Property titles and the urban poor: from informality to displacement?

Abstract An extensive literature opposes the provision of property titles for the residents of informal settlements. One concern is that titling leads to commodification and the market-driven displacement of the original inhabitants. Another is that it propagates the ideology of private ownership, undermines collective solidarity and demobilises social movements. This article, based on observations from Mexico City and Guadalajara, finds little evidence of displacement but highlights the importance of location. It supports the view that formalisation undermines resistance, but argues that titling does so by meeting rather than creating the desire of the urban poor for private property and homeownership.


Introduction
The titling of properties in informal settlements has become a battleground of ideas. Opposing positions were already clearly defined well before Hernando de Soto started to disseminate his ideas about formalising the informal as a means to combat world poverty (de Soto, 2000;de Soto, Ghersi, Ghibellini, & Instituto de Libertad y Democracia, 1986). The case for providing titles for "squatter" housing was already sufficiently developed by the early 1980s as to feature in a World Bank publication arguing that titles would bring tenure security, housing improvements and economic benefits, including rental income, capital gains on sale and access to credit (Linn, 1983). The case against was formulated in the 1970s and 1980s by neo-Marxist theorists and political activists from Latin America (Connolly, 2013). De Soto's proselytising about the use of titles to secure loans and enhance productivity, however, brought the debate to a broader audience. As an expression of neoliberal thought, it polarised the debate to such an extent that commentators are now routinely assigned to his camp unless they subscribe unquestioningly to the litany of opposing arguments.
Different positions on titling are therefore lumped together under the "for" or "against" heading. Opposition to formalisation produces unexpected bedfellows. For judges, recognising the rights of "slum dwellers" is like "rewarding pickpockets" (Harvey, 2008, p. 35, writing about India's Supreme Court). City planners and environmentalists worry that it perpetuates informal development (Wigle, 2013), as do activists supporting social production of housing by organised collectives (Ortiz Flores, 2012).
In the battle between entrenched views of titling we rarely hear what residents have to say. 1 One objective of this article, therefore, is to consider residents' opinions of titling in Mexico. With over 2.5 million homes legalised between 1974 and 2013, Mexico has the longest history of mass regularisation in Latin America, and therefore offers a unique opportunity to assess the impact of and responses to titling over a period of four decades. 2 Given the volume of writing focused on de Soto, I adopt a different approach here. 3 I also avoid rehearsing the evidence on housing improvements and access to credit (for overviews, see Gilbert, 2002Gilbert, , 2012. Instead, the article focuses on the main arguments against regularisation. I explore the theoretical framing of these arguments, both as originally advanced and in the current literature, before considering new empirical evidence from Mexico. The economic case against titling hinges on the question of displacement of the original residents that is believed to follow incorporation into the formal market. The political case hinges on the assimilation of residents into an ideology of property ownership and the depoliticising effects of that ideology. There appears, however, to be a potential inconsistency between these arguments. If the consequences of titling are harmful, it is unclear why it would attract support from residents and undermine their propensity to resist. The empirical findings discussed in this paper therefore consider first the evidence for gentrification following regularisation in the Federal District of Mexico City, and second the views of residents from two neighbourhoods that started out as informal settlements in the city of Guadalajara.

Commodification and displacement
The most common argument against titling is what some regard as its key advantage: the incorporation of informal property into the formal market. For its supporters, this ensures fungibility and offers investors a guarantee that they can recoup the value of loans secured against property. For its critics, it means the conversion of homes into commercial assets.
Criticisms of titling appeared almost as soon as the case for supporting "self-help" housing was formulated by advocates such as John Turner (1972). emilio Pradilla (1983, p. 107) argued that titling would lead to property interests "evicting the occupants by purchase" [desalojar a los ocupantes por la vía de la compra], 4 a phrase that anticipates current concerns about the "market-driven displacement" of the original residents (Durand-Lasserve, 2006, p. 207;Payne, Durand-Lasserve, & rakodi, 2009;Smolka & de Larangeira, 2008). Little has changed in the drivers of displacement envisaged. They include the need to pay property taxes and charges both for titling itself and for infrastructural upgrading and consumption of urban services (Alonso et al., 1980;Pradilla, 1983;Durand-Lasserve & Selod, 2009). Tenants see rents increase as landlords pass these costs on to them (Payne, 2001). At the same time, incorporation into the formal land market unleashes "middle class down-raiding" (UN-HABITAT, 2003, p. 35) or gentrification (Payne et al., 2009, p. 449). Poor households may be persuaded to sell and move to more remote, still affordable, areas as a result of these "market-based evictions" (Harvey, 2008;UN-HABITAT, 2007, p. 126).
Consequently, critics argue that "freehold is not the best option" for the urban poor (Payne & Durand-Lasserve, 2012, p. 53). Titling is to be discouraged (Payne et al., 2009). Instead, incremental approaches to regularisation will avoid sudden price hikes and discourage higher-income "raiders". Group title prioritises collective over individual concerns and will "control transfers and discourage speculation" (Payne, 2002, p. 21). 5 The key is to protect residents -against "down-raiders" but also against themselves, by preventing transfers deemed to be against community interests (de Souza, 2004).
The theoretical base of arguments against titling was first elaborated in the 1970s and 1980s by Marxist urbanists from Latin America. regularisation meant "the penetration of commodity relations in land, where hitherto they either did not exist or where they were undeveloped (precisely because they were illegal)" (Burgess, 1978(Burgess, , p. 1120. This was part of a broader critique of self-built housing: by reducing the cost of reproduction of the industrial labour force, self-help kept wages down (Pradilla, 1976). The state would tolerate illegal occupation of land that, given its location or lack of services, was of little interest to landowners. Over time this land would be valorised by the labour invested in building and in obtaining urban services from the state (Alonso et al., 1980). With regularisation the state could recover infrastructural investments by charging taxes, and property/construction interests could capitalise on the valorised asset (Pradilla, 1983). Latin American theorists emphasised, however, that regularisation did not create a market or a commodity where none had existed before.
More recent critiques draw on a different vein of Marxist thought to evoke an image of informality as existing outside the market. They cite David Harvey's assertion of the continuing relevance of "accumulation by dispossession" i.e. by means of "predation, fraud, and violence", including: the commodification and privatization of land and the forceful expulsion of peasant populations [accompanying] the conversion of various forms of property rights (common, collective, state, etc.) into exclusive private property rights (Harvey, 2003, pp. 144−145).
Formalisation of informal settlements has subsequently been described as an instance of accumulation by "enclosure and dispossession" (Porter, 2011, p. 118). It recalls the enclosure movement and its transfer to Britain's colonies (Home, 2004). Such parallels are consistent with the emphasis on collective ownership as an alternative to titling: informal settlements are regarded as "urban commons" (Porter, 2011, p. 117), where rights are based on "social tenures" (Lemmen & Augustinus, 2011, p. 6). An empirical focus on urban Africa reinforces this idea (Varley, 2010). It transfers to the city an endorsement of "customary" tenure informed by a discourse of "African/traditional/good versus Western/new/bad" (Whitehead & Tsikata, 2003, p. 101). Interpretations of titling as dispossession thus draw post/decolonial arguments into the debate, strengthening the tendency to oppose a lost eden of collective values to the pernicious individualism of private property. Titling implies forced dislocation from a communal past.
equating informality with urban communalism is empirically questionable, 6 and casting the formal/ informal distinction in moral terms runs the risk of converting residents into a new type of "noble savage" (Varley, 2013). Pragmatically, it is not clear why, if the intention is to dispossess, people are first given titles, since it is easier and cheaper to evict those who lack registered rights. Both strands of criticism fail to develop a fully convincing account of how displacement occurs. The argument about "market-driven" displacement rests on the assumption that middle-class "down-raiders" or commercial interests will seek to acquire property and that residents will sell, but they appear to do so as expressions of an inexorable structural logic. Arguments depicting regularisation as state support for capital accumulation have been described as instrumentalist and mechanistic; they overlook the variety of state responses as well as the agency of different social groups (Duhau, 2015;de la Peña & de la Torre, 1993). The market is simply all-powerful (Varley, 2002). empirically, the examples of displacement by titling reported are few in number and "unconvincing" (Gilbert, 2012, p. x). even if gentrification is observed, the task of distinguishing the impact of titling from that of upgrading still remains. In addition, evictions following exclusion from a titling programme are sometimes attributed to that same programme, raising questions as to whether titling has gone "too far or not far enough" (Hirsch, 2011, p. 10). The literature nonetheless confidently describes displacement as a "commonly observed" result of formalisation (Durand-Lasserve, 2006, p. 212), as demonstrated by "evidence from most countries" (Otto, 2009, p. 189). Belief in the self-evident nature of displacement resulting from titling has thus become an orthodoxy for anyone wishing to count as a critic of neoliberal urbanism.

Regularisation and resistance
Critics of titling have also decried its ideological and political uses. regularisation recruited new disciples for the "petit-bourgeois" ideology of private property from among the "working masses" (Pradilla, 1983, p. 107). Other theorists saw more political potential in urban social movements (Connolly, 2013); but formalisation has had significant consequences for grassroots organisations and the urban popular movement. The promise of title has been employed to undermine collective solidarity and willingness to confront government agencies in the struggle for infrastructure and services. As each household has had to negotiate individually to acquire title, formalisation has served as a "divide-and-rule" strategy and this atomisation has demobilised opposition (Castells, 1977;Durand, 1983). At the same time, clientelist practices have redirected demand-making along more conservative channels (Melé, 1988;Varley, 1998). Such arguments have, again, been criticised for treating residents as "purely passive" and overestimating the state's regulatory capacities (Duhau, 2015;de la Peña & de la Torre, 1993, p. 105).
The main thrust of the arguments about titling's political uses has remained similar over time. Formalisation saps the energy from community initiatives and reinforces the "socially divisive" idealisation of private ownership (Frediani, 2009, p. 12), leading to a "breakdown of the social fabric" (van Horen, 2004, p. 23). reference to neo-Marxist theorists is now accompanied, however, by another set of theoretical influences, drawn from a postcolonial urbanism shaped by the nomadic thought of Deleuze and Guattari (1987).
In their introduction to six articles investigating the lines of connection between cities in the "global South", Simone and Boudreau (2008, p. 991) note approvingly how the authors "speak of mobilitiesmigration, immigration, resisting colonial emplacement, yearning for a home -and of the crushing weight of state domination, efforts to regularize, normalize, settle down, erase the past and entrench spatial hierarchies". Although none of these authors discusses titling, the quotation nonetheless captures the objection implicit in recent work on urban informality, which employs images of mobility to convey the vitality of informal settlements. Jacques (2002), for example, sees rio de Janeiro's favelas as in-between spaces, constantly moving toward the as yet unknown. Favelas evoke "the idea of shelter [rather than] that of dwelling", because "shelter belongs to the order of time and the provisional; dwelling … the fixed and the permanent" (Jacques, 2002, pp. 20, 57). Informality characterises the "kinetic" city, as conceived by Mehrotra (2010, p. xii): "a fluid and dynamic city that is mobile and temporal (often as a strategy to defeat eviction)".
Mehrotra's suggestion about defeating eviction, although more relevant to street vendors, points towards a broader interest in informality as resistance to emplacement. Authors deploy the idea of nomadic identities to counter the fixity of categories such as "the urban poor", assigned to their place in the "slums" in a binary division between the formal and informal city. Deleuze and Guattari (1987, p. 481) envisage "shantytowns of nomads" as "turning back against the town", and residents of informal settlements are accordingly depicted as urban nomads journeying through these examples of "smooth space which emerges within the striated space of the city but refuses to conform to the rules it attempts to enforce" (Hernández & Kellett, 2010, p. 9;Jáuregui, 2010). Those rules (striation) include formalisation and property tax registration.
The understanding of informality as resistance also resonates with Henri Lefebvre's notion of the "right to the city". Lefebvre (1991, p. 373) cites Latin America's "shanty towns" as exemplars of the defiant appropriation of space generating a rich social life. They assert the primacy of creative use and inhabitance over exchange values, such that the right to the city cannot be "reduced to a positivist legal right" (Butler, 2012, p. 148). Property rights, by contrast, "alienate urban space from inhabitants" (Purcell, 2013, p. 149). It would be difficult, then, to cast Lefebvre as a supporter of titling. Some housing movements have, nonetheless, adopted the right to the city as a banner in their struggles to counter eviction (Huchzermeyer, 2014;Ortiz Flores, 2012). As their aim is to achieve security of tenure, titling is not inconsistent with that aim, insofar as it can provide security of tenure.
The demand for formalisation and for "rights to the city" is itself read as resistance in James Holston's (2008) study of "insurgent citizenship" in São Paulo. Holston describes the demands from informal settlements for services and legal tenure as a claim for "contributor rights" that will acknowledge the contributions residents make to the city via their labour, consumption and taxes (p. 260). Property ownership is the cornerstone of these demands, but also one of the criteria by which citizenship is differentiated: "the very same historical sites of differentiation … are also the conditions of its subversion" (p. 9). In other words, insurgent citizenship both destabilises and perpetuates differentiation. This theorisation draws on a complex understanding of agency as both resistance and "entrenchment, persistence, and inertia" (p. 13).
The following sections consider empirical evidence relating to both of these strands of criticism by exploring first the question of gentrification and displacement in Mexico City, and second, residents' views on titling in Guadalajara.

Titling in urban Mexico
In Mexico, regularisation has been most consistent on ejido land, the product of the country's agrarian reform: some 28,000 ejidos and 2,300 recognised comunidades comprised approximately half the national territory by the 1980s (Assies & Duhau, 2009). Although informal settlements on private holdings and public land have also been regularised, a focus on the ejido is merited because of the scale of urban development on this type of property, and because of the federal government programme set up in 1973 to expropriate ejido land, at relatively low levels of compensation, for the Comisión para la regularizacion de la Tenencia de la Tierra (COreTT). 7 COreTT has arranged for titles to be issued after householders have paid the costs, which have generally been affordable, even though residents have had to pay twice for their land (once in an illegal purchase from the ejido and once for regularisation) (Azuela & Tomas, 1997).

Has titling led to displacement in Mexico City?
The earliest ejido expropriations for regularisation in Mexico City took place at the end of 1970, when land was expropriated from nine ejidos for housing and industrial projects and for the titling of informal settlements. The expropriated areas were located in the State of Mexico municipalities which, together with the Federal District, make up the Metropolitan Zone (Figure 1 inset). 8 Three years later, 374 hectares were expropriated for regularisation in the south of the Federal District. In 1975, one of the first expropriations for COreTT also concerned land in the south of the city. By 2001, over 144,000 titles had been issued by COreTT's Federal District office (Huamán, 2010).
It is, then, 45 years since expropriation started to be used for regularisation in Mexico City, providing an unusual opportunity to observe the consequences in the longer term. Another unusual opportunity stems from the Federal District government making a social development index available at block level for the entire built-up area of some 720 km 2 . Census microdata are used to calculate this index, and the results are classified into four quartiles: very low, low, medium and high. 9 The social development classification of neighbourhoods expropriated for regularisation might suggest that the original residents have been displaced if these neighbourhoods register values associated with middle-class areas. It should be remembered, however, that with the exception of some pockets of enduring poverty, the social heterogeneity of informal settlements "increases gradually and persistently" anyway, because of intergenerational change in educational achievements for example (Connolly, 2003;Duhau, 2014, p. 154). The results presented in Table 1 and Figure 1 show that the great majority (91%) of regularised neighbourhoods fall within the "very low" or "low" categories, accounting for one-half of all housing blocks in the Federal District. There is, then, little sign of outcomes that might reflect the arrival of middle-class newcomers as a result of titling.
Given that the Federal District is the core of Mexico City, with generally higher levels of socio-economic development, it seems unlikely that anything different would emerge from a similar exercise in the State of Mexico.
To gain an idea of whether or not gentrification might have resulted from titling, in 2012 I revisited four areas in which I had worked in the 1980s, in both the Federal District and the State of Mexico. Figures 2 and 3 demonstrate how similar San Agustín, in the municipality of Naucalpan, appears today to how it looked in 1982. The ravines have been filled in, but the main housing difference is the growth in the number of homes with several storeys and the appearance of some three-or four-storey blocks of apartments for rent. These processes are a product of owners building additional accommodation to house the next generation and of the death of the original owners, when heirs sell the property to avoid family conflict over who should live there (Gilbert & Varley, 1991;Varley, 2010). (Similar processes are observed throughout Latin America: see, for example, Ward, Jiménez Huerta, & Di Virgilio, 2015). The increase in renting over time is well recognised, but tenants and owners are generally from similar backgrounds (Gilbert & Varley, 1991); the apartment blocks in San Agustín showed no sign of being occupied by more affluent residents. Another settlement expropriated in the 1970s lacked the apartment blocks, probably because it is in the relatively remote area of Cuautepec, almost surrounded by mountains marking the northern limit of the Federal District; again, it did not appear dramatically dissimilar from how it had been in the 1980s.
One other area that I revisited, Jardines del Ajusco, would certainly, at first glance, seem to fit an apocalyptic vision of gated communities displacing informal settlement. Many streets were closed off, with guards controlling entry. The area offers no support, however, to the displacement thesis. The one neighbourhood in the "high" category in Table 1, it had always been the preserve of wealthy residents, including a former mayor of Mexico City and a Supreme Court judge in the 1980s. They were no doubt attracted by the large (and inexpensive) plots on a mountainside location with sweeping views over the south-west of the city, and could afford their own solutions to the problem of urban services. The final area was particularly interesting: if any Mexico City neighbourhood were to suffer titling-induced displacement, it would surely be this one. In the 1980s San Jerónimo Aculco was an unusual mix of low-and higher-income housing. The settlement occupied 70 hectares next to the south-west  section of the ring road (periférico), a particularly favoured location -indeed, in the 1950s the ejido had supplied the land for nearby Jardines del Pedregal, one of the city's best-known upmarket neighbourhoods. Before the area was expropriated the avenue to the north was already lined with expensive houses, but there were also many poor-quality houses, a few vecindades (rental tenements), and the odd plot where maize was still grown.
In the 35 years since San Jerónimo was expropriated, the periférico has attracted high-end commercial development, meaning that pressures for redevelopment should have been greater here than elsewhere. Some properties have indeed been redeveloped, particularly close to the periférico, where multi-storey apartment and office blocks have appeared. Despite such changes, San Jerónimo is still a mixed area, with contrasting types of housing.
A survey in 2015 asked residents about how they had acquired their property and about changes in the area. 10 One in ten reported an increase in the population, and a similar proportion, rising property values. For some, the fact that newcomers were from wealthier groups was welcome, bringing "less problematic" people, "development" and more jobs; for others, it meant changes in local shops, higher prices and different lifestyles. The most common complaint -raised by one in three respondentsconcerned the poor water supply, blamed by some on the new office and apartment buildings, but almost as many complained about higher property taxes, attributed by one respondent to increased property values (although tax authorities claim to allow for enclaves of different-value properties, as well as the age, size and quality of individual houses). One person believed higher taxes were forcing families to move out.
These changes bear the hallmarks of gentrification. They do not, however, mean that titling caused displacement. Table 2 and Figure 4 show how and when the 210 owner-occupied properties surveyed were acquired. Two in five were bought as land: probably from the ejido in most cases. The fact, then, that a significant proportion of homes have not changed hands implies that many of the original owners have resisted the lure/pressures of the market. In addition, Figure 4 shows no acceleration in house sales after the 1980 expropriation. A dramatic rise occurs much later, around the turn of the millenium. 11 The heterogeneous nature of San Jerónimo Aculco leads to its allocation to the "medium" social development category: one of 13 neighbourhoods in this category (Table 1). If this is the best evidence for gentrification of regularised areas available in Mexico City, it lends little support to the displacement-by-titling thesis. The findings also fit with earlier ones showing that housing market turnover remains slow after legalisation (Varley, 2002;Ward, 2012).
San Jerónimo offers still less support to any suggestion that titling is intended to incorporate such areas into the land market. There were other, much easier, means for capital to access unoccupied ejido land, such as the exchange by which San Jerónimo surrendered land for Jardines del Pedregal (Varley, 1985).

What do residents think of titling in urban Mexico?
If there is little evidence to support the thesis that, in Mexico, regularisation was intended to incorporate informal settlements into the land market, the same cannot be said of its political uses (Melé, 1988;Varley, 1998). However, one question that rarely surfaces about titling as a political strategy is why is it effective?
The most likely answer is that residents already embraced the values of private property and homeownership. That residents welcome titling soon became apparent when I first asked for their views in Mexico City. One much-repeated phrase was "ya somos dueños", "now we are owners": regularisation confirmed the legitimacy of their ownership. The message never varied in subsequent research, but work in Guadalajara in the late 1990s provided an opportunity to pursue the reasons for this response. Fieldwork involved two social surveys and discussion groups in four neighbourhoods, two of them on ejido land: Los encinos, less than 10 years old, and San Mateo, dating from the late 1950s (neighbourhood names are pseudonyms). A survey of women householders and one discussion session (with separate groups of women and of men) elicited owners' views of titling. In San Mateo, most had title, but Los encinos was still untitled. Table 3 summarises survey responses. Fewer than 2% of respondents complained about paying regularisation charges or property taxes. In the discussion groups roberto (Los encinos) complained that "the point of the title is to pay taxes, that's what title's for", but no-one else mentioned this. In Mexico, property taxes have often been charged after services have been installed, but before titling (Durand, 1983;Varley, 1987). 12 Tax demands have been recognised by COreTT as one type of document by which people can demonstrate possession.
Most respondents welcomed titling. Although many did not say why, 27% of respondents in San Mateo, and 47% in Los encinos, mentioned security of tenure; 35% and 28%, respectively, mentioned being recognised as an owner ( Figure 5).
"Security of tenure" meant enjoying the protection of the law. As Francisco reported in Los encinos, "it's 100% yours now -who's going to take it off you? you've got the proof in your hand, it's your authority [el poder tuyo]". The most common response was "now they can't take it off you"; but who are "they"? It turns out that "they" are people occupying the house temporarily, as tenants or because they have been lent accommodation (as a favour or while the owner is away). As rodolfo argued, to rent without a title, "it has to be someone you can really trust, because if there are no deeds someone can (…) put it down in the name of whoever's living there". This finding supports the suggestion that the supply of rental accommodation might expand once potential landlords have a title (Gilbert, 2002). It also shows that the threats that worry residents come not so much from the state or big business as from rather closer to home (Varley, 2002(Varley, , 2010. They are not idle threats. One discussion-group member in San Mateo had lost a property when he lent it to someone who then got hold of papers recording himself as the purchaser. The notion of title as protection against dispossession is different from title as recognition of what I have called "authentic ownership" (Figure 5). Title "gives the house value [valor]", but not necessarily monetary value. In Los encinos, Pancho asserted that "without title, the house has no value", to which someone replied "yes it does, if you know it's yours". In San Mateo, Adolfo said that "someone who doesn't have title is building on thin air [literally, on a vacant plot, baldío]". When said of a person, tener valor also means "to be courageous", and participants ascribed a moral dimension to title. It was reflected in comments about being a "legitimate" owner, having right on one's side, acquiring title as the "proper" or "correct" thing to do, or even benefiting from "a proper legalisation" (Figure 5). 13 Policy debates about titling focus on function, and research participants certainly mentioned the practical consequences; but Table 3 shows how much weight they placed on the ideal of ownership. These different dimensions of ownership are not independent. Figure 6 reveals how ideas about security or being able to sell overlap with moral concerns about one's identity as a named homeowner and about the authenticity of that self-possession -about property as "propriety" (rose, 1991, p. 223). The clearest illustration of property as propriety comes from Margarita's equation of her title with her birth certificate, without which she is "nobody". Property is not just "a bundle of rights": it means having a recognised place in the world, a position in the social hierarchy, as a self-contained and sovereign individual (Davies, 1998, p. 147;Davies, 2007). A property owner is not just "anybody", or (for women) una cualquiera: a term which also points to questions of sexual propriety (and arguably, given the significance of marriage in Mexico's colonial history, to racial "propriety" too) (Haney, 2012;Varley, 2010). residents' enthusiasm for titling also expresses a deep-seated appreciation of dwelling as positive value and, conversely, a normative rejection of mobility (Varley, 2015). They use the phrase andar rodando ("to go rolling around", like a stone rolling downhill) to convey their dislike of renting or sharing with relatives. By contrast, they express their satisfaction with homeownership in terms of rootedness: "you're on firm ground, no-one can move you (…) It's like planting a tree" (2015, p. 288).
These narratives of home as mobility and dwelling demonstrate, moreover, that the concepts are more complex than recent celebrations of mobility suggest. Andar rodando is an undesired mobility, which should not be understood literally, but rather in terms of the threat of eviction. It means, at the same time and paradoxically, a lack of movement, an inability to develop one's own (shared) narrative of home: running to stand still, going nowhere fast. Those who subsequently became homeowners recount their experiences in ways that suggest a release from this running in place and an ability to "move on". Home ownership, for them, represents what has been called the "movement inherent in dwelling" (Pratt, 1999, p. 159).
Two qualifications are needed. First, rejecting unwelcome mobility does not mean people would tolerate being prevented from moving, as demonstrated by the failure of restrictions on the sale of social housing in Mexico (Azuela & Duhau, 1998). Second, there is growing evidence that the intestate death of the original owners is leading to a "new informality", in Mexico and elsewhere (Varley, 2010, p. 92;Dubé, 2015;Ward et al., 2015). To interpret this as lack of interest in title is mistaken: people want public recognition of ownership, whatever their private arrangements (Varley, 2002(Varley, , 2010.

Conclusion
The evidence from Mexico City considered in this article yields little support for the thesis that titling leads to displacement. Fears of the rich or of businesses queuing up to buy out the original residents once they have title seem unlikely to be realised in urban Mexico at least, unless, perhaps, informal settlements occupy exceptionally attractive sites. This article has identified one very well-located area in Mexico City that is now experiencing gentrification, but it remains a mixed area 35 years after expropriation, and location, not title, is the key to what is happening in San Jerónimo. The great majority of regularised settlements on former ejido land remain in the low or very low social development categories. It seems likely that the exceptions never featured in these categories: San Jerónimo and Jardines del Ajusco started out as a mixed and a wealthy area, respectively.
Location, then, is crucial, but the sites on which informal urban development is tolerated are generally of scant interest to capital. The investment of inhabitants' labour in building and consolidation does not create a rent gap large enough to attract rival land uses.
The significance of location is underlined by the case of one Latin American city where gentrification of informal settlement is currently being reported: rio de Janeiro. The location of some of rio's favelas adjacent to wealthy neighbourhoods, on sites with extraordinary views of the coast, means that their "pacification" has unleashed strong gentrification pressures. Titles are not, however, to blame: successive efforts to regularise tenure in rio's favelas have achieved very little (Perlman, 2016). 14 Only where unusual locational advantages beckon, then, might commercial interests find it worth their while to buy out the owners of large numbers of small individual properties. The likelihood of their being willing to do so is reduced where there are means of obtaining unoccupied land cheaply: for example, by taking advantage of ambiguous "public interest" clauses in expropriation laws (Varley, 1985). expropriation can also be used to assemble land for redevelopment in existing urban areas. If the demand arises, a state-led process of expropriation is surely a more likely route to displacement than a multitude of individual market transactions.
Critics have argued that, rather than enhancing security of tenure, titling may diminish it. Four decades after a national programme of ejido land regularisation got underway in Mexico -decades during which people have had plenty of opportunity to hear about the experiences of parents or even grandparents -they still appear unconcerned about any such threat. residents actively seek titling (Varley, 1998). This might seem curious, given that the inhabitants of informal settlements on ejido property have generally enjoyed a relatively high degree of de facto security (Azuela, 1989;Varley, 1987). That they should nonetheless be so positive about title underlines the extent to which, far from refusing to conform to the rules of the urban order, residents actively uphold the values of ownership. They do not do so as a result of regularisation. The process -including the public ceremonies where government officials, including the president, distribute titles to homeowners -reinforces those values, telling people they were right to struggle for a home of their own and offering compensation for the suffering they have endured. Titling does not create those values in the first place.
In the final instance, regularisation cannot address the scarcity of affordable land, and it should never be a substitute for the provision of urban services. What is needed, ultimately, is indeed "a broader notion of property that includes not only the rights of the owner but also … the rights of the community" (Blanco, 2011, p. 22). The idea that informal settlements should serve as a laboratory for the emergence of a new model of property is, however, problematic. It risks creating "second rank" property rights for the urban poor (Durand-Lasserve & royston, 2002, p. 254). Where residents seek or welcome such alternatives, these experiments may rightly come in to their own; but where they do not, a reconsideration of blanket opposition to titling is in order.

Notes
1. For an exception, see de Souza (2001). Others summarise survey results but do not discuss individual views. 2. In Mexican cities, legalisation or titling is called "regularisation". I also use "formalisation", although it can encompass alternative approaches (see text). The figure refers to ejido land (legally inalienable land granted to rural communities in Mexico's agrarian reform), but settlements on both private and public land have also been regularised (Mexico, 2013). 3. De Soto's arguments are scrutinised by Bromley (2004), Fernandes (2012), Home (2004) and Benda-Beckmann (2003). 4. All translations are mine. 5. Disaffection amongst those prevented from selling means the approach "may not be viable over the long term" (Durand-Lasserve & Selod, 2009, p. 25). 6. For example: the work of de Sousa Santos (1977) on legal pluralism in Brazil is sometimes taken to mean that favela property systems express different values; but they are "based on respect for the substantive principle of private property" (p. 89). 7. regularisation of private land has been a local and more ad hoc undertaking. In the last few years, less effective approaches have also overtaken expropriation of ejido land (Salazar, 2012). 8. In January 2016, México Distrito Federal officially became Ciudad de México (CDMX). 9. Components and their weighting are: housing quality and size 0.338, access to healthcare and social security 0.291, education 0.244, consumer goods 0.060, sanitation 0.038, and access to electrical grid 0.029 (Consejo de evaluación del Desarrollo Social del Distrito Federal, 2011, p. 39). 10. each house in 16 randomly selected blocks was visited. In 10.7% of 280 houses visited, occupants refused to participate; in another 10.0% no-one was in. Of 222 households surveyed, 210 (94.6%) were owner-occupiers. The 2000 census recorded 7,918 people in 1,919 households (79.3% owners); 39% of those employed received two minimum salaries or less, but 22% received more than five (data from http://www.sideso.df.gob.mx/documentos/ut/MAG_08-036-1_C.pdf ). 11. This dramatic rise may reflect the introduction in December 2000 of a Federal District policy restricting new development in peripheral delegaciones such as Álvaro Obregón, where San Jerónimo is located, resulting in an immediate marked increase in land values (Guadarrama Flores, 2007). 12. The same applied in the areas of São Paulo studied by Holston (2008, p. 262). 13. São Paulo's "insurgent citizens" shared the ideal of authentic ownership: "I do not want to live for free; I want to be the owner of my lot; I want to be the legitimate owner" (Holston, 2008, p. 174; added emphasis). 14. By contrast, a study of two regularised favelas in metropolitan São Paulo revealed no gentrification or displacement of owners, although the authors speculate that tenants might have been evicted prior to regularisation, to prevent their claiming acquisitive prescription (Silva & Mautner, 2016).
anonymous referees and Charlotte Lemanski pushed me to think further and prompted me to revisit the 1980s literature; both were rewarding exercises.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
The research in Guadalajara was funded by the UK economic and Social research Council, research Grant r 000 23 6808, "Gendered Housing: Identity and Independence in Urban Mexico".

Notes on contributor
Ann Varley is Professor of Human Geography at UCL (University College London). She has worked on security of tenure for Mexico's urban poor since the 1980s. She was awarded the Busk Medal, one of the senior awards of the royal Geographical Society, in recognition of her fieldwork in low-income communities in urban Mexico.