In post-revolutionary Mexico, transport infrastructure meant nation-building for the state, but a threat to identity and traditions for many Indigenous communities.
Large infrastructure projects are a classic tool of nation-building. From schools in Kenya (Miguel 2004) to resettlement programmes in Indonesia (Bazzi et al. 2019), governments use public goods to foster a shared identity and bring remote communities ‘into’ the state.
However, much less is known about what happens when parts of the population do not want to be incorporated on the state’s terms and have the organisational capacity to push back.
Our research (Elizalde et al. 2026) looks at this question in Mexico. After the Mexican Revolution, governments tried to unify a fragmented country by expanding transport infrastructure. For the state, this project was a public good that would modernise the countryside and forge Mexicanidad (i.e. a shared national identity). For many Indigenous communities, however, new transport infrastructure looked more like a public “bad”: a way to undermine autonomy, dilute local institutions, and erode distinct languages and traditions.
We show that where Indigenous communities had a stronger legacy of political organisation, they were more effective at resisting this nation-building push, and that the result was a lasting underprovision of transport infrastructure.
Nation-building through transport infrastructure in post-revolutionary Mexico
Before the Revolution (1910–1920), many Mexicans lived in rural areas, often in semi-autonomous Indigenous communities with their own languages and traditions. The railways built under the Porfirio Díaz regime mainly linked Mexico City to major ports and mining areas, leaving many Indigenous regions physically and politically distant from the state.
Post-revolutionary governments saw this fragmentation as a problem. Their ambition was to reconstruct the nation by unifying citizens under a single national identity and language, namely Mexicans and Spanish language. A central tool was transport infrastructure, which could connect remote areas to cities, markets, and state institutions. Between 1930 and 1960, Mexico built its first national highway network, adding tens of thousands of kilometres of new roads on top of the existing transport infrastructure.
This policy was openly framed as the ‘‘social reconstruction of the nation’’ (Bess 2017). Transport infrastructure was meant to bring remote Indigenous communities into “more direct contact with market forces and the institutions of the state” (Waters 2006). At the same time, historical accounts describe Indigenous communities that opposed such infrastructure because they feared the arrival of outsiders, increased state control, and, above all, the deterioration of their existing forms of social organisation and traditions.
Our research then asks: when and where did such resistance systematically slow down Mexico’s nation-building through transport infrastructure?
Measuring Indigenous communities’ capacity to mobilise
We do not observe, in any systematic way, which communities protested or blocked transport infrastructure as such data is not available. To overcome this, we use history to proxy for the capacity of Indigenous communities to mobilise.
At the time of first contact with Europeans, Indigenous societies in what is now Mexico differed widely in their political organisation, from large, centralised polities to small, fragmented groups. Crucially, these institutional differences did not disappear with colonisation. Colonial authorities concentrated Indigenous populations into small settlements (known as Pueblos de Indios) and often ruled them indirectly through local institutions, allowing traditional governance structures to persist. Indeed, recent work shows that pre-colonial political institutions still influence contemporary outcomes in many parts of the world (Michalopoulos and Papaioannou 2013, Angeles and Elizalde 2017).
We use these differences in pre-colonial political centralisation as a measure of historical institutional capacity: how effectively communities could coordinate and enforce collective action.
We combine this historical variation with a new, digitised map of Mexico’s transport network from the late 19th to the mid-20th century. For every municipality and decade, we track how much area is covered by historical roads, railways, and later highways, and examine where transport infrastructure expanded more or less rapidly. We then compare how infrastructure grew in municipalities with more descendants of historically centralised societies versus those linked to fragmented groups, before and after the nation-building push. In econometric terms, this is a difference-in-differences approach, which lets us see how the policy affected communities with different mobilisation capacities.
There is less infrastructure where mobilisation is stronger
The following findings emerge.
First, there is no evidence that the Mexican state systematically neglected connectivity in Indigenous areas. Before the Revolution, municipalities with Indigenous populations actually had more transport infrastructure on average than those without. This trend continued after the 1920s within municipalities that reported Indigenous populations in the early 20th century.
Second, the divergence in transport infrastructure appears during the nation-building period – and it is driven by Indigenous institutions. During this period, the state expanded infrastructure more where communities had less historical capacity to organise, and less where they were better able to mobilise collectively.
We show that this pattern is not easily explained by geography, state discrimination, or other policies. We control for geographic characteristics that shape the cost and feasibility of building transport infrastructure, and for other nation-building tools such as schools and local bureaucrats. We also construct a counterfactual road network that simply maximises connectivity between major cities. Even after accounting for this “optimal” network, we still see lower infrastructure expansion in municipalities with stronger Indigenous mobilisation capacity.
The evidence points to an important role for Indigenous resistance: where communities were better organised, they were more able to push back against projects that they perceived as a threat to their identity and autonomy.
Why resist transport infrastructure? Identity and collective action
Why would communities resist a public good like a highway? Our reading of historical sources suggests two complementary mechanisms.
Identity preservation
Many Indigenous communities viewed new transport infrastructure as a threat to their identity and traditions. Such infrastructure does not only carry goods; it carries people, ideas, and state agents. If transport infrastructure was seen as an instrument for national integration, communities with greater historical capacity for collective mobilisation would have stronger incentives – and better means – to protect their identities and own ways of living.
We explore this mechanism by focusing on language as a marker of identity. Using census microdata, we find that Indigenous individuals whose ancestors came from more historically centralised societies are less likely to adopt national identity markers, such as speaking Spanish. At the municipal level, these same areas show higher shares of monolingual Indigenous speakers during the mid-20th century.
Stronger collective action
Pre-colonial political centralisation also meant stronger capacity for collective action. Communities with established decision-making rules are better able to organise around shared interests. In earlier work, Elizalde (2020) shows that politically centralised Indigenous communities were more successful at securing the restitution of ancestral land during Mexico’s land reform.1 In our study, we also see a similar pattern but where infrastructure was more intensive, the advantage of centralised communities in land restitution weakened. This suggests that a strong state presence may erode the organisational capacity of Indigenous institutions.
Policy lessons: Building with, not over, Indigenous communities
Our findings speak to broader debates about infrastructure, state-building, and Indigenous rights. They highlight that public goods are not always seen as ‘goods’ by everyone. When transport infrastructure is tightly linked to projects of cultural assimilation, well-organised communities may rationally resist them, leading to slower infrastructure expansion and potentially higher conflict.
They also emphasise the importance of recognising and engaging with existing local institutions. Indigenous communities are not passive recipients of policy; they are political actors with their own governance structures. Policies that try to build over these institutions, rather than with them, are likely to face pushback, especially where historical capacity for collective action is strong.
For today’s policymakers, the message is not that transport infrastructure should be abandoned in Indigenous regions. Rather, it is that how infrastructure is planned, discussed, and implemented matters enormously. Genuine consultation, co-design with local authorities, and respect for cultural markers are not just normative concerns; they are also pragmatic strategies to ensure that infrastructure is actually built, used, and seen as beneficial.
Our results suggest that historical political institutions continue to shape the space for policy today. Understanding where communities have strong internal organisation can help governments anticipate resistance, design better engagement strategies, and avoid repeating past mistakes where nation-building came at the expense of Indigenous traditions.
References
Angeles, L and A Elizalde (2017), “Pre-colonial institutions and socioeconomic development: The case of Latin America”, Journal of Development Economics 124: 22–40.
Bazzi, S, A Gaduh, A D Rothenberg, and M Wong (2019), “Unity in diversity? How intergroup contact can foster nation building”, American Economic Review 109(11): 3978–4025.
Bess, M (2017), Routes of Compromise: Building Roads and Shaping the Nation in Mexico, 1917–1952, University of Nebraska Press.
Elizalde, A (2020), “On the economic effects of Indigenous institutions: Evidence from Mexico”, Journal of Development Economics 147: 102530.
Elizalde, A, E Hidalgo, N Salgado, and S Kampanelis (2026), “Public good or public bad? Nation-building and Indigenous institutions”, Journal of Development Economics 179: 103652.
Michalopoulos, S and E Papaioannou (2013), “Pre-colonial ethnic institutions and contemporary African development”, Econometrica 81(1): 113–152.
Miguel, E (2004), “Tribe or nation? Nation building and public goods in Kenya versus Tanzania”, World Politics 56(3): 327–362.
Waters, W (2006), “Remapping identities: Road construction and nation building in postrevolutionary Mexico”, in M K Vaughan and S E Lewis (eds), The Eagle and the Virgin, Duke University Press.