“[W]hile there are always things to differentiate between national groups,
for me the social question has most to do with where you live”
(Gabriel, Paraguayan migrant in Buenos Aires)

Buenos Aires is incredibly divided; a ‘global city with global slums’. Consistently stigmatised, its villas miseria (urban ‘informal settlements’) are home to well over half a million people, many of whom are migrants from nearby countries. But these are also spaces of hope, resistance, and prefiguration, home to an inspiring array of movements from the piqueteros emerging in the 1990s, to today’s powerful feminist movement.1 As the opening quotation highlights the spatiality of the villas is extremely important: if, where, and how you live in the villas can shape not only attitudes towards you, but also the nature of your interactions with others. The contours of the ‘informal’ housing market are central to this. Rising rents are increasingly embedding the villas within processes of accumulation, housing conditions shape political (in)activity, and the reliance on kinship networks to secure housing reproduces national divisions. Or, put another way, housing is central to technical, political, and social compositions.

By exploring these issues, this article shows the importance of a spatial composition analysis, not only for periodising previous cycles of resistance, but also for helping to build for future struggle.2 Autonomist thinking has long had an implicit spatiality, with a focus on urban politics and the idea of the ‘social factory’ where class struggles are diffused across the city.3 Crucial as these spatial insights are, however, they have tended to remain underdeveloped, and so the spatial compositions framework thus combats this, helping to develop the idea of the social factory: it shows that while housing (as well as broader infrastructures of social reproduction) is vital to class recomposition, it can also drive class decomposition. Focusing on migrants and their struggles this article also shows how these spatial compositions are fundamentally intersectional and can both transcend and reinforce differences. And linked to this, the focus on Buenos Aires is part of a wider project to push autonomist thinking out of its geographic comfort zone, hoping to rework and ultimately improve it.4 To make these points the article begins by introducing the concept of spatial composition, which is then used to analyse migrants’ housing situation in the villas, based on nine months of research with activists and migrants. After this the article moves beyond housing to explore the everyday infrastructures of social reproduction that make up the social factory, highlighting how – although these often extend and deepen capital’s control over labour – if reworked they are potentially fertile grounds for spatial recomposition.

From class to spatial composition

Steve Wright5 has argued that class composition analysis is the key contribution of the autonomist project, with its focus on technical composition (how capital organises labour power) and political composition (how the working class organises – or not – against capital). Technical compositions thus seek to discipline labour and maximise accumulation, while political compositions exist immanently to these. But class composition analysis reflects Mario Tronti’s ‘Copernican inversion’ of the dialectic between capital and labour: rather than capital being on the front foot and labour responding, the polarity is reversed, and technical compositions are responses to labour militancy.6 Class composition analysis therefore seeks to trace capital’s shifting conjunctures (technical compositions), locating weak points and the forms of struggle that are most appropriate to that point (political compositions). When the working class is on the offensive, uniting our struggles and gaining concessions from capital, this is understood as ‘recomposition’, while the inverse is ‘decomposition’. And while a focus on recomposition is understandably alluring, honest and reflexive analyses of decomposition is similarly crucial, helping us to learn from past mistakes.

The framework has been rightly expanded through a focus on ‘social composition’, emphasising the role that social reproduction plays in shaping working-class subjectivities (as a site of both control and resistance), as well as through research which highlights the intersectional nature of all compositions.7 But despite the implicit spatiality of much important autonomist work – the layout of conveyor belts being central to (post)Fordist technical compositions; Negri’s focus on the ‘metropolis’;8 and, most famously, the ‘social factory’ – the geography of class compositions has remained relatively underexplored. Recent work, however, is starting to combat this by focusing on ‘spatial composition’ in detail.9 The idea here is that urban space is never neutral, instead reflective of prior intersectional class struggles. Urban space can be created to try and simultaneously limit struggle and maximise accumulation (a technical spatial composition), but diverse forms of urban struggle (political spatial compositions) are always immanent.

Examples from Buenos Aires help to make this point. In an attempt to limit the power of Peronists and the trade unions, the military dictatorship of the 1970s and 80s actively deindustrialised Buenos Aires, not only limiting the organisation and power of the working class, but by shifting circuits10 to a focus on housing and real estate ushered in a new period of accumulation. This deindustrialisation and spatial shift of capital, however, created the conditions for the piqueteros and unemployed workers’ movements to emerge in the 1990s, who targeted the circulation of capital with explicitly spatial forms of class politics. Their road blockades and neighbourhood organising played a central role in bringing down the government in 2001, and also had links to the expanding worker-occupied factory movement, which again emerged because of, not despite, deindustrialisation. These movements drove a powerful period of recomposition, developing extensive alternative economies, ensuring significant state subsidies, and undertaking ‘everyday revolutions’11 under the banner that the ‘neighbourhood is the new factory’. Adopting explicitly spatial forms of class politics, these movements emerged in-against-and-beyond a seemingly totalising and brutal state-capital nexus. Compositionally speaking, therefore, the political spatial composition was immanent to the technical spatial composition, with recomposition following sustained decomposition. But there is always a counter-movement, technical class compositions shift in response to working class militancy, and technical spatial compositions are no different.

Housing, ‘informality’, and migration in Buenos Aires

Following the 2001 economic crisis and default, Argentina saw a sustained period of growth and consolidation (although from a low base). But while levels of unemployment and extreme poverty fell, levels of inequality rose – the number of people living in the villas and those deemed to be middle classes had each grown by more than 50% by 2011. This ‘elite-led recovery’ was intimately linked to what has been described as a form of expanded extractivism. As Argentina’s rural areas developed a renewed reliance on oil and gas, this was matched by the growing importance of rent in urban environments. The federal capital has seen average rents increase over 1500% since 2001 – from $12.7/m2 to $224.1/m2 – forcing the urban poor into either the conurbano (the urban periphery) or/and the villas. This gentrification has been policed violently, targeting racialised communities in an attempt to whiten the city and enable rents to increase further. A prime example of capital switching to an urban circuit, this has created an incredibly uneven housing crisis. As mentioned, the pressures placed on the ‘formal’ housing market have driven a larger number of people into the ‘informal’ sector, and consequently ‘informal’ rents are not only rising faster than those in the ‘formal’ sector, but my research found that they can also be higher. This has disastrous impacts, especially for migrants, who often lack the savings and documentation to rent ‘formally’ – the people who can least afford these rent increases, are having to deal with them most. But the consequences go far beyond this, showing the importance of a detailed analysis of technical spatial composition.

Migrants’ housing is typically reliant on kinship networks, especially for those who have recently arrived and lack the knowledge/contacts for anything else. This has two immediate, interrelated, consequences. First, it can consolidate national differences between migrant groups, and second it empowers a small selection of migrant landlords. The first of these therefore shows how housing can play a central role in strengthening the social composition within individual migrant groups, but consequently interaction between different groups is limited, ultimately leading to an intersectionally-striated class decomposition. But even more troublingly, the bonds created among migrant groups cut across class lines, creating a false sense of solidarity among landlords and their tenants. The spatiality of these political and social compositions is even more significant, however, when focusing on Bolivian migrants working in the textile industry.

Talleres clandestinos (clandestine textile workshops) are big business in Buenos Aires, employing well over 300,000 people, including 80% of Buenos Aires’ ‘working age’12 Bolivians. The talleres have a particular and intentional technical spatial composition, being spread out across the city in small units where many people live and work in the same building. This has a number of consequences. First, it makes it harder for workers to share their experiences and organise, being so fragmented and isolated. Second, it insulates the owners of the talleres from state-led closure, as workshops are easily moveable and effectively sacrificial if the authorities need a tokenistic appeasement. Third, housing migrants within their workplace affords workshop-owners not only an extra level of control over their workforce, but also the ability to extract further rent. In this sense the talleres have a distinct technical composition that adopts a clear strategy to minimise resistance and maximise accumulation, but this technical composition is explicitly spatial. Housing a workforce in the talleres makes these migrants even more vulnerable and beholden to their employers/landlords who, in many situations, have also been involved with the trafficking of these very migrants, leveraging debts incurred from this cross-border process to lock them into extremely exploitative work.

Given the housing crisis in Buenos Aires, and migrants’ inability to avoid the ‘informal’ housing market, a perfect storm is created. Living and working in the same buildings means that migrants suffer from an almost total subsumption, limiting the possibility of political recomposition – something worsened by the false sense of cross-class national unity promoted by workshop owners to try and convince those they exploit that their shared Bolivianess makes other nationalities the enemy. What we see here, therefore, is the crucial role that housing plays, not only in Buenos Aires’ technical spatial composition, but also in driving a nationally-bound social spatial composition which limits the potential of any political spatial recomposition. As such, not only should housing be considered more centrally in compositional analyses, but it is clear what (spatially) compositional analyses can bring to political discussions of housing. But it is important to think beyond housing and consider other facets of (migrant) life – the everyday socially reproductive infrastructures that are simultaneously sites of resistance and control.

The social factory

For migrants and the racialised urban poor, life in Buenos Aires has real spatial constraints. Their use of ‘public’ space is lethally policed, and their presence in much of the city is tolerated only in so far as it fulfils supposedly servile roles. Places and networks of comfort and security are therefore extremely important, helping not only new migrants settle in, but to also provide those who are more settled with crucial respite. As Mayra, a Bolivian migrant, told me: “it is only [in places like this] that we feel happy and safe to meet, to plan, to talk, to sing, to dance”. But important here are not only the neighbourhoods where large migrant populations live, but also the shops, markets, street stalls, restaurants, radio stations, health clinics, and discos which migrants use. These infrastructures of social reproduction are central to the social life/composition of migrant groups, enabling them to explore affinities outside of either the home or workplace (which, for many, are synonymous). Analysis of housing is, while necessary, therefore insufficient when unpacking socio-spatial compositions. It is vital to consider these wider social infrastructures, especially given the fundamental ambivalence of the social factory.

As shown, housing can foster solidarity within, but animosity between, migrant groups, and Buenos Aires’ social infrastructures consolidate this. The discrimination and pressures migrants face drive this reliance on internal networks, but consequently these migrant groups can become even more hermetic, as Marcos, a militant and researcher, explained:

Migrant collectives are normally quite separate…they don’t often mix spatially…Because migrants come here, and there […are] very defined places for them to live…and [they have] their own networks, their own radios. They keep themselves to themselves, and they have their own expressions of culture. So it is very easy to arrive to the country and the first people you join with are the same as you, and you talk about politics…or football back in your country. It is really easy to get into that dynamic and not interact with other groups.

Social compositions are thus not only explicitly spatial, but fundamentally framed by housing situations and the wider networks these are part of. For Bolivians, however, just as the majority of their housing is owned by their cabal of employers, so is much of this social infrastructure. Once more, this has multiple consequences. Not only does it allow these landlords/employers/business owners to further expand extractivism, recapture wages, keep money circulating within, and thus further consolidate the Bolivian community. But it also extends the level of control exerted over Bolivian workers, many of whom live, work, and relax quite literally under the watchful eye of powerful community members - this also has important consequences for analysis of social reproduction, further challenging neat binaries of work/home or the public/private sphere.

This shows the panoptic qualities of the social factory, where a deliberate socio-spatial strategy has been developed to limit dissent and increase accumulation (housing and social infrastructure are thus part of technical spatial compositions). And detailed spatial composition analysis can therefore emphasise how this is tending towards decomposition, as political and social spatial compositions are obfuscating class divides within migrants’ communities and limiting solidarity between them. At the heart of compositional analyses, however, is the immanent potential of political compositions. In the face of even the most punitive technical compositions, green shoots of recomposition can be found, and this is no different in Buenos Aires.

Potentially the most powerful example here is the inspiring feminist movement, leading the struggles for abortion rights and against femicide, while also organising millions-strong women’s strikes. Many of the most powerful groups in this movement are based in the villas, and migrants make up much of the movement’s membership. One of the reasons the movement has been so successful is that the focus on gender can transcend nationally-bound differences perpetuated by the processes described above, emphasising the need for compositional thinking to always be intersectional. While another factor is the focus on social reproduction, making often overlooked forms of labour visible, and showing how housing and social infrastructure (the social factory) is central to these movements. Another key area of struggle and recomposition has been the recent series of migrant strikes, where, although variegated, some the groups involved have started to push back against the owners of the talleres. To do this they have leveraged aspects of these social infrastructures – most obviously the radio networks – to organise. Migrant groups I worked with explained the centrality of story, film, and music workshops to their political organising, and how they built class awareness through communal meals and events. This not only reminds us of the important leap from social to political compositions, but that these are always spatialised and immanent to their technical counterparts.


Housing is vital yet ambivalent. The home is a site of struggle, where we see oppression but also great resistance, and as capital switches circuits rent extraction is increasingly important to wider processes of accumulation. This article has also shown how in the case of Buenos Aires housing crises can be exploited by landlords-cum-employers to combat militancy and obfuscate class structures along national lines. Housing is thus central to wider technical compositions, but as this article has shown these can be helpfully understood as technical spatial compositions, showing the importance and value of geographically-attuned autonomist thinking and political organising. Moving beyond the focus on technical compositions, housing in Buenos Aires’ ‘informal’ urban settlements also fundamentally frames a social spatial composition that, in many cases, leads to political spatial decomposition. National differences are reproduced by the housing situation in Buenos Aires, and so migrants’ class compositions are not just spatial, but also intersectional. Housing is, however, also part of wider social infrastructures, and for migrants in Buenos Aires these can perpetuate exploitation and increase extraction. Vital support and survival networks are often owned by the same compatriots who exploit migrant workers at work and at home, reinforcing the contours of these national groups. But this infrastructure can be repurposed and weaponised, and it is these socially reproductive struggles that have the strongest recompositional potential. Social infrastructures therefore make up the ‘social factory’, a key, yet at times nebulous, autonomist insight, which shows how both resistance and control are spread across society and well beyond the workplace. Spatialising compositional analysis enables us to hone and unpack the social factory, helping to realise the concept’s powerful potential, and to emphasise the mutual benefits to better dialogues between geographical and autonomist thinking.

  1. Marina Sitrin. 2012. Everyday Revolutions: Horizontalism and Autonomy in Argentina. London: Zed Books; Raúl Zibechi. 2012. Territories In Resistance: A Cartography of Latin American Social Movements. AK Press: Oakland, CA. 

  2. For a more detailed discussion of this concept and overall argument see: Nick Clare. 2019. Composing the social factory: An autonomist urban geography of Buenos Aires. Environment and Planning D: Society and Space, 37(2): 255–275. https://doi.org/10.1177%2F0263775818805096 

  3. Neil Gray. 2018. Beyond the Right to the City: Territorial Autogestion and the Take over the City Movement in 1970s Italy. Antipode, 50(2): 319–339. https://doi.org/10.1111/anti.12360 

  4. Stevphen Shukaitis. 2014. Learning Not to Labor. Rethinking Marxism. 26(2): 193–205. https://doi.org/10.1080/08935696.2014.888835 

  5. Steve Wright. 2017. Storming heaven: Class composition and struggle in Italian autonomist Marxism (Second Edition). Pluto Press: London. 

  6. Mario Tronti. 2019. Workers and Capital. Verso Books: London. 

  7. David Camfield. 2004. Re-Orienting Class Analysis: Working Classes as Historical Formations. Science & Society, 68(4): 421–446. https://www.jstor.org/stable/40404795; Elise D. Thorburn. 2017. Cyborg witches: class composition and social reproduction in the GynePunk collective. Feminist Media Studies, 17(2): 153–167. https://doi.org/10.1080/14680777.2016.1218353 

  8. Antonio Negri. 2018. From the Factory to the Metropolis: Essays Volume 2. Edited by Ed Emery. Polity Books: Cambridge. 

  9. Clare. 2019; Gray. 2018. 

  10. For an overview (and update) of Henri Lefebvre and David Harvey’s original arguments on capital switching see: Brett Christophers. 2011. Revisiting the Urbanization of Capital. Annals of the Association of American Geographers. 101(6) 1347–1364. https://doi.org/10.1080/00045608.2011.583569 

  11. Sitrin. 2012. 

  12. Although given the huge number of cases of children working and even dying in the talleres the idea of a ‘working age’ is contentious. 


Nick Clare

Nick Clare is a currently based at the University of Nottingham. He is interested in the intersections between autonomism and geography, and has carried out research in Latin America and the UK.