Scarce states are not always weak states
Political scientists often measure state weakness as state absence: where we count fewer bureaucrats, where we see fewer roads that enable state penetration, or where citizens have fewer state-issued documents like IDs or birth certificates, it seems reasonable to assume states have more limited power to influence society. Zones of state absence become the “brown areas” on our maps of statehood.
Conceptualizing weakness as absence shapes how historical political economy (HPE) scholars study places with stereotypically weak states. Consider two examples from the thriving literature on precolonial and colonial legacies in Africa. Scholars often explain correlations between contemporary outcomes and precolonial institutions, such as political centralization, through arguments that presume these institutions have persisted in the face of relatively absent states incapable of fully displacing them. Along these lines, data on precolonial institutions in George Murdock’s famed Ethnographic Atlas is regularly interpreted at face-value as characterizing current ethnic group attributes. Elsewhere, studies of colonial policy often use subnational regions with less state activity as control groups for estimating effects of more extensive interventions elsewhere. The implicit assumption is that zones of state absence provide a good counterfactual for what a more undisturbed society would have looked like.
My new book, The Scarce State: Inequality and Political Power in the Hinterland, zooms in on quintessential zones of state absence: rural hinterlands. Focusing on Northern Ghana alongside shadow cases, I track the long-term effects of isolated state actions in a place that otherwise experienced limited state attention and investment since the outset of colonial rule. The book suggests that absent states can be more powerful than we often think, including through an underappreciated ability to create elements of the “strong societies” that then turn around and seem to dominate these weak states.
The book’s evidence has particular implications for HPE. I show how precolonial institutions changed massively from even fleeting contact with the modern state, suggesting a need to rethink how we understand evidence of precolonial persistence. I also explain how regions of state absence were at times changed more by state actions than areas that received much higher doses of those same policies, with hinterlands that often comprise our control groups in studies of long-run relationships far from the static reference category they are assumed to be.
Large effects of scarce states
In the rural periphery of the developing world, the state is often not just absent, it is scarce. Even where they have few officers on the ground, developing states can oversee the allocation of resources far more valuable than those available in underdeveloped local economies. Classic accounts focus on the state’s role as an extractor, even a predator or bandit. But in subnational regions with little wealth worth attempting to extract – something I argue is characteristic of many hinterlands in developing countries — states often give out more than they take back. In these regions, the state’s actions and the resources it controls are in short supply relative to demand for them in rural society. Much as the scarcity of a good in a marketplace increases its price, the scarcity of the state increases the relative impact its actions can have.
Scarce states step into society with real weight behind their footsteps. Their steps leave large imprints even if they take few of them overall. In fact, that scarce states only attempt a few steps into society itself is what makes these actions so impactful. In contexts with limited baseline access to resources, isolated state actions confer windfalls on narrow sets of beneficiaries in society, generating new forms of local inequality. This inequality sticks and becomes entrenched, changing allocations of power within societal institutions, if the state takes few subsequent actions that allocate new windfalls to other recipients that could offset the compounding effects of its earlier windfalls.
The lucky few with privileged contact with an otherwise scarce state get an economic head start in a context in which – partly because there’s such limited state investment — there are few other paths available for anyone else to catch up. Seemingly small steps into society can then have big effects on who comes to wield local economic and political power. Counterintuitively, the same state actions can have smaller effects in places where states do them at a larger scale, governing more, and in the process, spreading economic opportunity more widely.
The book focuses on tracing out the long-run effects of three modern state actions in a hinterland it mostly otherwise ignored: early colonial efforts to stand up new forms of chieftaincy to govern Northern Ghana indirectly; belated attempts to build out a rudimentary public education system in the independence era; and post-colonial efforts to delegate land ownership to local actors, in part to reduce the state’s administrative responsibilities. Each action delivered targeted windfalls more valuable than what was previously available in the local economy, and in the process, remade society.
The combined effect has been persistent economic inequality, which in turn has helped create political dynasties, clientelism, and the grievances underlying political violence. Elite capture, clientelism, and violence are outcomes we usually take as evidence that a state is weak. But in this case I suggest they were caused by the power of the state over society, suggesting that “weak” might be the wrong adjective for thinking about states in the rural periphery in the first place.
Implications for precolonial legacies
The book complicates how HPE scholars think about precolonial institutions. For Northern Ghana, the Ethnographic Atlas – a central data source for many scholars — paints a wildly inaccurate picture of current institutions. This is even though this is exactly the type of place – a zone of persistently limited state presence – where classic works would expect the most continuity to the pre-colonial period.
Many local ethnic groups coded as pre-colonially acephalous – lacking chieftaincy or similar institutions of political centralization – are not acephalous at all today. Chieftaincy positions artificially imposed by the colonial state have produced centralized systems of traditional governance now just as meaningful as those in nearby pre-colonially centralized groups. In fact, I show that state-invented chiefs have become more effective clientelist intermediaries – a common role assumed of chiefs in the political science literature – than those whose positions date to precolonial institutions.
And among a smaller set of pre-colonially acephalous ethnic groups that still remain nominally acephalous today – lacking chieftaincy positions legally recognized by the state – I document and explain a grassroots social movement to proactively invent their own chieftaincy institutions from scratch, an effort that continues right up through the present. Realizing that chieftaincy and intra-ethnic centralization could attract the scarce state’s attention and improve access to its resources, group members are reimagining the past and “inventing tradition” by themselves. Rather than fixed legacies of precolonial state-building frozen in time by colonial indirect rule, as the standard account in the literature often claims, Northern Ghana’s “traditional” institutions remain highly in flux in response to incentives created by the modern state’s continued scarcity.
Correlations between precolonial characteristics and present-day outcomes cannot always be interpreted as active effects of institutions that still persist. It is safer to assume that these institutions have evolved significantly over time through interaction with the state. The mechanisms behind relationships between precolonial institutions and contemporary variables are likely much more complex: colonial and post-colonial states treated communities with different initial features differently, exposing them to different bundles of state action that cumulatively and, potentially quite indirectly, have determined present-day conditions. Rather than “compressing history,” and skipping from the deep past to the present, HPE scholarship will benefit from more carefully investigating what’s happened in between through the gritty on-going process of state building, exploring how local institutions change over time in response to state intervention and absence.
Implications for the control group
A second implication for HPE scholars is to show how and why state policies – such as the introduction of a formal education system – may have non-linear effects across subnational regions with different degrees of state scarcity.
Moving from no schooling to limited access to schooling, as in Northern Ghana in the mid-20th century, may have larger and substantively distinct effects on socio-economic inequality and the distribution of grassroots political power than moving from a little schooling to a lot, as occurred instead in Southern Ghana, a region in which the state funded the creation of orders of magnitude more schools, including indirectly via missionaries largely kept out of the North. Comparing Ghana’s hinterland with its wealthier cash crop zones and urban areas suggests that even though vastly fewer state resources were transferred to society in absolute terms, the relative distributional effects within society could be more severe.
This type of non-linearity complicates our ability to interpret comparisons of places that vary in state presence. As zones of state absence, hinterlands may seem like intuitive reference groups and counterfactuals for evaluating the effects of state-building interventions implemented at a grander scale somewhere else. But the book shows that we can’t look at regions where the state has done less in total and assume it has mattered less for making society look like what it does today. The ostensible treatment and control groups in such a comparison are both moving targets, each potentially changing in different ways in response to different doses of the state. Places with less state are not more free of state effects.