under these conditions, societies are able to allocate part of their initial resources to grow the economy and, given their military capabilities, the remainder is enough to create a deterrence and protect the fruits of production.


The model offers a new understanding of the rise and fall of civilisation and states, and helps to rectify the lessons from contemporary failed states. The paradigmatic civilisational breakthroughs in Sumer and Egypt are processes that were more militaristic than is currently acknowledged by anthropologists and historians, and solutions to state failure are processes that will be less institutional than is assumed in contemporary recommendations by the international community.

Since the work of V. Gordon Childe (1950), archaeologists have accepted that the rise of civilisations was primarily driven by an exceptional potential for food production both in terms of endowments and technology. The availability of fertile soil, abundance of edible plants and animals, and the development of irrigation combined to make possible the Agricultural Revolution, an economic transformation that in many ways is more fundamental than the Industrial Revolution (Diamond 1997). Without a substantial surplus, it is not possible to fund the tangible components of civilisations. However, surplus production is only half of the story. The other half is surplus protection. As our model shows, without adequate protection, prosperity can actually be self-defeating. Historian William McNeil observed that “the relatively enormous wealth” created by the Agricultural Revolution made the first surplus-producing societies “worthwhile objects of attack by armed outsiders” (McNeil 1979). We are not aware of any work that has analysed the rise of Sumer and Egypt under the perspective of the paradox of civilisation. Out of thousands of agrarian villages scattered around the world between 8,000 and 2,000 BC only the select few that managed to deter predatory attacks were able to develop into full-fledged civilisations. The model predicts that high productivity must go hand in hand with strong defence for civilisation to arise. Natural protection provided by the desert in Egypt and man-made defences erected in Sumer were crucial components in their developmental paths. In Sumer and Egypt, irrigation systems were the single most important investment in surplus production. In Sumer, irrigation was coupled with perimeter walls, the other large-scale public good, which was a partial substitute for the Egyptian desert in terms of surplus protection.

The model produces specific lessons for policy interventions in failed states. First, the balance between military and economic aid is more relevant than the sequence. Economic aid alone can actually be counter-productive for the purposes of state formation as higher income will spur the voracity of predatory groups. New economic resources must be accompanied by new capacities, indigenous or foreign, to provide some protection. Second, force is more important than institutions. The notion that institutional design will provide a solution to state failure derives from an erroneous policy application of a foundational finding in the field of political economy. The finding is that liberal institutions like property rights, the rule of law, and electoral accountability are the pillars of modern development. However, failed states lack the resources to enforce any type of institutional setup, either liberal or illiberal. The priority for failed states is more power rather than better rules.

Liberal institutions were a key precursor of the Industrial Revolution in Western Europe and North America. Nevertheless, since pre-historical times, the relevant large-scale economic transformation that made polities viable was the Agricultural Revolution, the success of which rested on the physical deterrence of predatory groups. The Agricultural Revolution in Europe and elsewhere preceded the rise of liberal institutions by thousands of years, but it is largely pending in substantial sections of tropical Africa and was essentially undone in large portions of the Middle East. Failed states are placed in a more basic developmental trap than the one faced by the North Atlantic economies on the eve of the Industrial Revolution: failed states do not prosper because, without security, prosperity exacerbates conflict.


Dal Bó, E, P Hernández, and Mazzuca, S (2015), “The Paradox of Civilization: Pre-Institutional Sources of Security and Prosperity”, NBER working paper 21829

Diamond, J (1999), Guns, germs, and steel: The fates of human societies, WW Norton & Company

Fukuyama, F (2006), Nation-Building. Beyond Afghanistan and Iraq, Baltimore, Johns Hopkins University Press

Fund for Peace (2015), Fragile States Index 2015, Washington DC.

Childe, V G (1950), “The urban revolution”, Town Planning Review, 21 (1), 3

Herbert, S (2014), Sequencing reforms in fragile states: Topic guide, Birmingham, UK GSDRC, University of Birmingham

Herz, J H (1950), “Idealist internationalism and the security dilemma”, World politics, 2 (2), 157-180

Jervis, R (1978), “Cooperation under the security dilemma”, World politics, 30 (2), 167-214

McNeill, W H (2013), The pursuit of power: Technology, armed force, and society since AD 1000, University of Chicago Press

Sayre, R F (1994), American lives: An anthology of autobiographical writing, University of Wisconsin Press

Tilly, C, and G Ardant (1975), The formation of national states in Western Europe, Volume 8, Princeton University Press.


[1] Since only survivor states have a well-documented history, the magnitude of the ratio of failed states to successful states remains colossally underestimated.

[2] It is useful to differentiate the paradox of civilisation from the classical ‘security dilemma’ in international relations theory, which refers to the fact that a country’s efforts to become more secure threatens the security of the rest, which respond with identical efforts, resulting in enhanced conflict and insecurity for all (Herz 1950, Jervis 1978). In the paradox of civilisation, societies are able not just to increase their military force but also to make productive investments to grow their economies. Crucially, the second type of investment is not security-neutral because increased prosperity will provide neighbours with an incentive to invade. Finally, in anticipation of the geopolitical dangers created by prosperity, societies may avoid economic investments, forgoing its potential for growth in exchange for military security.

[3] By an irony of history, ISIS is carving its territorial domain from the most dramatic cases of failure, Syria and Iraq, which occupy the same geographical region as the first civilisations.

6,299 Reads