The Bureaucratic Politics of Urban Land Rights: (Non)Programmatic Distribution in São Paulo ’ s Land Regularization Policy

How do bureaucrats implement public policy when faced with political intermediation? This article examines this issue in the distribution of land rights to informal settlements in the municipality of São Paulo, Brazil. Land regularization is a policy established over three decades, where politicians ’ requests for land titles to their constituencies play a relevant role. Based on interviews and documents, this study finds that bureaucrats adopt a twofold approach to regulate distribution: they document informal settlements, enacting eligibility criteria; then, they manage and prioritize beneficiaries, accommodating qualifying political demands. In this process, they enforce eligibility rules consistently across cases, constraining political intermediation to a rational scheme. Therefore, bureaucrats reconcile nonprogrammatic politics and policy rules by separating eligibility assessment from beneficiary selection. This paper bridges urban distributive politics and street-level bureaucracy literature by revealing that policy implementers may use technical expertise to curb political influence and negotiate conflicting interests and constraints.

facilitating procedures and skipping steps (Krishna et al. 2020).Yet, the case of land regularization in São Paulo confounds this expectation.
In Brazil's largest municipality, with a population of 11.5 million, about three million people live in informal settlements lacking legal tenure.The local government carries on a policy of land regularization for low-income communities, granting titles to households in favelas and irregular subdivisions (Marques and Pulhez 2021).In the city, brokerage is a common practice adopted by municipal councilors to provide services-including land regularization-to their constituencies and build political representation (Hoyler 2022).These municipal legislators intermediate citizens' demands for land tenure security and push the housing bureaucracy to deliver land titles to the informal settlements among their electorate.Such "particularistic" requests may conflict with rational-bureaucratic procedures, which entail a standardized and technical treatment applied to all cases.However, bureaucrats implementing land regularization in São Paulo break down this tension by accommodating some political demands without circumventing rules.How do bureaucrats enact policy rules and partially accommodate nonprogrammatic requests?
This paper argues that the conflicts between distributing land titles based on political intermediation or programmatic criteria may be conciliated by the dissociation of assessing eligibility rules and allocating land titles.Bureaucrats control the scrutiny of qualifying cases, preventing political influence at this point.As the number of eligible informal settlements is larger than the bureaucracy can process, cases queue up.The selection of recipients then entails a compromise between political intermediation, bureaucratic discretion, and legal administrative constraints.
This raises yet another question: what explains why a noninsulated implementation bureaucracy may impose limits on political influence?Bureaucrats use technical expertise and legal accountability to safeguard the assessment of qualifying cases from political interference.They enact eligibility rules in documentation practices, that is, in the technical procedures of collecting, creating, and analyzing records of informal settlements.Municipal civil servants are legally responsible for these documents and use both their capacity to operationalize them and liability to deter external influence.In doing so, bureaucrats uniformly apply these overarching rules to all cases, thereby containing political demands to a rational scheme.These findings are grounded in documents (e.g., legislation, plans, reports) and interviews conducted with bureaucrats, executive officials, city councilors, civil society, and law enforcement actors in São Paulo.
This study sheds light on the interface between politicians and bureaucracies, a key issue for Latin American political scholarship, seldom examined in the context of policy implementation (Polga-Hecimovich and Trelles 2016;Polga-Hecimovich 2019).Urban distributive politics literature demonstrates the relevance of brokerage networks for citizens to make claims on the state in Global South cities.Still, it is unclear how and why political intermediation influences bureaucratic action.The street-level bureaucracy literature, in turn, suggests that where brokerage is common, frontline workers may engage in predatory behavior or put their principals' interests first, hardly withstanding political influence.This paper bridges these two strands of literature by analyzing bureaucrats' interactions with political intermediation in the case of land regularization in São Paulo and revealing that policy implementers may use technical expertise to curb political influence and negotiate conflicting interests and constraints.Moreover, it provides a novel account of land regularization, exploring processes overseen by studies mainly based on documents.
The article proceeds as follows.After this introduction, section 2 discusses the extant literature and section 3 presents the methodology.Section 4 provides a background on land regularization policy in São Paulo and its bureaucracy, followed by the main empirical results.Section 5 shows how bureaucrats enact eligibility criteria in documenting informal settlements; section 6 reveals how bureaucrats manage and accommodate nonprogrammatic demands; and section 7 considers the dissociation of selecting eligible and beneficiary cases and explores the drivers of bureaucrats' control over the first.The conclusion ensues.

Brokerage for Services and Land Tenure in Global South Cities
A burgeoning literature on distributive politics in Global South cities underscores the relevance of brokerage for the delivery of goods and services (Post 2018;Le Galès 2021).Brokers help to operate informal land markets (Paller 2015;Krishna et al. 2020;Tellman et al. 2021) and their distribution and density may affect residents' claims-making capacity, as Auerbach (2019) reveals in India.Similarly emphasizing citizens' agency, Alvarez Rivadulla (2012) shows that squatters in Uruguay seize opportunities to access basic needs such as water and electricity by strategically engaging in political networks.
These recent works review and expand previous research on politics in Latin American cities, which explained the roots of clientelism among the urban poor (Auyero 2000), and the varying types of relations between informal settlement dwellers and politicians (Gay 1994).As the quid pro quo typically equated to clientelism is rarely found in empirical accounts (Hicken and Nathan 2020), scholars underscore the lack of conceptual fit and explore the persistence of nonprogrammatic politics in varied and complex local arrangements (Rivadulla 2012;Hoyler 2022), revealing that legislators and other politicians may provide constituency services to allow citizens to benefit from services they are entitled to but can hardly access (Bussell 2019;Hoyler 2022).As Heller and colleagues (2023, 84) argue, research on brokerage may thus benefit from differentiating between "(a) highly dependence-inducing and autonomy-eroding (classic clientelism), (b) asymmetrical but strategic and competitive, or in fact (c) approximates [of ] constituency service," considering their divergent consequences for democracy.
Previous literature finds that brokerage is relevant for securing land tenure in informal settlements in at least two ways.First, it is associated with protection against evictions.Politicians may opt not to apply land use regulations and property rights, guided by welfare purposes and the minimization of substantive electoral costs related to enforcement (Holland 2017).Insecure land tenure can lead residents to rely on political brokers for protection against evictions and access to services when these are conditional on land rights.In these situations, land titling may contribute to breaking this relationship of dependency (Larreguy et al. 2018).
Second, brokerage plays a crucial role in facilitating access to formal land tenure through land titling or regularization.These procedures are often much more complex than commonly assumed.They are subject to various regulations, developed in stages, and can be time consuming.Krishna and colleagues (2020) find brokers expedite land titling in Bengaluru, India, by reducing the time bureaucracies take to process cases, shortening and sometimes skipping stages.Opportunities for political intermediation arise from the ambiguities of this complex process, which involves various tenure statuses documented in multiple papers issued by different, uncoordinated authorities.They conclude that "[i]ntermediation, rather than objective criteria and legal provision, helps explain which slum gets provided with what type of paper."(Krishna et al. 2020(Krishna et al. , 1991)).
Latin American urban scholars have made related claims regarding land regularization.Varley (1994) documents the early experience of land regularization in Mexico in the 1970s, calling attention to the targeting of specific groups of beneficiaries: residents of informal settlements affiliated with the Institutional Revolutionary Party (PRI) and squatters organized into social movements.Similarly, other authors assume the particularistic allocation of land titles to residents of informal settlements in Brazilian or other Latin American cities (Fernandes 2011;Gilbert and De Jong 2015), suggesting that political intermediation opportunities lie in bureaucratic deadlocks (Rolnik 2019).
In summary, prior research highlights the significance of context-specific brokerage relations within urban settings, and specifically, their role in securing land tenure and delivering land titles.While this body of literature primarily delves into the interactions between brokers and citizens, there remains a knowledge gap concerning the dynamics between bureaucrats and brokers.For intermediation to work, it needs the cooperation of the bureaucracy.Besides political influence, bureaucrats are constrained by legal rules and norms of practice.Still, they often retain a degree of discretion when it comes to implementing policies.The following section discusses the literature on bureaucratic politics and street-level bureaucracy relevant to the present case.

Bureaucracy and Politics at the Front Line of Public Service
Bureaucracies in Global South countries are seldom insulated from politics (Dasandi and Esteve 2017), impacting daily activities, such as enforcement (Amengual 2016;Holland 2016) and worker appointments (Grindle 2012).While traditional views deemed political influence on the bureaucracy detrimental (Hicken 2011), recent scholarship recognizes its potential contribution to policy delivery (Toral 2023).Importantly, the presence of high-capacity agencies, regardless of their degree of autonomy, within overall low-performing countries, indicates that political influence is neither all-encompassing nor inherently counterproductive (Bersch et al. 2017).
Previous research revealed sources of bureaucratic autonomy in policymaking.Besides civil service's insulation from politics, bureaucrats may rely on expertise to assert control in decisionmaking.Appointed high-profile technocrats are commonly insulated in Latin American governments and may advance reforms that counter political interests (Dargent 2014).Bureaucrats in Brazil have sought alliances with civil society actors to advance their policy agendas (Rich 2013) or engaged in contentious causes and activism from within the state to shift the direction of policies (Abers 2019) or open up space for participation (Silveira 2022).The question is whether similar processes take place in policy implementation.
Despite this growing literature on bureaucratic politics in the Global South (Pepinsky et al. 2017;Brierley et al. 2023), we know less about the politicianbureaucrat interface at the front line.Street-level bureaucrats interact directly with citizens and exercise some discretion in policy implementation, acting as de facto policymakers as they interpret rules and allocate goods and services, typically in resource-constrained environments (Meyers and Vorsanger 2003;Lipsky 2010;Hupe et al. 2015).
The challenges of frontline work are exacerbated in the Global South by three main factors: severe resource constraints within the public sector, greater social inequality, and considerable political influence on bureaucracies (Lotta et al. 2022).As a result, different coping mechanisms may emerge among street-level bureaucrats, including improvisation to assist clients and compensation for institutional weaknesses or engagement in predatory behavior linked to clientelism (Peeters and Campos 2022).In this case, frontline workers may put personal interests first and practice selective enforcement when they have more discretion.Conversely, the literature suggests that when they have less autonomy, they focus on job survival, engaging in "alienative commitment" (Peeters and Campos 2022).
Prior research on the interface and tensions between politics and bureaucracy in the Global South thus emphasizes political oversight, with fewer instances of bureaucratic autonomy.These are typically associated with professional bureaucracies or insulated technocracies at higher levels of government.It is not clear, however, how frontline bureaucracies, which are more permeable to political influence, may safeguard some autonomy over allocational decisions.

Political Intermediation and Bureaucratic Work in Land Regularization
Traditional approaches to distributive politics treat brokerage and programmatic distribution as binary categories.The former refers to the distribution of public goods or benefits mediated by political intermediaries, not following programmatic criteria -established in public debates or internal governmental processes-as in the case with the latter (Stokes et al. 2013).However, by examining land regularization implementation as a process, from the starting claims to the issuance of land titles, a more nuanced picture emerges.This analysis unpacks the intricate interaction between intermediaries and bureaucrats, revealing that allocating titles involves negotiating competing interests and constraints.Simply put, neither bureaucrats nor politicians wield complete control over the distribution of land titles.
I argue that bureaucrats solve the tension between distributing land titles based on political intermediation or programmatic criteria by dissociating the assessment of eligibility rules and the allocation of land titles.Bureaucrats control the first part, preventing political influence.After guaranteeing eligibility criteria are verified, they may accommodate qualifying claims by managing the queue of recipients, which remains for lengthy periods to be processed.Indeed, as Krishna and colleagues (2020) underscore, intermediation may influence the time bureaucracies take to deliver land titles.However, in our case, it may allow for queue jumping but not the overriding of policy rules, as bureaucrats keep the elaborate documentation procedures away from political influence.
Expanding upon the previous discussion, I sustain that frontline bureaucrats may also rely on technical expertise to prevent political interference, even if they are not insulated from politicians.Bureaucrats interpret and amalgamate a series of eligibility rules, which are based in disperse laws from different levels and in norms of practice, and operationalize them in the technical procedure of documenting informal settlements.In doing so, they oversee the eligibility of informal settlements for land regularization.Even though politicians may appoint workers of their preference and shift statutory civil servants' positions, they refrain from doing so.Bureaucrats' knowledge is indispensable in the execution of sophisticated public policies, rendering them essential to politicians (Dargent 2014, 5).
In contrast to what Peeters and Campos (2022) find in the existing literature on street-level bureaucracy in the Global South, the land regularization bureaucrats studied do not engage in predatory behavior, red taping, or a lack of commitment in the face of brokerage.Instead, they leverage political intermediation in favor of citizens and even use legal deadlocks to avoid negative outcomes.Besides, they use some of their discretionary space to favor collectively organized residents.These behavioral patterns are related to bureaucrats' embeddedness in civil society, as discussed after the methodology section.

METHODOLOGY
This paper draws on qualitative research and document and interview data.I collected the 46 semistructured interviews used in this paper during 13 weeks of fieldwork between 2021 and 2023.Respondents included bureaucrats, politicians, law enforcement actors, and civil society actors.I recruited participants based on their professional or leadership roles, using snowballing to expand the sample.For the executive branch, I identified persons who were in or had previously occupied key positions in the municipal executive, notably Housing Secretary officials, bureaucrats (statutory or otherwise) who worked as coordinators of land regularization, directors of departments, and other members of the land regularization and housing staff.I interviewed housing and land rights specialists in the Public Prosecutor's and Public Defender's offices, as well as leaders of housing movements, neighborhood associations, and nongovernmental organization (NGO) members supporting them.Municipal legislators were suggested by other respondents for their demands of land regularization for their constituencies.
More than half of the respondents were appointed or permanent bureaucrats with previous or current experience with land regularization in the housing authority (Municipal Secretariat of Housing-Secretaria Municipal de Habitação).Most of them were female architects and engineers, lawyers, or social workers and many had a longterm experience with land regularization.Their profile, summarized in table 1, relates to the overall characteristics of the land regularization bureaucracy. 2In some cases, respondents could be affiliated to more than one category, so I considered their primary role (e.g., some civil society actors had a little experience as advisors in the municipal executive or legislature and some bureaucrats engaged with social movements).
Respondents were treated as key informants for their unique experiences implementing, demanding, monitoring, or benefitting from land regularization.Questions were open-ended and adapted to cover respondents' experience and interactions with other actors.Bureaucrats revealed the intricacies of the land regularization procedure, formal and informal rules of managing policy implementation, their participation in making new regulations, and the influence and interests of actors in this policy.These interactions were also explored with executive officials, councilors, civil society actors, and public prosecutors.Testimonies of civil society and law enforcement actors also contributed to a better understanding of the broader context of informal land tenure and land rights struggles.Interviews were conducted in Portuguese on-site or online, if requested by respondents.
Respondents provided oral consent at the beginning of interviews after I explained the research purposes, data processing, and protection measures.Such procedures were agreed with the ethics and data protection unit of the authors' affiliated institute.Qualitative content analysis was conducted with the transcripts of interviews.The excerpts in this paper are referenced by randomly assigned numbers within each group listed in table 1, with information that could identify respondents suppressed or modified.Quoted interviews and documents were translated by the author.
Besides interviews, I analyzed documents to trace relevant legislation applicable to land regularization, examine the background of this policy in São Paulo, and verify statements of fact provided by interviewees.Documents analyzed include the municipal and federal laws, decrees, ordinances, resolutions, communication brochures, administration target plans, and official reports available online or in the Housing Secretariat library archive and websites of São Paulo municipality.Academic literature on the subject also provided contextual material.

LAND REGULARIZATION IN SÃO PAULO
Before examining bureaucratic work, this section provides a background to land regularization policy and describes its bureaucracy in the municipality of São Paulo.Latin America's history of informal urban settlements is relatively long, associated with socioeconomic macroprocesses and national and local policies and politics (Gilbert 1994;van Gelder et al. 2016).In São Paulo, informal urbanization took form predominantly via land occupations and unauthorized subdivisions of land into plots for sale, both typically lacking services and shaped by the autoconstruction of dwellings (Fernandes 2011;Caldeira 2016).These are respectively referred to as favelas and irregular subdivisions (loteamentos irregulares), where approximately 3.2 million people live in 16,421 hectares or 18% of São Paulo's urban area (CEM and SEHAB 2016).In 2020, 27.5% of the population and 21.4% of households in the municipality of São Paulo were estimated to live in these settlements (Marques and Minarelli 2021). 3These are the types of informal settlements typically concerned in land regularization.

Land Regularization Policy
The first objects of regularization in São Paulo were irregular subdivisions.These were developed mostly between the 1970s and 1990s (Pasternak 2010).In 1981, the municipal government of São Paulo created a unit called Serla 4 to regularize irregular subdivisions on private land and carry on diligences to prosecute illegal developers.
The first national law on urban land subdivision was issued in 1979, 5 establishing legal concepts, rules, and parameters for land development (e.g., basic infrastructure required, restrictions on the subdivision of land in "improper conditions") and made it a crime to subdivide plots not following these rules (Tierno 2020).The 1979 law created conditions for municipalities to intervene with land regularization when developers did not comply.Implementation was concentrated in the early 1980s, and titles were issued without clear procedures, generating some discord that remained unsolved for decades (Interviewee 9; Interviewee 25).
After the end of the military dictatorship, with the election of the first left-wing mayor in 1988, regularization gained relevance as a municipal policy with a department created in 1990 in the housing authority.The department, Resolo, 6 systematically analyzed the irregular subdivisions and opened administrative proceedings to formalize land tenure via the municipality.This first mandate of the Workers' Party (Partido dos Trabalhadores [PT]) in the city of São Paulo  was also marked by a slum upgrading program in favelas (Marques and Pulhez 2021).These are denser and more often result from land occupations without the gridded subdivisions of lots and streets present in market-oriented irregular subdivisions.Favelas are more frequently located on public lands, such as areas destined for leisure or institutional use within a larger subdivision (Interviewee 5; Interviewee 7).In this period, while land regularization was conducted in irregular subdivisions in private land, the policy in favelas was aimed at the provision of basic services, by a distinct unit in Habi 7 (Interviewee 4; Interviewee 20; Cities Alliance 2004).
Land regularization of informal settlements in public land only became significant a decade later, during the second PT mandate in the municipality .After the federal government enacted in 2001 a new instrument of concession of use rights for housing purposes, 8 the municipal housing authority created a large-scale program to issue concession titles to informal settlements in public land, targeting areas subject to repossession lawsuits (Interviewee 2; Interviewee 10).Since the early 2000s, land regularization on public land has been carried out throughout the terms of both sides of the political spectrum, with more emphasis in PT mandates (Marques and Pulhez 2021).New laws have been issued at municipal and federal levels in the past 15 years and rules were updated, consolidating a procedure for this policy in the city.
This gradual institutionalization of land regularization is inscribed in a broader process of incremental policy change in São Paulo.Since the return to the democratic regime in Brazil in the late 1980s, the municipal government of São Paulo has amplified and diversified the policies addressing housing precarity (Marques and Pulhez 2021), achieving more equality in the distribution of housing (Bradlow 2022).Influenced by academics, bureaucrats, and professionals linked to the urban reform movement (Fernandes 2007;Bonduki 2017), this sector was further developed at the local level, encompassing new policy instruments, such as housing plans and zoning tools (Sette Whitaker Ferreira et al. 2020;Nastari Fernandes 2022) and an information and classification system (Gonçalves and Gama 2020; Coelho 2013), and benefited from international and national funding schemes (Denaldi and Cardoso 2021;Marques and Pulhez 2021).New legislation at the national level (Macedo 2008;Tierno 2020) also contributed to this institutionalization process, which was not linear, as shown by Marques (2023a;2023b) in this and other urban policy sectors.

Land Regularization Bureaucracy
Even though land regularization bureaucrats are in contact with policy users, they differ from the typical street-level bureaucrats portrayed in the literature as they are not as distant from supervisors.The Coordination of Land Regularization (Coordenadoria de Regularização Fundiária [CRF]) was created in 2013, combining the organizational structures and staff of Resolo and Habi previously dedicated to regularization.CRF's coordinator reports directly to the housing secretary and oversees four subunits. 9The municipal administrative apparatus is compressed in comparison to central governments typically portrayed in the literature, and therefore, the distance between frontline workers and secretaries is not as great.
CRF is primarily dedicated to policy implementation as its bureaucrats are responsible for processing all administrative cases of land regularization.Each case refers to one informal settlement, an area comprising multiple plots titled individually by the end of the procedure.Besides handling casework for land title distribution, bureaucrats perform activities not typical of street level: they discuss complex cases and their interpretations of legislation, promote new initiatives, and propose regulations, as discussed ahead.CRF thus combines characteristics of street-and mid-level bureaucracies, with bureaucrats who act as implementers and, eventually, rule-makers.
The bureaucrats working with land regularization typically have technical expertise and long-term experience with this policy.The unit convenes career civil servants selected in open and competitive exams, as well as temporary employees appointed by politicians as nonstatutory civil servants or hired through private contractors. 10Only civil servants (statutory or otherwise) are eligible to be nominated as heads of the coordination, departments, or divisions and to approve documents as technical managers.Career civil servants typically serve in the unit for several years, while temporary employees often have intermittent involvement, which relates to changes in the political party in power (Interviewee 10).For instance, some bureaucrats participated in the early land regularization initiatives in São Paulo and other metropolitan municipalities during the PT administrations (Interviewee 12; Interviewee 24); others did the same later in center-right mandates.Despite these links, most bureaucrats are not affiliated to parties.Politicians may even nominate experienced bureaucrats based on their track record of policy delivery, regardless of their ideological affinities with another party (Interviewee 30).This differs from the usual approximation of political appointments to patronage (e.g., Toral 2023) and suggests that they may also be guided by workers' expertise.
Another noteworthy aspect of the land regularization bureaucracy is its connections with civil society.Bureaucrats participate in housing policy debates and forums, some worked in NGOs dedicated to housing and land rights and, occasionally, they collaborate with social movements (Interviewee 4; Interviewee 15; Interviewee 23; Interviewee 25).Most of them share affinities with the urban reform agenda, which mobilizes activists, academics, bureaucrats, and NGO members (Marques 2023b).Similarly, housing activists also held positions in the municipality of São Paulo during PT mandates.For instance, between 2001 and 2004, these activists played a pivotal role in establishing the municipal housing council (Interviewee 45), a core agenda of housing movements (Tatagiba and Teixeira 2016).The same opening to social movements happened later in the PT presidency (Abers et al. 2014).This reveals the permeability of the São Paulo municipality to social networks and its embeddedness in civil society (Coslovsky 2015;Bradlow 2022;Marques 2023a).

Documenting for Eligibility
The production of documents and representations is a process that renders a complex social phenomenon legible to the modern state (Scott 1998).It entails a simplification in the selection of certain aspects of interest.Bureaucrats document informal settlements to assess various eligibility criteria and operationalize tenure security recognition to demarcated areas.Bureaucrats assess about twenty different documents for each administrative proceeding, including field visits reports, topographic surveys, cadasters, environmental and risk assessments, records of overlaying plans and property deeds, and a consolidated land regularization plan.A large part of the production of documents is outsourced to private management contractors hired via public bidding (Pulhez 2016).
The three most common standards applied to informal settlements mentioned by respondents refer to the main bottlenecks dividing cases that get to move forward.First, informal settlements (or parts of them) in areas with identified risks of landslides or floods are not eligible for regularization.Second, a cutoff date establishes the time frame for eligibility.That is, settlements are only qualified for regularization if developed before a specific moment-a similar regulation is observed in other cities (Ren 2018).Third, informal settlements are assessed regarding the availability of infrastructure and services (street lighting, pavement, sewage, water, and electricity).They are only eligible for land titling once this "essential infrastructure" is in place.Besides, land regularization in certain environmental conservation zones is possible, requiring a specific procedure of documentation and evaluation.These rules have different origins and implications regarding the path ahead for each case.
The interdiction to carry out land regularization in disaster risk areas stems from prohibiting any land subdivision in these areas, as established in the 1979 law.With the continued expansion of informal settlements and intensification of environmental disaster risks, the salience of this criterion for land regularization increased.CRF currently rejects requests for land regularization in these areas unless risks are classified as low and mitigated before the attribution of land titles.Other agencies in the municipality may eventually displace and resettle residents in these areas.
The cutoff date is another rule established in law (municipal and federal), but this does not imply that this is a fully clear or stable rule.The development of legislation at federal level independently from the municipal level may create ambiguities that require the interpretation of bureaucrats (Interviewee 12).This situation happened with the change of the land regularization federal law in 2017, pushing the cutoff date ahead and defining new procedures.It was only followed by a matching update in municipal law in 2022.Between 2017 and 2022, bureaucrats defined shared practices to comply with both regulations (Interviewee 12).These changes also reveal that legislators pushed toward less strict rules at the central level, in contrast to other Latin American cases (Holland 2020).If an informal settlement does not comply with the cutoff date, the case is rejected.Still, the residents can wait "until the day someone changes the law and changes the date" (Interviewee 17) and then file another request for regularization.The excerpt below illustrates how these diverse rules require the interpretation of bureaucrats: Besides criteria established in federal or municipal legislation, documentation also embeds bureaucratic practices.This is the case for the essential infrastructure criterion: although the legislation did not explicitly limit regularization to settlements with infrastructure installed, in São Paulo, this has been a relatively stable practice, justified by some respondents based on past experiences that have produced a mismatch between land titles and the "ground truth."Because servicing informal settlements often requires changes in plot layout and, in some cases, partial displacement or reallocation of the residents, the regularization proceedings are left on standby until the construction works are finished.The statement below illustrates this practice and its rationale.
Interviewee: Although the federal law says that it is possible to regularize in any situation, without infrastructure, etc., from the City Hall's experience, we try not to do this.[ : : : ] According to the law, we can distribute titles, but many times, these titles will not match [the plot layout], because you will need to do some construction works, and this will change the characteristics of the land subdivision.[ : : : ] So this is why we are reluctant to give titles in default, without the missing works.Because if [the infrastructure works] will de-characterize the subdivision, we will have to rework. 12Interviewer: Is this a rule of yours here?Or is there an explicit directive?
Interviewee: An explicit directive would be the law.It is more like : : : due to the knowledge of what has happened in the past here in São Paulo that we end up not doing this, guaranteeing that we will not repeat the past mistakes.(Interviewee 25) Prior documents and testimonies reveal this has been a well-institutionalized practice.The Municipal Housing Plan 2009-24 differentiates informal settlements requiring urbanization from "those that were urbanized and, therefore, should be encompassed by land regularization and registration programs."(Prefeitura de São Paulo 2010, 22).Furthermore, it defines "an urbanized settlement [as having] 100% of all infrastructure networks in place, namely water, and sewage networks, paving, drainage, street lighting, and containment of risk areas."(22).As these documents demonstrate, the housing authority prioritizes urbanization (understood as infrastructure) first and legal land title as a subsequent step.Such a requirement differs from land titling procedures in other countries, where formal land tenure is a prerequisite for infrastructure provision.This long-standing practice in São Paulo stems from the land subdivision legislation, which crystallized the idea that large portions of land are urbanized upon their subdivision in plots and in public space (with a percentage of land being earmarked for that), and equipped with infrastructure.Relatedly, the notion of land regularization as entailing both urbanization and legal land tenure was consolidated-even if, in practice, it means assessing infrastructure as a condition for titling.
The requirement of infrastructure has consequences as to which settlements achieve regularization.Even though the policy focuses on low-income neighborhoods, more consolidated settlements, with the infrastructure required, are typically at the higher end of this group.That means that the eligible cases are those closer to full security in the "tenure continuum" (Payne 2001;Issar 2022).Even though these cases have lower risks of being evicted, full tenure allows for residents to have a formal address and deeds that can be passed on to future generations. 13 Such bureaucratic practices may eventually turn into formal rules.The crafting of legislation sometimes takes on existing arrangements and actions, turning them into formal regulations.Bureaucrats themselves seek to have practices recognized by formal rules to legitimize their decisions by legal standards.This was the case of the infrastructure eligibility criterion, achieved in 2022.Bureaucrats report that internal debates and workshops subsidized the drafting of the bill.Civil servants contributed, for example, by including the criterion of infrastructure requirement (Interviewee 25; Interviewee 11).The bill was discussed and developed with the municipal housing council, which then sent it to the municipal legislature (Interviewee 14).As a result, since January 2022, the updated municipal law conditions titling on the implementation of infrastructure. 14

Documenting for Ordering
Given the large sum of administrative proceedings to carry out and the long time they take to be concluded, a queue is inevitable.One interviewee reported that there are over five thousand open administrative proceedings (Interviewee 17).The bureaucracy processes only a portion of these administrative cases at a time.In 2023, CRF was working on about 500 cases (Interviewee 17).Consequently, determining which cases move to the front of the queue becomes another critical issue.
Besides eligibility criteria, the documentation of informal settlements also implies a form of ordering administrative proceedings.Lists with priority settlements are prepared at the beginning of an administration or when preparing the bidding for private contractors (Interviewee 8; Interviewee 10).Bureaucrats provide an initial input by analyzing the stage of development of the administrative proceedings.In other words, they assess which cases have most documents necessary for concluding the land regularization procedure and whether these documents are up to date.
The administrative proceedings are divided between the technical staff [i.e., bureaucrats].
[ : : : ] So, when the beginning of the administration comes, we define the targets for the four years [ : : : ].Each technician has their own areas to take care of, so s/he knows more or less the progress of each case.So, when the cabinet asks: "We need to set up the targets, what do you suggest?"Then, the technical team makes their suggestions, as the people working with the administrative proceedings on a daily basis.(Interviewee 25) Giving priority to cases with more advanced documentation contributes to achieving the administration targets.Public servants described dividing their workload between many proceedings with more advanced documentation and fewer, more complex cases, not leaving behind the last but still focusing on meeting the targets (Interviewee 10; Interviewee 25; Interviewee 12).Besides assessing the stage of documentation, bureaucrats include in the lists cases in which the municipality is ordered to carry out regularization because of a conviction by the Public Prosecutor's Office (Interviewee 25). 15Besides meeting administration targets and legal obligations, bureaucrats also manage the requests linked with claims-making, which will be discussed below.

POLITICAL INTERMEDIATION: PUSHING DEMANDS WITHIN A PROGRAMMATIC FRAMEWORK
Municipal legislators of different parties ask the housing authority to deliver land titles to informal settlements in their constituencies' territories and follow up on these demands themselves or with the support of brokers (Interviewee 25; Interviewee 12; Interviewee 8; Interviewee 10).As advisors report, many municipal legislators operate in an ad hoc electoral-district logic (Interviewee 36; Interviewee 38), "with electoral constituencies clearly defined" (Interviewee 37) in the territory.Councilors send requests to the housing authority calling for priority for their constituencies' areas. 16 The bureaucratic rules discussed in the previous section set a logic of allocating land titles according to rational principles anchored in legal, administrative, and technical bases, contrasting with political intermediation.Bureaucrats follow the first rationale primarily but must still manage the second one.In other words, while the bureaucracy's rationalization procedures account for classifying and ordering informal settlements eligible for land regularization, room must be made to accommodate such requests.Nevertheless, testimonies demonstrate that bureaucrats find ways of conciliating the two logics to a certain extent.
Interviewee: What we do here most is answer councilors and prosecutors [ : : : ].
Interviewer: What are they demanding?Interviewee: Information, they want to know why : : : they ask to prioritize, do you understand?They [councilors] bring the community here to talk, then we show them the proceedings and explain the situation : : : Interviewer: And are you able to absorb this demand?
Interviewee: What we do is put it on our list.When this is the case, it is included on the list.When the bidding process [for private contractors] is prepared, we select the areas that will be included in the contract.Then, eventually, when we identify that it [the councilor's demand] is an area that can be regularized and that we have inspected to see if it has all the infrastructure, if the situation is OK : : : [it goes on the list].(Interviewee 8) The excerpt above suggests how the bureaucracy channels nonprogrammatic demands into programmatic goals: they evaluate the case to check if it fits the eligibility criteria.If affirmative, they include it in their pipeline.Therefore, the bureaucracy processes the nonprogrammatic requests and constrains them to a rational scheme.Still, the two distribution logics may occasionally conflict because the

Claims from Civil Society
Indeed, political intermediation is not the only way for informal settlements' residents to claim land rights.Neighborhood associations can directly pressure bureaucrats as they "work on these [more than five thousand] administrative proceedings based on the demand" (Interviewee 17).This interaction between bureaucrats and citizens improved as the 2013-16 administration established a citizen service desk for anyone interested in filling a land regularization case or following up an existing one.Bureaucrats may use their discretion to prioritize mobilized groups, considered by some as more deserving: The requests from neighborhood associations also enter here [in the list bureaucrats prepare for the administration targets].There are loads of claims for these cases, and every month, people are here.I always tell them: "Look, I have more than two thousand proceedings : : : which one will I give preference to?"We can't work on all of them at the same time.When the civil society is organized and comes claim to the city hall, monitors, and is always close, these proceedings end up being priorities.Why will I deal with a case that nobody is looking for if I can deal with a case that everybody is interested in?So : : : [as if speaking to them] "get organized and do come and claim!"I, at least, end up organizing the demand considering the claims of associations and communities' movements to choose the cases.But there is also a lot of political demand.(Interviewee 25) Assessing and incentivizing the collective organization of residents of informal settlements has been a relatively long-standing practice of the housing authority in São Paulo.Surveys from 1992 identify residents' forms of collective organization, and guidance leaflets also encouraged it a decade later: "With the organized action of the resident families and the social movements fighting for housing and for urban reform, throughout the years, several improvements and public services have been implemented in the informally occupied areas.[ : : : ] Participate in your community organization.You and your area will only win!" (Prefeitura de São Paulo 2008).Even though most bureaucrats interviewed share similar perspectives, some of them raise concerns about most vulnerable communities having a lower capacity of collective organization (Interviewee 26).
Bureaucrats' testimonies and nonverbal communication during interviews suggested that most of them disapprove of the influence councilors and politicians have on policy implementation.Many respondents avoided talking openly about politics, and portrayed a clear-cut division between "technical" (or rationalbureaucratic) and "political" decisionmaking, implying a normative difference between both.Even so, some of them recognize that councilors are in a legitimate position to demand services for their constituencies if they are responding to the needs of citizens: "For electoral purposes or not, they come to us to intermediate with the residents.[ : : : ] And in a certain way, I don't think it is so bad because these are doing their job, right?" (Interviewee 25).Requests by neighborhood associations and other collective organizations of residents, in turn, are taken as entirely legitimate.

BUREAUCRATIC STEERING: DISSOCIATING ELIGIBILITY AND ALLOCATION TIMING
The evidence presented in previous sections portrays how both nonprogrammatic politics and policy rules are relevant for the allocation of land rights in São Paulo.The housing bureaucracy has an extensive set of classification mechanisms-based on legal rules and long-standing practices-for rendering informal settlements legible and identifying eligible cases.At the same time, bureaucrats must address politicians and are pressed to fulfill (at least in part) the demands of councilors.The accommodation of these requests does not mean that they override the rational-bureaucratic policy procedures.On the contrary, the bureaucracy steers nonprogrammatic demands to policy goals, accommodating only requests that fit eligibility criteria.
How, then, do bureaucrats enact policy rules and accommodate nonprogrammatic demands?This conciliation is possible by dissociating the eligibility assessment from the allocation of land titles.In other words, bureaucrats evaluate all cases considering technical criteria and identify those qualifying for land regularization, thus preventing political influence at this point.Once qualifying cases are screened, the selection of recipients results from a compromise between achieving administration targets and meeting the claims of political brokers and collectively organized residents of informal settlements.The remainder of this section discusses why bureaucrats can keep political influence at bay while screening informal settlements.
Eligibility rules are well defined and enacted in documentation practices.As far as this research could assess by reviewing administrative proceedings and interviews with other actors, bureaucrats follow closely these implementation rules.More than that, they were relevant actors in the production of these rules, as the recent change in municipal legislation illustrates.However, the ordering and queueing of land regularization cases are more malleable and not specified by legal norms.This is a key point to understand why bureaucrats resist political pressure.Municipal bureaucrats (statutory or otherwise) are legally accountable for the technical documents subsidizing administrative cases and, ultimately, land titles awarded.Political and legal accountability may conflict with one another (Hupe and Hill 2007), and bureaucrats stand by the latter.When asked about whether political pressure may influence the technical assessment of cases, an interviewee said: There is, and there will always be [pressure], especially because of the councilors.But [ : : : ] the people here and I always say: "If you're not comfortable signing, don't sign!" Regardless of pressure from councilors, I think it's our name out there, right?So, no.(Interviewee 24) Being "comfortable to sign" means that one attests the technical and legal responsibility for a document determining whether the case is eligible for land regularization.In doing so, bureaucrats depoliticize the distributive issue, shifting it to the technical domain, whereby they have the upper hand.Even if politicians can appoint bureaucrats more aligned with them, the latter tend to follow the "technical" view.One of the few appointees belonging to a ruling party coalition explained that when bureaucrats refused to take on cases of ineligible settlements, "the pressure was all on me.It was that 'if you don't do it, we'll ask for your job' kind of thing.[ : : : ] What could I do?It's a technical question" (Interviewee 17).As another bureaucrat puts it, "the technicians [or bureaucrats] want to see that problem solved, but in a way that their functional safety is guaranteed, which is the right thing to do" (Interviewee 4).This suggests that noninsulated implementation bureaucracies may also mobilize technical expertise to deter political influence, even if they have less autonomy than high-level technocracies.
Given the long list of eligible settlements and the lengthy process, the management of time is a central aspect of land regularization.Manipulating the beneficiaries' queue impacts the timing of land rights distribution, with cases benefiting from political intermediation more likely arriving first, as previous research showed (Krishna et al. 2020).In the case of São Paulo, the embeddedness of bureaucrats in civil society makes them more sympathetic to collectively mobilized groups, who may benefit from bureaucrats' discretion to get a place higher up in the line.
Nonetheless, these are not the only factors affecting the timing of allocation.Land regularization is a long process, taking from a few years to many decades to be concluded.It is not uncommon for administrative proceedings to remain on standby due to a lack of complete infrastructure.Bureaucrats may also manipulate time to avoid adverse outcomes for communities in informal settlements.For example, they may leave administrative cases pending to prevent the harmful effects of legal deadlocks, as described below.The case in the excerpt may now be solved with new instruments introduced by more recent legislation (Interviewee 4).
We still have proceedings that we sometimes maintain [open] here because there is an administrative proceeding rule that if the request comes from the interested party, you either approve or reject it.[ : : : ] I have cases, for example, that have generated property deeds of the lots at the land registrar's office, but the plan that I have in the proceeding is different from what was opened in the registry.So, if I dismiss the proceeding, it is as if I were asking for those titles to be canceled.And I can't accept it because what is here differs from what is implanted.[ : : : ] So sometimes, we end up holding back many things to avoid harming the residents that are there.(Interviewee 4)

CONCLUSION
This article analyzed the bureaucratic work of land regularization in São Paulo, Brazil, to understand how bureaucrats allocate land rights to informal settlements when faced with political intermediation.It demonstrated that bureaucrats partially control the distribution of land rights by enacting eligibility criteria.In the process of documenting informal settlements, bureaucrats integrate these requirements that are scattered across different legal pieces and norms of practice.Bureaucrats avoid political influence at this point by asserting their technical and legal responsibility.Given the backlog of land regularization cases, the bureaucracy only works on a portion of the administrative proceedings at once.Political intermediation and claims-making by neighborhood associations are influential in the selection of cases to be at the front of the queue, among other constraints.The article concludes that the conflicts arising from distribution based on nonprogrammatic or programmatic criteria are conciliated by dissociating the selection of eligible cases from beneficiary ones.In this case, bureaucrats control the first and negotiate the latter.
These findings converge with previous research arguing that political intermediation typically reduces the timing of land titling (Krishna et al. 2020) and expands these findings by showing that bureaucrats may still control a substantive part of this policy as they master its technicalities, and may fully define eligible cases.Also, this study finds that a noninsulated implementation bureaucracy may rely on technical expertise and legal liability to keep politicians' influence partially at bay.At the same time, bureaucrats embedded in civil society recognize citizens' claims as legitimate and are likely to be responsive to them, giving priority to these cases.Interestingly, bureaucrats may leverage political intermediation by negotiating resources from legislators that can be used to fulfill eligibility criteria in certain informal settlements.These interactions of frontline workers and politicians differ from patterns found in other Global South street-level bureaucracies (Peeters and Campos 2022).
The empirical account of the entanglement of brokerage and policy processes in this paper addresses calls for further research on the politics of urban land regulations and informal settlements (Auerbach et al. 2018;Post 2018) and highlights the central role that may be played by bureaucrats in distributive issues in cities.The findings suggest that there is a need for greater examination of the interplay between politicians and bureaucrats at the front line and for attention to variation across agencies within cities and countries.Further investigation of the interaction between nonprogrammatic politics and policymaking may explore the unfolding of this relation in time-for example, investigating what happens to brokerage after processes of institutionalization or shifts in bureaucratic autonomy.Alternatively, future research may explore whether political intermediation incentivizes the continued implementation of certain policies in contexts where there are significant disruptions due to political shifts.All in all, research on this topic will benefit from further contextualization and a careful understanding of whether and when nonprogrammatic distribution undermines policy delivery.
It's always like this.The municipal law says the date limit is April 2000.Then the Master Plan [municipal law] says it is July 2014.But then comes the [federal law] 13.465 and says that for granting tenure legitimation-which is basically the registry in the name of the occupant-it is December 2016.[ : : : ] 11 They [the legislators] are always pushing ahead [the cutoff date].[ : : : ] And there are always new people trying to convince you that they occupied the area before 2016, even if the aerial photographs show there wasn't anything.(Interviewee 12)

Table 1 .
Interview Respondents **One respondent participated in two interviews.Therefore, 46 interviews were collected with 45 respondents.***Includes degrees in administration, economics, social sciences, and one respondent without a degree.8 LATIN AMERICAN POLITICS AND SOCIETY https://doi.org/10.1017/lap.2023.35Published online by Cambridge University Press LATIN AMERICAN POLITICS AND SOCIETY https://doi.org/10.1017/lap.2023.35Published online by Cambridge University Press