The Governance of Public Budgeting: a Proposal for Comparative Analyses - the Cases of São Paulo and London ^*

Abstract

This paper aims at understanding the governance of public budgeting in large metropolises with the use of comparative analysis. The analysis is focused on budgetary governance in London and São Paulo and uses qualitative and quantitative data from 2008 to 2019 to understand whether analytical categories such as incrementalism of expenditures, complexity of budgetary rules, bureaucratic hierarchy, bargaining, and muddling through are useful to compare two metropolises, especially to determine the discretionary power of mayors in making budget allocation decisions. The analytical categories are derived from the studies of theorists of economics and political sociology, notably Wildavsky (1975, 1969), Wildavsky and Caiden (2004), Schick (2009, 1976), Caiden (2010) Lascoumes and Le Galès (2005), Baumgartner and Jones (2005), and Fuchs (2012, 2010). The main argument of the paper is that, despite the administrative and political differences between London and São Paulo, similar dimensions can explain decisions about budget allocation and the political discretionary power of mayors. The study shows that mayors have little discretionary power, particularly in contexts of fiscal austerity; it also highlights the importance of property tax as a means to protect such power.

Public budgeting; governance; London; São Paulo; fiscal austerity

During a public hearing in the city of São Paulo in 2013, a social-movement representative asked an official of the Planning Secretariat how was it possible that, with a US$11 billion budget, the mayor could do nothing to improve public safety. During the tense pre-election period in November 2019 in London, several protesters complained about the reduction of social and security policies and condemned the mayor's inaction. Despite the differences between the two cities – London has a budget about 10 times larger than that of São Paulo and a completely different administrative policy structure – are there any similarities in the governance of public budgets in São Paulo and London? What are the determinant factors explaining the autonomy and discretion of mayors over budget allocation and public policies implementation?

This paper aims at understanding this issue by investigating, in a comparative analysis, the governance of public budgeting in large metropolises. The analysis is focused on budgetary governance in two global metropolises, London and São Paulo – the former being in a developed and central country in the global North; the latter, in a federal and underdeveloped country in the global South. Despite their structural differences, both cities are economic capitals subject to annual budgetary processes. The analysis thus focuses on revealing these differences and similarities in order to understand whether analytical categories such as the incrementalism of expenditures or the complexity of budgetary rules are useful in the comparative analysis of two major metropolises, particularly in determining the discretionary power of mayors over budget decisions.

The comparative analysis is guided by analytical categories – structured in five dimensions – which were derived from studies of theorists of economics and of the political sociology of budgets, notably Wildavsky (1975 , 1969 ), Wildavsky and Caiden (2004) , Schick (2009 , 1976 ), Caiden (2010) Lascoumes and Le Galès (2005) , Baumgartner and Jones (2005) , and Fuchs (2012 , 2010 ).

The study focuses on the governance of public budgeting in São Paulo and London and on their contentious processes – not just on the fiscal results of the dispute involved. The main argument is that, despite the differences between the administrative and political structures of distinct countries, similar dimensions can explain budget allocation decisions and the level of political and discretionary power of mayors. The analysis shows that, although mayors have little of this power, especially under fiscal austerity, it is a very symbolic form of power in the political differentiation of governments. Discretionary power is limited in both metropolises, since most of the expenses are defined either by constitutional rules in São Paulo or by central grants in London. In a context of economic expansion, important increments may come from tax revenues and be used to advance the political programs of mayors. However, under fiscal austerity policies, most increments may only be enough to maintain ordinary services, if so. Thus, discretionary decisions about increments are very important and symbolic. Similarly, this study also highlights the importance of property tax in both metropolises as a means for guaranteeing mayors’ discretionary power.

The remainder of this text is organized in four sections, followed by the conclusion and final remarks. In the first section some brief remarks are made on our methodological choices. The second section presents the theoretical framework on which this study draws to determine the analytical dimensions used to compare the budgetary governance in the two metropolises. The third tries to shed light on the main administrative and political characteristics of these two cities. The fourth presents a comparative analysis for each dimension highlighted in the theoretical framework. Finally, the conclusion discusses two relevant points about the political discretion of mayors in the governance of both metropolises. The first point refers to the fact that this discretion is very much associated with the mayor’s degree of autonomy in determining tax policy, as observed in Slack (2017) , but also with his level of control over incrementalism of expenditures ( Wildavsky, 1969 ). The second point is that the less political discretion there is (as of percentage of the total budget), the more relevant this this discretion is and the greater the need to resolve conflicts in global cities like São Paulo and London.

Methods

This project is based on a method of international comparison of cases, which, by definition, allows for the analytical comparison of cases by similarities and differences ( Lijphart, 1971 ). Rather than making an analytical generalization, this study seeks to increase the external validity of the theoretical framework; in adding a new case, it expects to draw a clearer picture of budgetary governance. The analytical dimensions described in the next section have already been used to investigate São Paulo and Paris ( PERES and ALLÉ, 2021 ) and ( PERES, 2021 ).

This study examines two metropolises: one representing the global North (London); the other, the global South (São Paulo). They represent two very different cases. The first, the municipality of São Paulo, is in a federal country in South America. The overall population has been growing and the economic activities have been developing in this economic capital, which has led to increased demands for infrastructure and public services (transportation, water supply, waste management, healthcare, among others). The second metropolis, London, is in a very centralized country, the United Kingdom. Its budgetary governance is organized in two administrative levels: the GLA – Greater London Authority - and other 33 local authorities (32 councils and the City).

Despite these differences, these metropolises have important similarities: both are considered global, dynamic, unequal, segregated, and, most of the time, ungovernable ( GORDON AND TRAVERS, 2010 , HIGGINS et al., 2014 ; MARQUES, 2014 ; and SANTOS, 1994 ). From the standpoint of an international comparison of cases, these metropolises could thus be analyzed by combining the comparative methods of similarity and difference, as pointed out by Fernandes (2002) .

In order to improve the comparative analysis, the qualitative analysis was conducted in tandem with a quantitative data survey, in a ten-year series ( FERNANDES, 2002 ). The ten-year cut-off for data collection – which enhances the sample from the longitudinal extension – follows the suggestion of Lijphart (1971) . Another relevant choice is the analytical cut-off, i.e., budgetary governance. Thus, for these metropolises to be compared, however different (or similar) they may be, they must have an annual budget process, as discussed by Caiden (1994) .

In the case of São Paulo, our data include official balance sheets from 2008 to 2018 and documentary analysis of legislative budgetary processes. As for London, the analysis is based on budgetary data from 2011 to 2019, legal documents, specific literature on budgeting in England and public finance, and on interviews with experts, academics, practitioners, and civil servants. Between September 2019 and January 2020, 19 interviews were conducted with academics in the fields of public budgeting, tax policy, local government, public policy in comparative analysis, decentralization in the UK; interviews were also conducted with civil servants from central and local governments and with budget analysts linked to independent institutes of budget evaluation and thinktanks. It is worth noting that the author has been employed by the municipality of São Paulo in different periods and that the analysis is also based on direct observation.

Analytic dimensions of the governance of public budgeting

Each year, during the same period, following a predefined calendar, civil servants working in all national and subnational administrations, such as São Paulo, London, or Paris, set in motion a processual/legal routine to prepare budgets.

Although budgeting processes undergo changes, these tend to be marginal – expenditures in budgets thus usually grow incrementally ( Wildavsky, 1975 ). More significant moments of rupture can be seen following changes in priority, and budgeting processes will adapt to the new political and economic forces until a new incremental process develops ( Baumgartner and Jones, 2005 ).

Various comings and goings between the budget office, the frontline areas, and the executive and legislative branches are necessary to reach a decision and approve the administration’s annual budget. Moreover, there are social and economic actors outside the government structure who are monitoring, pressuring, and disputing this process. As such, budgets should be considered a sovereign space of distributive conflict.

Decisions over spending allocations occur throughout the process of establishing revenues and expenses. This process defines the priority levels, who is to finance public goods and services, and who is to receive funding. The budget as a public policy instrument ( Lascoumes and Le Galès, 2005 ) is extremely important for economic and social policies. Allocations in budgets reflect decision-making processes and they also establish the starting point for the process of determining the objectives in each public policy area ( Rezende, 2015 ).

Because of these characteristics, defining and describing the budget process and the outlines and details of its governance is not an easy task. Budgeting is not a completely technical process, despite all the systematic rules and legal structuring involved. The budget process is eminently political.

One of the main analytical concerns of budgetary theory is the margin for discretion that actors involved in the budget process use to balance funds (current and capital) and allocate them over the years. In this case, the analysis stems from the premise that incrementalism ( LINDBLOM, 2009 ) and rigidity are institutional elements of the budgetary process. Wildavsky (1969) shows that the budget incrementalism perceived throughout a period under study is part of the decision-making processes, and as so, it allows for public policies to be adjusted progressively. As shown by Baumgartner and Jones (2005) , however, it is important to consider that there are moments of budgetary rupture with significant paradigm shifts.

The theoretical debate based on Wildavsky, (1975 , 1969 ), Wildavsky and Caiden, 2004 , Lindblom (2009) , Shick (1976), Baumgartner and Jones (2005) , Caiden (2010) , Fuchs (2012) , and Rezende (2015) identifies several crucial elements to comprehending public budgeting. These elements are found in the macro- and microlevel of the institutional environment of public budgeting and they make up the formal and informal rules that affect and are affected by the behavior of governmental and non-governmental actors ( MARQUES, 2013 ) who have influence over public budgets. The interweaving of these rules and actors generates the governance of public budgeting.

As developed in Peres (2018) , budgetary governance combines political, economic, and institutional elements originating from distinct periods and paradigms, based on the search for rationalization, transparency, and accountability of the budget process, as well as on the recognition of civil and social rights (constitutional and legal rules, based on social welfare systems) that are characteristic of the macro-institutional environment. All these elements, along with the level of resource scarcity, play a role in determining the possibilities for revenue and spending ( SCHICK, 1976 ). With respect to the microlevel of the institutional environment, the literature suggests a strong influence of incremental and inertial elements on both the formulation and implementation of budgetary policies.

The intention here is to analyze the governance of public budgeting in both metropolises by comparing four different dimensions of this governance, namely: complexity of budgetary rules, bureaucratic hierarchy of budgeting, incrementalism of expenditures and bargaining, and muddling through. From the analysis of these four dimensions, it is possible to identify an important fifth dimension: the discretionary power of mayors. These five different dimensions are presented and explained as follows:

Complexity of budgetary rules

According to Bezes and Siné (2011) and Core (2006) , the budget process is complex, permeated by regulations, mechanisms, and instruments ( Lascoumes and Le Galès, 2005 ) that affect revenues and expenditures. Wider society has little oversight and/or understanding of the abstract flows deriving from this complex process. On the one hand, these rules may lead to rigidity in budget formulation, leaving little room for discretionary decision and negotiation. On the other hand, as complexity renders the process less comprehensible to society, it can be used by mayors to retain control and increase his or her discretionary power.

Bureaucratic hierarchy of budgeting

As a result of this complexity, only a selected group of actors comprehends the budget, usually those who are close to the political core of decision-makers – that is the departments of finance, government, and planning ( Schneider, 2005 ). This bureaucracy is frequently highly specialized with a degree of stability in its posts over the years and administrations, creating a kind of ‘bureaucratic elite’. The day-to-day requests for funding are assessed by the budget bureaucracy before proceeding to the decision-making core, imposing a hierarchical structure on the other departments. To bypass the hierarchy, a project must have a new, strategic status and the political backing of the core of government or of key actors. Even the strategic projects must be analyzed by the core of budget bureaucrats, who, together with the core of political actors, is responsible for defining the percentage of discretionary spending in each budget, after calculating the basic budget ( Wildavsky and Caiden, 2004 ). In this scenario, the budget process is likely to involve a debate over the allocation of resources in closed arenas dominated by bilateral discussions and bargaining ( Schick, 1976 ). If mayors have control over the budget bureaucracy, they gain control of the budget, which in turn provides them some discretionary power.

Incrementalism of expenditures

Incrementalism and inertia are two other fundamental elements of the budget process ( Wildavsky and Caiden, 2004 ). Cutting budget expenditures is generally more difficult than obtaining approval for their inclusion. Budget settlement – that is, the relationship that spending creates between beneficiaries and suppliers, the legal support of pre-existing contracts, and the constitutional (or legal) regulations that ensure spending on wages, debts, and other outlays – takes much of the basic budget, making it incompressible and dependent on increments. Some of these expenditures increase as a result of specific legal rules ( Rezende, 2015 ), others do so simply because of the natural inertia of organizations ( PERES, 2018 ). Incrementalism of expenditures reduces the room for new discretionary projects within budgets.

Bargaining and muddling through

Financial and economic difficulties, combined with the growing demand for new public policies and rights, produce a shortfall in financial resources at different levels ( Schick, 1976 ); it is thus upon the core of decision-makers to take on the dominant role in approving or rejecting actions and projects. In this context, this core can better control the budget process through various micro-institutional mechanisms: negotiation, bargaining, and muddling through ( Lindblom, 2009 ). Although this classic view of budgetary decision-making process remains relevant, one must admit that the negotiation between the areas representing spenders and guardians – initially described by Wildavsky (1975) – has changed in the post-managerialist budgetary governance focused on fiscal austerity. According to Kelly and Wanna (2000) , after changes in favor of a more managerialist administration, the guardians – who are tasked with guaranteeing that the fiscal rules imposed by the highest levels of the budget hierarchy are followed – have been largely empowered. As such, even spenders have started to adopt the behavior of guardians, seeking cost reduction in their areas, also as a way to advance future bargaining ( Kelly and Wanna, 2000 , p. 43).

The more complex and rigid the budget rules are, especially those aimed at fiscal austerity, the more difficult it is to negotiate and bargain over the budget. Thus, negotiations will occur if there is room for adjustments and if the budget bureaucracy allows for this negotiation. In this case, the more hierarchical the budget bureaucracy is and the greater control the head of the executive has, the more discretionary power she or he will have.

Political discretionary power

The discretionary power of the head of the executive still plays a role in the budget process, although it is only exercised over a small proportion of the budget. As highlighted by Lindblom (2009) , the policy process evolves incrementally – and so do public budgets ( Wildavsky, 1969 ). Incrementalism, combined with the overlapping of complex budgetary rules and scarcity of funds, leads to rigidity in the definition of public spending. The mayor, however, holds discretionary power over a small part of the budget – such proportion is contingent on the level of scarcity ( SCHICK, 1976 ) and on the degree of control exerted by the core of budget bureaucrats over the baseline budget. In addition, moments of systemic change do exist when policy formulators are faced with urgent and immediate problems, or when major political shifts occur – in such instances, awareness on a specific issue is raised, sometimes leading to budget reallocation ( BAUMGARTNER and JONES, 2005 ).

Budgetary governance involves a natural tension between, on the one hand, the limits imposed by the rules and, above all, by incrementalism – both of which lead to the maintenance of the status quo, that is, to controlled spaces in public funding – and, on the other hand, the pressure of new demands to be included in these spaces. This tension is resolved through the discretionary power of the executive branch. In São Paulo, administrations of different party affiliations resolved this conflict differently, in terms of both how to collect revenues and which type of expenditures would be approved and where. In London, especially in the GLA, different mayors have shown that political strategy and urban projects matter, as demonstrated by Livingstone¹ ( TRAVERS, 2008 ). Institutions and bureaucratic rules are very important, but the political context of a city may also define what mayors can do ( FUCHS, 2012 ).

This dimension is thus crucial for budgetary governance, and it certainly depends on the other dimensions. Moreover, the more complex the budgetary rules are and the more hierarchical the budgetary bureaucracy is, the more centralized the power of the ‘top of the hierarchy’ is. Incremental processes tend to restrict the room for discretionary spending over time ( REZENDE, 2015 ), especially in times of acute scarcity ( SCHICK, 1976 ). In such situations, as budget guardians are empowered, spenders’ ability to bargain for an increment is reduced. Space for discretion is reduced, making this amount of discretionary power more symbolic and politically relevant. According to Fuchs (2010) , in this case, it is crucial to centralize mayors’ budget decisions.

In the next two sections, as the governance of São Paulo and London is compared, relevant elements will be presented in an effort to clarify the characteristics of this governance in reference to the first four analytic dimensions and how these dimensions affect the fifth, the mayor's political discretion.

São Paulo and London: two different cases?

Understanding the distinct state structures is central to comparing these two global metropolises. São Paulo is a city of 12 million inhabitants in a federal country, Brazil, whereas London, a city of 8.9 million inhabitants, is the capital city of the United Kingdom, a unitary state.

Nevertheless, both cities provide mandatory services in a similar fashion. These services involve urban policies such as street cleaning, public transportation, environmental stewardship, public security, housing, and infrastructure. In São Paulo, the education and healthcare policies – which are the city’s most important policies – are formulated and financed within the administrative and budgetary structure of São Paulo. In London, however, these two policies are formulated and financed by the central government, despite being implemented and coordinated at the local level.

An analysis of federal and local laws in Brazil and of central and local regulations in the UK shows that both São Paulo and London are responsible for a wide range of public policies. The city of São Paulo, however, is legally responsible for the education and healthcare policies. In Brazil, these policies have been systematized and controlled by the federal government over the last decades, especially since the 1988 Constitution, when expenditure rules came to be defined nationally and the tasks were divided between subnational governments. In addition, the Constitution defines that these policies are funded with revenues earmarked at the local level, i.e., at least 25% of local taxes must finance education and at least 15% should be allocated to healthcare ² .

Therefore, as a result of the national systematization of these policies, a significant share of the resources in the annual budget of local governments shall be used in preschool, in elementary education, in primary healthcare, and in the financing of medical specialties and hospitals. Although the federal government allocates funds to these two policy areas, federal resources do not reach 15% of São Paulo’s expenditures. Consequently, in São Paulo, almost half of the municipality’s tax revenues are taken even before the budgetary law is prepared; in this sense, increasing tax revenue does not automatically lead to more discretionary power.

As São Paulo, London is partially responsible for the education and social policies (especially the boroughs); although the GLA and the boroughs are not obliged to spend a predefined share of local taxes, most of this funding comes from central grants and is already allocated to specific policies. In some way, expenditure management is restrictive in both São Paulo and London. Public services in São Paulo are jointly organized by the mayor's office, by 35 municipal bodies, and by 32 other sub-municipalities into which the city is divided. In London, the responsibility for the municipal policies is decentralized and involves the GLA, 32 borough councils, and the City as shown in Figure 01 . Despite the coincidence in their numbers, São Paulo’s sub-municipalities do not have the same level of autonomy as the London borough councils, nor the same budget. In London, the head of the GLA – the Mayor – and the council leaders of each borough are independent and elected separately and differently, so they may have different political party affiliations. In São Paulo, sub-mayors are appointed by and report to the mayor.

Figure 01
São Paulo’s and London’s regional divisions: sub-municipalities and boroughs.

Moreover, sub-municipalities in São Paulo have very limited responsibilities compared to London’s boroughs. They do not have tax autonomy; in practice, they are only responsible for janitorial services such as street cleaning and landscaping. In London, on the other hand, the boroughs are in charge of implementing important social policies, such as education, healthcare, planning, housing, cultural policies and related services, highways, and environmental and regulatory services. Additionally, they are allowed to collect council taxes.

As for the metropolises’ political organization, both São Paulo and London have their own executive and legislative branches. The legislative bodies of both cities participate in the budget process: they receive and analyze the budget. The London Assembly, however, does not have the power to exert the same control and to bargain as the City Council of São Paulo does. It is quite difficult for the Assembly to have an amendment proposal accepted, and it is only possible to reject the budget with a two-thirds majority. In São Paulo, a simple majority is required for this voting process and the budget is generally subject to amendments. In any case, mayors in this city usually have a majority in the City Council, which is guaranteed by broad coalitions – more than 30 political parties can be represented in São Paulo. In forging these coalitions, mayors can ensure governability and, to a certain degree, a leading role in the budgetary process ( Marques and Hoyler, 2018 ). This leading role is also true for the head of the GLA in London, whose administrations follow the British tradition of a parliamentary majority in the Assembly³ - such majority is usually determined by one party, or by a coalition of two parties at the most.

With respect to public revenues, the degrees of tax and financial autonomy also differ between the two metropolises. As part of the Brazilian federation, the city of São Paulo has autonomy over local taxes and levies – the city’s main sources of revenue are the property tax (IPTU) and a tax on services (ISS), as shown in Figure 02 . In addition, the city of São Paulo receives tax transfers from the state of São Paulo (vehicle tax (IPVA) and a tax on goods and services (ICMS)), which vary according to the rate charged by the municipality. It is important to notice that the property tax (IPTU) as a share of São Paulo’s main revenues have increased in recent years due to a law proposed by Mayor Haddad (Worker's Party) in 2013 to adjust property values by up to 30%. Although this measure was quite unpopular and politically costly for Haddad, it is perceived today as fundamental for preserving the city's revenues during fiscal crises.

Figure 02
São Paulo’s main taxes and transfers as a percentage of current revenue* – 2008/2018.

Finally, São Paulo also receives grants from the federal government, chiefly for funding the healthcare system (SUS), but also other policies. Nonetheless, these grants make up a small share of São Paulo’s budget, as shown in Table 01:

Table 01
São Paulo’s and London’sFunding - Annual Budget

São Paulo’s Expenditure per inhabitant: £870	São Paulo’s Annual Budget	London’s Expenditure per inhabitant: £8,890	London’s Annual Budget
Federal Level Funding	0.7 billion (6,7%) (1)	Central Gov (NHS, Housing and others)	48 billion (56%)
State Level Funding	0.5 billion (4%) (2)	GLA and GLA Group	16.6 billion (18%) (4)
Municipal level	9.8 billion (89,3%) (3)	Boroughs	22.4 billion (26%) (4)

• Notes:(01) Corresponds to transfers shown in Figure 02 , which include the FPM (Municipality Participation Fund), transfers to SUS (Unified Health System), and Fundeb, a funding for education; (2) corresponds to state taxes shared with the municipality, shown in Figure 02 : ICMS and IPVA; (03) corresponds to local taxes: IPTU, ISS, and ITBI; (04) corresponds to council tax collection, business rates (presented in Figure 03 ) and, eventually, loans.
Source: Author’s creation based on data presented by Prefeitura de São Paulo (2018) and by UK Public Spending (2018) and UK Government (2012).

In contrast, the budgets of the GLA and the London boroughs rely heavily on central government grants, which are defined by the National Treasury based on local needs and on formulas for distributing annual revenue among local governments. Part of the resources allocated to London are managed by the GLA, which has five public bodies: MOPAC (Mayor’s Office for Policing and Crime), LFC (London Fire Commissioner), TfL (Transport for London), LLDC (London Legacy Development Corporation), and OPDC (Old Oak and Park Royal Development Corporation) ( GLA, 2019a ). Another share of the resources is managed by the borough councils, which are responsible for delivering most of the urban public services and social policies. Finally, the central government and its ministries are responsible for the remaining resources, which are chiefly allocated to education and healthcare. Thus, as presented in Table 01 , London’s funding depends mostly on central grants, which represent 56% of the public funds.

Public funding in the UK is structured in accordance with the country’s highly centralized tax system. The GLA and the boroughs are only allowed to collect council tax (equivalent to property tax) and business rates (non-domestic property tax). All the remaining taxes, such as income tax and national insurance contributions (NICs) – which are the two main sources of tax revenue in the UK, as shown in Figure 03 – are controlled by the central government.

Figure 03
Evolution of the composition of UK net taxes and NIC Receipts*.

Local taxation is thus quite limited. With respect to the council tax, which represents 4.3% of the total tax revenue in the UK ( Institute for Fiscal Studies [IFS], 2016 ), the central government defines the tax rate bands every year, while the tax base (property) is defined and evaluated by the central government. The last time this was done was in 1993 ( Slack and Bird, 2014 ). Business rates, which make up 4.5% of total tax revenues, are also quite controlled by the Treasury, which receives 50% of this tax to be redistributed according to fiscal balance criteria ( Sandford, 2019 ).

In what concerns tax revenues, it is worth noting that São Paulo manages a higher number of taxes and has greater autonomy than London, which is more dependent on central government grants.

Table 02 summarizes the general comparisons made in this section between São Paulo and London.

Table 02
São Paulo and London: general comparison of political, administrative, and financial structures.

	São Paulo	London
State Structure	Federal	Central State
Administrative Decentralization	Mandatory local public services and political responsibility	Mandatory (Statutory) local public services (The GLA and 32 Boroughs +City) and political responsibility
Political Decentralization	Elected members of the executive and legislative branches / constitutional autonomy - 32 Sub-Municipalities indicated by the Mayor	Executive Mayor (elected by citizens, Legislative Power (Assembly) with few powers (GLA) - 33 Boroughs/Councils: elected mayor/leader and cabinet or committee system 5
Financial Decentralization	-Partial Budgetary Autonomy: budgetary laws & earmarked funds -Partial Fiscal Autonomy -Tax Autonomy -No Bank Autonomy (no access to credit)	-Budgetary Autonomy: annual budget and mandatory competences -Low Fiscal Autonomy -Low Tax Autonomy -Bank Autonomy (access to credit)
Two economic capitals	-Privileged access to state and federal administrations; - Sub-municipalities - low budgetary autonomy; -Pressure from interest groups;	-Privileged space in central funds (capital city) - High Dependency; -Boroughs – autonomy to deliver public services – social pressure; -Pressure from Central Gov to reduce expenditure -Pressure from interest groups

Source: Elaborated by the author’s.

São Paulo’s and London’s budgetary process and governance

After providing an overview of the administrative and political structures of the two metropolises, in this section we offer a comparative analysis of the dimensions of the governance of public budgeting presented in section 02.

Complexity of budgetary rules

The budgetary process in both São Paulo and London is a complex rite with pre-established rules. In São Paulo, while the Constitution establishes the budget triad (Multi-Year Action Plan, Budget Guidelines Law, and Annual Budget Law), the city's Basic Law stipulates the target programs. All elements of the budget are drafted by the executive branch according to fixed legal calendars, which are assessed, approved, and controlled by the legislative branch with the support of the municipality's court of auditors. Since 2000, São Paulo’s budget – like that of every city in Brazil – is regulated and defined by a federal law, the Fiscal Responsibility Law (LRF). This piece of legislation is a very strong policy instrument ( Lascoumes and Le Galès, 2005 ) that has been used by budgetary guardians (both federal and local) to reduce indebtedness and constrain personnel expenditure, which leaves the government facing fiscal and political constraints.

In London, the budget rules have been established mainly by the British Parliament and adjusted in recent decades in a centralized manner, contrary to the more intensive devolution processes for local authorities during the last decades of Labor government ( Wilson and Game, 2011 ). The exception is the case of the GLA, whose rules were created in 2000 by the ‘Greater London Authority Act 1999’, which filled the void left by the absence of a metropolitan authority for about 20 years⁶ . Both the GLA and the borough councils are in charge of the annual budgeting process, which relies to a great extent on central grants from the Ministry of Housing, Communities, and Local Government. The Treasury is responsible for establishing budgetary rules and the annual expenditure parameters for all legal authorities. According to Jones et. al. (2013) , the main budgeting framework for expenditure control is the ‘Consolidated Budgeting Guidance’, published annually by the Treasury. With respect to expenditure parameters, an important pluriannual (general) parameter is the Spending Review, with which the Treasury sets long-term targets for current expenditures and investment and announces eventual changes in taxation ( Ferry and Eckersley, 2011 ). Finally, there is the Annual Budget, which consists of two statements. First, the Autumn Budget, an annual budget set by the Treasury for the following financial year, with public-sector revenues and expenditures. Second, the Spring Statement, also known as the ‘mini budget,’ made according to adjustments in the economic forecast ( HM TREASURY, 2020 ).

It is worth highlighting that, as the Brazilian Fiscal Responsibility Law, the Spending Review became the main instrument for implementing the fiscal austerity policy adopted by the Conservative/Liberal Democratic coalition in 2010 ( Ferry and Eckersley, 2011 ). This budget instrument sets deficit reduction targets and is used as a governance instrument ( Le Galès, 2016 ) by the Treasury to define targets and to control the actions of local authorities, including the GLA. The power of this instrument is not only imbedded in budgetary complexity, but also in the public acknowledgement of fiscal austerity.

Hence, in the last 10 years, based on the austerity parameters defined by the Treasury, which are established in the Financial Act passed by the Parliament, the GLA and the borough councils have been developing their annual plans and presenting them to the London Assembly, in the case of the GLA, and to the Councils, in the case of the Boroughs. Each plan is formulated by the respective relevant financial areas. In the case of the GLA, this is done by the Group Finance Manager, which is responsible for consolidating the budgets of the five public bodies; in the Boroughs, such responsibility lies with the Cabinet member in charge of the finances. Standardizing the budget language between the two publications is not a concern, probably because these two levels of government are independent.

In São Paulo, the budget process starts with a four-year budgeting forecast known as the Multi-Year Action Plan (PPA), a specific law approved by the City Council during the first year of the mayors’ term in office, to be implemented in the next 4 years.

Thereafter, in São Paulo, annual guidelines are established in the Budgetary Guidelines Act (Lei de Diretrizes Orçamentárias - LDO). The Annual LDO determines how the Annual Budget will be formulated, if any tax will be increased and which will be the fiscal goals. After creating this frame of reference, a discussion about the budget begins within the Secretariats/Departments. It starts with the definition of parameters by SUPOM (The Subsecretariat of Planning and Municipal Budget) according to the guidelines established in the PPA and the Annual LDO. Each department must then prepare its individual budget and discuss its needs with the sectoral councils and with civil society – in London, in contrast, this discussion may occur in the boroughs, but it is not required by law as in São Paulo. After the period of consultation with the public and with the sectoral council, the head of each department submits a sectoral budget to be approved by the Financial and Budgeting Board (JOF). Discretionary expenditures are the number-one issue of contention between the target areas and the JOF.

The Mayor deliberates the final budget with the JOF, after which the executive branch sends the Annual Budget Project to the City Council by September. Then, city councilors in the Council Finance Committee discuss and propose amendments. As part of the budgeting game, the Mayor and the Mayor’s office must negotiate these amendments with the city councilors. The Annual Budget process is concluded when the São Paulo City Council votes the budget. If the Mayor has a solid majority in the Council, this represents a merely symbolic stage. In London, the Financial Commission in the London Assembly normally approves the budget without amendments.

Understanding the rules and the budgets is a difficult task in both metropolises. Although several organizations and think tanks have analyzed the budgets and tried to translate them to civil society in London – as pointed Sandford⁷ (2019) – and in São Paulo, the budget remains a complex and hard subject, and thus a restricted domain.

In these two cases, the fact that budget rules are complex increases the discretionary power of mayors, since they can control the budgetary bureaucracy. Because most of the budget expenditures are earmarked by constitutional rules and by either federal rules, in the case of São Paulo, or central government rules, in the case of London, the space for discretionary decision is reduced. Discretionary expenditures have to be negotiated either in a broad participative fashion or determined by a decision of the mayor; due to the complexity of the rules, the latter tends to prevail.

Hierarchy of the budgetary bureaucracy

In this subsection, I present some similarities – and also some important differences – between the two cities. In São Paulo, the annual budget process is overseen and established annually by the budget bureaucracy, i.e., the Subsecretariat of Planning and Municipal Budget (SUPOM), which is under the Finance Secretariat. It provides bureaucratic support to the Financial and Budgetary Board (JOF), which consists of four secretariats: finance, government, administration, and justice. This board makes all decisions on spending parameters with the support of the budgetary bureaucracy (SUPOM) and the treasury bureaucracy, which estimates the revenue collection for each year. Thus, for new expenditure to be included in the budget of a given secretariat, housing for example, the creation or supplementation of the budget must be formally required; the request must be analyzed by the budgetary bureaucracy and approved by the JOF. Generally, over the year, only the Secretary of Treasury, the Board, and the mayor have information about revenues. In this budgetary hierarchy, the mayor always has the last word on the budget.

In the case of England, there are two types of day-to-day hierarchies: one that is internal to the GLA; another between the local levels and the central government. Regarding the decision-making hierarchy of the GLA budget, it is important to mention the existence of a program for authorizing internal expenditure: ‘Mayoral Decision-Making in the Greater London Authority’ (Great London Authority [GLA], 2019b). This program establishes four different hierarchical levels for decision making on expenditures depending on the type and value of expenditure, ranging from a program expenditure of £10,000 resulting from a managerial decision, to a project that costs more than £150,000 and has to be decided by the Mayor, to initiatives that involve something novel, contentious, or repercussive (NCR) or that are categorized by the law as ‘Reserved Mayoral Matters’.

Regarding the existing hierarchy connecting local authorities to the central government, as mentioned previously, the Finance Group in the GLA and the Cabinet member in the Borough councils are responsible for consolidating the budget expenses of their units. The Treasury defines the values of the central grants to be transferred annually to the Borough Councils and the GLA. After defining these values and factoring in the information on necessary expenditures, the financial areas of the GLA and the Borough Councils organize their annual budgets and define the amount of council tax collection they need to cover the expenses⁸ . It is important to say that both the Mayor and the Cabinet Leaders in the Councils are ultimately responsible for the budget, although the Mayor in London has no influence on the Council budgets.

Given the Brazilian federal arrangement, São Paulo has greater autonomy than London in its day-to-day budget process, despite being subject to the hierarchy of the federal bureaucracy and to the rules of public finance set by the Ministry of the Economy. London’s bureaucratic hierarchy - even the GLA - is linked to the Treasury. For some authors, such as Gordon and Travers (2010) , this dependence and tax limitation imposed on London by the Treasury, despite being an effort to ensure tax equalization, represents a response to the fear that an independent and autonomous London City Hall could be politically threatening to the central government, as was the case with the former Great London Council - GLC ( TRAVERS, 2004 ).

It is worth mentioning that, in the context of this vertical hierarchy, the Spending Review is used by the central government as the main policy instrument for controlling both the public deficit and the actions of local authorities.

Incrementalism of expenditure

As mentioned in the theoretical section of this paper, incrementalism of public spending is one of the most studied subjects in budget theory. In the case of São Paulo, as shown in Peres (2018) , personnel expenditure (especially pensions) and expenditure on third-party services increased more than the growth in revenues between 2003 and 2014. Although revenue has increased in this period, this increment in the baseline budget led to a reduction in investment expenditure, especially in urban infrastructure.

Between 2014 and 2018, São Paulo’s revenues were affected by the Brazilian economic and fiscal crises, leading to other budgetary adjustments. Given the existence of incompressible expenditures (see Figure 04 ), the amount of budgetary resources has been the same across three different administrations, while the amount of discretionary resources in the budget was reduced. This has happened because most of the expenses are mandatory – or, as Rezende (2015) says, they are incompressible.

Figure 04
São Paulo Annual Budget: 2010-2018.

As highlighted by Rezende (2015) , incompressible expenditures have become consolidated in recent decades: they cannot be cut without a legal decision, and the funding for these expenditures are earmarked in the public budget. The main incompressible expenditures are personnel expenditures, healthcare and education expenditures (which are funded with earmarked tax revenues)⁹ , and debt payment to the federal government. In this category are also included the expenditures of the legislatures and court-ordered debt payments. Finally, in this group there is also room for credit agreements with the federal government (Programa de Metas – Target Program) and for Private-Public Partnerships (PPPs). These expenditures are also incompressible because credit for urban infrastructure must be used by mayors within parameters that are predefined by the federal government; also, mayors must provide matching funds.

As shown in Figure 04 , most of the expenditures are incompressible – about 80% of net current income. As a result of all expenditure rules, in 2010 approximately 7,5 billions out of 10 billions pounds were incompressible expenditures, and in 2018 the figures were 09 out of 11 billions of pounds, which means that less then 02 billion pounds were left for compressible spending.

However, compressible expenditures in São Paulo’s budget are not really considered discretionary expenditures. Compressible expenditures include expenses associated with urban services, that is spending on collective transportation, city waste management, and street lighting. These are basic services and cannot, therefore, be discontinued. The remaining 0,8 billion pounds are used to pay a myriad of expenses that – although discretionary – are necessary to keep essential services running. These services allow for the implementation of policies such as social welfare, human rights, culture, sports and leisure; they also maintain the target areas and services provided by the sub-municipalities.

Our analysis of the data on São Paulo’s budget indicates that there is only about 250 million pounds to be freely allocated, a discretionary increment that has been reducing over the past years ( PERES, 2018 ).

When comparing the budgets of 2010, 2014, and 2018 ( Figure 04 ), one can see that while compressible and discretionary expenses were reducing, the baseline budget – education, healthcare, and personnel – was increasing incrementally. The margin for discretion was thus reduced to balance the budget.

The target program from 2014 to 2018 was scaled down and the level of debt payments was reduced. The target program was scaled down because the federal government had cut back on its capital transfers, on which the program depends. This cutback was the result of the tightening of the fiscal austerity policy at the federal level due to the approval, in 2016, of Constitutional Amendment 95/2019, known as the Spending Ceiling Amendment. In view of this amendment, federal investments and capital transfers to municipalities were reduced. Indebtedness declined in São Paulo because Mayor Haddad managed to renegotiate with the federal government, which allowed for the debt to be recalculated and for the annual payment to be reduced. Moreover, if property tax collection had not increased (as mentioned in section 03), an even larger spending reduction would have been necessary to avoid deficit.

In London, the process of incremental adjustment of expenditure is more complex due to the events of the last 10 years. By analyzing budget data of the UK central government over a long period (1951 to 1996), John and Margetts (2003) sought to test whether the spending activity in England tended towards incrementalism or whether they were closer to Baumgartner and Jones’ (2005) idea of punctuated equilibrium. They found that there are punctuations, but only in part of the expenditures. These authors state that different budgetary strategies may be found across policy sectors. While some policies depend highly on human resources, and therefore most of their budget is consumed by expenditures on personnel, others are protected by entitlements, so that their variations are more unlikely to be punctuated. Thus, those bodies associated with healthcare (the NHS), education, and social security have expanded incrementally over the period, despite a series of political and legal changes. Also, there were more punctuations in urban infrastructure ( John and Margetts, 2003 , p. 429).

Figure 05 shows the evolution of expenditure (by tier of government: central government, GLA, and councils) in London in the last decades. Expenditures grew between the late 1990s and 2009 and, after a period of stagnation, they have grown again incrementally over the past five years, despite the central government’s fiscal austerity policy, which has been in place since 2010.

Figure 05
London’s Annual Expenditure

However, data on London’s budget are aggregated; as a result, variations in the two other tiers are concealed. When analyzing GLA and Council expenditures alone, some issues need to be highlighted.

Figure 06 below shows that, from 2013-14 to 2018-19, the level of expenditure of the GLA Group slightly changed, with only TfL showing some growth, while the other public bodies have maintained or reduced their expenditures. Especially relevant is the reduction in MOPAC, the agency responsible for policing. Considering that this agency's expenditure with salaries and payroll taxes continued to increase slightly and incrementally, this reduction in the agency’s overall expenditures meant that new employees were not being hired and that investment in equipment and public security programs was cut back, which was broadly criticized by Londoners and academics ( Dodd, 2019 ).

Figure 06
London GLA Budget: 2013-14 and 2018-19.

It is worth mentioning that, as in São Paulo, the GLA's budgetary situation would be even more complicated if the current mayor Sadiq Khan (Labour Party) had not increased the council tax annually, which can be adjusted by up to about 3%. Although revenue from council tax is low compared to the revenue from the central government, it gave the mayor a certain degree of discretionary power. In the previous term, Mayor Boris Johnson (Conservative Party) chose to keep the council tax frozen. According to Kahn’s speech during the budget plenary on January 29, 2020, this freeze had cost London £700 million.

Council expenditure is even more critical than that of the GLA, as shown in Figure 07 . In their analysis of the austerity policy, Ferry and Eckersley (2015 , 2012 , 2011 ) stressed that the consequences of the fiscal adjustment were felt mainly by UK local authorities. In London, some Councils have been affected by cuts of about 30% in the last 10 years, according to Harris at al. (2019).

Figure 07
Budget of London Borough Councils

Figure 07 shows that the London borough councils’ total expenditures were reduced between 2011 and 2019. The exception was expenditure on social care, which slightly increased. This increase, however, is deemed insufficient in view of the growing demand from beneficiary groups (especially adult care) in the last years ( Harris et al., 2019 ). Budget cuts have affected services and increased inequality in care, especially among the most vulnerable groups, as pointed out by Vacchelli et al. (2015) .

Thus, considering that the overall expenditure in London has increased (mainly due to the NHS), one can infer that council expenditures have been more severely reduced to allow for the necessary global adjustment ( Eckersley and Tobin, 2019 ).

According to an evaluation by the Local Government Association ( LGA, 2020 ), which was corroborated by the Institute for Fiscal Studies (IFS), councils cannot manage the reduction in central grants in recent years due to their limited autonomy in raising council taxes and using revenue from the business rate ( HarrIS et al., 2019 ). As a result, cuts were inevitable and have been felt mainly in policies related to urban infrastructure, such as housing. Affordable housing and public security are Londoners’ main current concern.

It is noteworthy that the austerity policy implemented by the Conservative/Liberal Democratic Coalition – which remains in place to this day – resonates with Schick’s 1976 analysis stressing the need to take into account the fact that there is a general lack of understanding of budgeting, as well as with Fuchs’ 2010 analysis, which suggests that, in moments of scarcity, Laswell’s question should be altered from ‘Who gets what?’ to ‘Who gets cut?’

Hence, by assessing the evolution of GLA and council expenditures, one can identify decrementalism rather than incrementalism in different expenditure categories. However, Baumgartner and Jones’ (2005) punctuation hypothesis does not apply very clearly in London’s case, since apparently there has not been any structural change to the budget in terms of public policies. The same policies are still being implemented, but with fewer resources and thus with precarious services.

What we can highlight in this section is that there is a set of rules (constitutional, in the case of São Paulo, and central grants, in the case of London) that defines much of the budget of the two metropolises and reduces the mayors’ margin for discretion. This margin may be further narrowed by an increment in compulsory expenditures or by criteria for fiscal austerity determined by the federal and central levels of government.

Bargaining and negotiation in the budgetary process

To make a comparative analysis of bargaining in the budget process, one should consider two levels of negotiation: vertical and horizontal. Vertical bargaining involves different administrative spheres, with local authorities seeking either a greater volume of intergovernmental grants, as in the case of Brazilian federalism, or the release of central grants, as in the case of London (demanded by the GLA or the Councils). As for the horizontal level, it involves a classic bargain between spenders and guardians, as defined by Wildavsky (1975) .

The budgetary process in the city of São Paulo is considerably rigid – up to 90% of revenues may be allocated to mandatory expenditures, depending on the level of capital revenue ( PERES, 2018 ). Thus, the room for horizontal bargaining in the budgetary process – between spenders and guardians – is limited, particularly between target areas and the decision-making body of the Board (JOF), and ultimately the Mayor, who is the centralizer of decisions ( Fuchs, 2010 ). The higher the level of scarcity is, the less room for bargaining there is and the fiercer is the conflict between areas ( Schick, 1976 ).

The London case is different to some degree in this aspect. Since the power in London’s political budgetary structure is concentrated in the Treasury and because there are predefined formulas for allocating the annual resources, bargaining becomes a difficult task. As reported by different interviewees, bargaining in trading may take place by changing the formulas for central grants or other rules, such as the recent request that 100% of the business rate could be retained by councils and not partly returned to the central government, as occurred until 2019. Negotiations for the authorization of new grants for housing, for example, are organized by the LGA (Local Government Association [LGA], 2020). However, it is harder for councils to bargain individually with the Treasury.

Hence, the GLA, represented by the Mayor, is more capable of gaining individual bargaining power. As observed by Gordon and Travers (2010) , and Travers (2008) , the mayor of London can exercise strategic leadership and, as such, attract private investments, which facilitates the negotiation of new projects with the central government. This political power was highly emphasized during Livingstone’s term in office ( Travers, 2008 ).

Discretionary power of mayors

In both São Paulo and London, discretionary revenues represent a small part of the budget, which compels the head of the executive to make political choices. Budgetary governance involves a natural tension between, on the one hand, the limits imposed by the rules and, above all, by incrementalism – which leads to the maintenance of the status quo, that is the controlled spaces in public funding – and, on the other hand, the pressure for new demands to be included in these spaces. This tension is resolved through the discretionary power of the executive branch. According to Fuchs’ (2010) analysis, the centralization of the budgetary process results in the mayor being the ultimate decision maker.

In São Paulo, as already indicated, the room for discretionary expenses has been reduced in recent years. The political discretion of mayors in São Paulo depends on the amount of expenses over which they have control, which in turn depends largely on capital revenue, i.e., credit operations, privatizations, urban operations¹⁰ , among others, since tax revenues are earmarked. In recent years, with the worsening of the city's fiscal crisis, São Paulo’s government has increasingly relied on resources from capital revenues, especially urban operations.

Recent changes in the city’s Strategic Master Plan enabled an increment in the expected collection for the Urbanization Fund (Fundurb), according to Paim (2019) . This could be a way to increase the political discretion of the Mayor. Additionally, the mayor can always increase tax rates, especially property tax, but most revenue is earmarked (approximately 65%), which leads to a fierce distributive conflict between the Mayor and the wealthiest sectors of the city ( Borges, 2018 ).

In what concerns the political discretion of mayors, there are differences between London and São Paulo. First, the mayor of London controls the resources of the GLA – the GLA represents the consolidation of five public bodies, two of which are strong bureaucratic structures in which the mayor hardly interferes: TfL and MOPAC. Furthermore, Borough Councils are not subject to the Mayor’s authority. At best the mayor may have some influence on the city’s urban policy, or act as a facilitator. According to Gordon and Travers (2010): “Underlying this notion of the Mayor as ‘facilitator’ is the fact that the lower tier (borough) authorities retained almost all their powers. Their relative strength is signalled by a level of collective spending which is twice that of the GLA, a balance which continues to differentiate London’s system of governance even from that of those other big cities, such as Berlin and Tokyo” ( GORDON and TRAVERS, 2010 , p. 05).

Thus, despite having the entire GLA structure at his or her disposal and despite the symbolic weight associated with leading one of the most important metropolises in the world, the Mayor of London does not have the fiscal autonomy to implement the urban policies necessary for developing the city. The mayor's strategy and political habilities could make a difference and translate into bargaining power in negotiations with the central government, as did Boris Johnson during the negotiation for investments for the 2012 Olympic Games ( Bennet, 2016) . The current mayor, Sadiq Khan, has also recently shown to have some bargaining power as he obtained £1.7 billion for a social housing program to be implemented by the Councils ( MAYOR OF LONDON, 2018) . Despite these successful negotiations, mayors are structurally constrained in their ability to exercise political discretion over London’s budget.

The London Plan, which defines planning and development strategies for Greater London could be a strong tool to boost resources that lack in capital revenues. However, according to Gordon and Travers (2010) , the Mayor’s London Plan: “lacks any capital budgeting component, since the Mayor has neither overall responsibility for, nor effective control of, the major investment resources on which implementation of strategies for a growing city would clearly depend” ( GORDON and TRAVERS, 2010 , p.06).

Therefore, despite the budget of the city of London being ten times larger than that of São Paulo, such financial advantage does not seem to translate into more political discretion for mayors in London; this is due to the current structure of the GLA and of the councils and also because of the strong control exercised by the central government over the city’s expenditures and taxes.

Conclusion and final remarks

This paper has presented a comparative analysis of the budgetary governance of two large metropolises, São Paulo and London.

First, it is important to mention some limitations of this study. One of these is that expenses with programs were not analyzed in either of the two cities – such step would be important to identify the incrementalism effect of educational or healthcare programs. So far, it has not been possible to work with these data in intertemporal analysis. These data are being analyzed in order to go deeper with this study and to enable a more accurate analysis of incrementalism effects based on econometric methods.

Despite this limitation, the quantitative and qualitative data allowed for the two metropolises to be compared based on the dimensions presented in section 02, and, with that, this study was able to draw some conclusions. A brief summary of the similarities and differences regarding the five dimensions is shown in Table 03 – the complexity of budgetary rules being one of these dimensions, with both cities tending to develop a ‘specialized’ bureaucracy. Another similarity is the hierarchical organization of the budget process and the high centralization of power in the head of the executive.

Table 03
Comparison of Budgetary governance in São Paulo and London

Dimensions	São Paulo	London (The GLA and The Boroughs)
Complexity of budgetary rules	Budget with two preparation rounds, a calendar, and legal rules; implementation power is concentrated in the executive branch.	Calendars, segmentation of the budgetary process, administrative decisions. Power is concentrated in the executive branch. (Rules have been implemented for centuries and they depend on definitions by the Treasury).
Hierarchical organization of the budget/ budget bureaucracy’s influence on the decision-making process	The budgetary process is structured similarly to the budgetary process involving the Ministry of Finances and other ministries at the national level; locally, however, the process is even more insulated. Power is concentrated in the Budgetary Board and in the mayor.	Structuration of the budgetary process depends strongly on Central Government decisions and, secondly, on the mayor or Borough leader. The Assembly scrutinizes the budget; to reject it, however, this legislative body needs 2/3 of the votes.
Incrementalism and inertia in the allocation of resources (Decrementalism)	Almost 90% of the expenses are already allocated. There is little negotiation about investments, especially at the beginning of new administrations. Current economic crisis has reduced various expenditures.	London’s budgets are in a decremental path since 2010 – with limits being set by the Spending Review”
Bargaining and muddling through	All secretaries concurrently try to bargain with the Budgetary Board.	There is some bargaining between the GLA and the Central Government. But, in general, funding for the boroughs depends on technical formulas.
Discretionary power of politicians in allocating a small part of the financial resources	The mayor decides on the allocation of a small but symbolic part of the financial resources; The Mayor must resolve the distributive conflict over public funding; Relative tax autonomy guarantees some discretionary power.	The mayor decides on the allocation of a small but symbolic part of the financial resources; The distributive conflict is partially affected by the funding and decisional power of the boroughs; Low tax autonomy reduces discretionary power. But the importance of London – as a metropolis – guarantees the mayor the informal power in strategic decisions (not always in the GLA budget).

Source: Elaborated by the author’s.

These two characteristics of the budgetary governance of London and São Paulo tend to increase the discretionary power of the mayor in what concerns his or her ability to shield and improve the budget bureaucracy. As budgeting is mastered only by a few people, the heads of the executive can easily handle the distributive conflict with the use of discretionary resources. This discretionary power tends to diminish in the face of pressures from other hierarchical levels of government – central and federal, respectively – and in the context of fiscal austerity rules (Spending Review in the UK; Fiscal Responsibility Law (LRF) and CA95/16 in São Paulo), as they undermine the chances of increasing the budget.

Fiscal austerity policies have effectively reduced incrementalism of expenditures. In both cases, expenditures are ‘subject’ to incrementalism, especially personnel expenditures. In the case of London, however, the fiscal austerity policy implemented in 2010 disrupted the process of incremental evolution of expenditures and initiated instead a process of decremental adjustments. This disruption was the result of a severe restriction imposed by the central government on the discretionary power of the Mayor and Council leaders to manage expenditures.

It is important to mention that although some increments were made in expenditures in São Paulo’s budget during the period under study (especially personnel and retirement), investments were reduced.

With respect to the political discretionary power of mayors, there are similarities and differences between the two municipalities. Regarding the differences, the discretionary power of the mayor of São Paulo is theoretically greater than that of the head of the GLA, since the final decision on the overall budget in the Brazilian city is centralized. In addition, the mayor of São Paulo has autonomy to change legislation regulating two important taxes, property and services, something impossible in London. However, the discretionary power of the mayor of São Paulo is intrinsically constrained. First because most expenditures are pre-defined by law and earmarked (specially education and healthcare), and thus cannot be easily reduced, so that increasing tax collection also means increasing expenditures. Also, raising taxes in São Paulo is not a trivial task; tax increases must be negotiated in the City Council and are met with fierce opposition from strong interest groups in the city. Nevertheless, despite the difficulties around these political and social negotiations, increasing property tax has been a means for retaining discretion over part of the expenditures in São Paulo’s budget as of 2014, after Brazil’s economic crisis.

Therefore, in that respect, there is a similarity with London’s budgetary governance, since Mayor Sadiq Khan’s decision to annually increase the council tax has given him some discretionary power over the budget, although it has also attracted criticism from the Conservatives in the political arena.

Taking into account the studied dimensions, we conclude that São Paulo has more autonomy in terms of fiscal revenue than London, but when it comes to mandatory spending like healthcare, education, and personnel, discretion is reduced by the central authority in both countries, although by different instruments: the Constitution and federal laws – in São Paulo’s case – and budget allocation through central grants, in London’s case. This finding shows that, even though Brazil and the UK represent different types of state, they apparently have a similar level of spending discretion.

In this way, the analysis of these two cases seems to reinforce the hypothesis, already discussed in the political science literature, that federalism or centralism are not sufficient explanations for the decentralization of public polices and autonomy of local governments. By further analyzing local tax policy and its relation to expenditure, scholars may better understand mayors’ discretion over public policies.

This hypothesis draws on the notion that, if top-down austerity policies and increments in expenditures have reduced the discretionary power of mayors in both cities on the one hand, the increasing of property tax, on the other hand, has been fundamental to offset this loss of power in the budgetary arena.

* Acknowledgments

The author would like to thank professors Anthony Pereira (Brazil Institute/KCL) and Patrick Le Galès (Sciences Po) for supporting this research and for all the suggestions to improve this paper. Thank are also due to all the colleagues from Center for Metropolitan Studies (CEM/USP) and the BPSR reviewers.

References

• BAUMGARTNER, Frank R. and JONES, Bryan D. (2005), The politics of attention: how government prioritizes problems. Chicago: University of Chicago Press. 304.
• Bennet, Asa (2016), Boris Johnson’s London legacy: a happy megacity loves by the world, but not by Britain. Available at ˂https://www.telegraph.co.uk/news/2016/05/05/boris-johnsons-london-legacy--a-happy-megacity-loved-by-the-worl/˃ Accessed on November, 05, 2019.
  » https://www.telegraph.co.uk/news/2016/05/05/boris-johnsons-london-legacy--a-happy-megacity-loved-by-the-worl/˃
• Bezes, Philippe and Siné, Alexandre (ed.) (2011), Introduction: gouverner (par) les finances publiques. Perspectives de recherche. In: Gouverner (par) les finances publiques. Paris: Presse de Sciences Po. pp. 17-111.
• BORGES, Luís Maurício Martins (2018), Imposto sobre propriedade: uma reflexão crítica do caso do município de São Paulo e da cidade de Nova Iorque. Revista Brasileira de Estudos Urbanos e Regionais. Vol. 20, Nº 03, pp. 570-588.
• Caiden, Naomi (2010), Challenges confronting contemporary public budgeting: retrospectives/prospectives from Allen Schick. Public Administration Review. Vol. 70, Nº 02, pp. 203-210.
• Caiden, Naomi (1994), Budgeting in historical and comparative perspective. Public Budgeting and Finance. Vol. 14, Nº 01, pp. 44-57.
• CIPFA (2019), Council tax increases in England to be second highest in ten years. Available at ˂https://www.cipfa.org/about-cipfa/press-office/latest-press-releases/council-tax-increases-in-england-to-be-second-highest-in-10-years˃ Accessed on November, 05, 2019.
  » https://www.cipfa.org/about-cipfa/press-office/latest-press-releases/council-tax-increases-in-england-to-be-second-highest-in-10-years˃
• Core, Fabiano Garcia (2006), Reforma gerencial dos processos de planejamento e orçamento. In: Planejamento e orçamento governamental. Edited by GIACOMONI, James and PAGNUSSAT, José Luiz. Brasília: ENAP. pp. 219-261.
• Dodd, Vikram (2019), Number of homicides in London climbs to 10-year high. The Guardian. Available at https://www.theguardian.com/uk-news/2019/dec/31/number-of-homicides-in-london-climbs-to-10-year-high Accessed on December, 05, 2020.
  » https://www.theguardian.com/uk-news/2019/dec/31/number-of-homicides-in-london-climbs-to-10-year-high
• Eckersley, Peter Mark and Tobin, Paul (2019), The impact of austerity on policy capacity in local government. Policy & Politics. Vol. 47, Nº 03, pp. 455-472.
• FERNANDES, Antônio, Sérgio Araújo (2002), Path dependency e os estudos históricos comparados. BIB. Vol. 53, pp. 79-102.
• Ferry, Laurence and Eckersley, Peter Mark (2015), Budgeting and governing for deficit reduction in the UK public sector: act three ‘accountability and audit arrangements’. Public Money & Management. Vol. 35, Nº 03, pp. 203-210.
• Ferry, Laurence and Eckersley, Peter Mark (2012), Budgeting and governing for deficit reduction in the UK public sector: Act 02 – ‘the annual budget’. Public Money & Management. Vol. 32, Nº 02, pp. 119-126.
• Ferry, Laurence and Eckersley, Peter Mark (2011), Budgeting and governing for deficit reduction in the UK public sector: act one 'the Comprehensive Spending Review'. The Journal of Finance and Management in Public Services. Vol. 10, Nº01, pp. 14-23.
• Fuchs, Ester R. (2012), Governing the Twenty First Century City. Journal of International Affairs. Vol. 65, Nº 02, pp. 43-56.
• Fuchs, Ester R. (2010). Mayors and money: fiscal policy in New York and Chicago. University of Chicago Press. 376 pp..
• Gordon, Ian and Travers, Tony (2010), London: planning the ungovernable city. City, culture and society. Vol. 01, Nº 02, pp. 49-55.
• GLA -Greater London Authority (2019a), The greater London authority consolidated budget and component budgets for 2019/2020. Available at ˂https://www.london.gov.uk/sites/default/files/mayors-final-budget-2019-20.pdf˃ Accessed on November, 02, 2019.
  » https://www.london.gov.uk/sites/default/files/mayors-final-budget-2019-20.pdf˃
• GLA - Greater London Authority (2019b), Mayoral decision-making in the greater London authority. Available at ˂https://www.london.gov.uk/sites/default/files/mdm_may_2019.pdf˃ Accessed on November, 02, 2019.
  » https://www.london.gov.uk/sites/default/files/mdm_may_2019.pdf˃
• GLA - Greater London Authority (2013), The greater London authority consolidated budget and component budgets for 2013/2014. Available at ˂https://www.london.gov.uk/file/9638/download?token=5Ga5zyUm˃ Accessed on November, 02, 2019.
  » https://www.london.gov.uk/file/9638/download?token=5Ga5zyUm˃
• HARRIS, Tom; HODGE, Louis, and PHILLIPS, David (2019), English local government funding: trends and challenges in 2019 and beyond. IFS Report. Available at ˂https://ifs.org.uk/publications/14563˃ Accessed on December, 09, 2019.
  » https://ifs.org.uk/publications/14563˃
• HM TREASURY (2020) HM Treasury. Spending Review 2010. Available at ˂https://www.gov.uk/government/publications/spending-review-2010˃ Accessed on November, 05, 2019.
  » https://www.gov.uk/government/publications/spending-review-2010˃
• HIGGINS, Paul; CAMPANERA, Josep, and NOBAJAS, Alexandre (2014), Quality of life and spatial inequality in London. European Urban and Regional Studies. Vol. 21, Nº 01, pp. 42-59.
• Institute FOR FISCAL STUDIES (2016), The changing composition of UK tax revenues. IFS Briefing Note BN182. Available at <https://www.ifs.org.uk/uploads/publications/bns/BN_182.pdf> Accessed on December, 05, 2020.
  » https://www.ifs.org.uk/uploads/publications/bns/BN_182.pdf>
• John, Peter and Margetts, Helen (2003), Policy punctuations in the UK: fluctuations and equilibria in Central Government Expenditure Since 1951. Public Administration. Vol. 81, Nº 03, pp. 411-432.
• Jones, Rowan; Lande, Évelyne; Lüder, Klaus and PORTAL, Marine (2013), A comparison of budgeting and accounting reforms in the National Governments of France, Germany, the UK and the US. Financial, Accountability and Management. Vol. 29, Nº 04, pp. 419-441.
• KELLY, Joanne and WANNA, John (2000), New public management and the politics of government budgeting. International Public Management Review. Vol. 01, Nº 01, pp. 33-55.
• Lascoumes, Pierre and Le Galès, Patrick (2005), Gouverner par les instruments. Paris: Le Presses/ Science Po. 370 pp..
• Le Galès, Patrick (2016), Performance measurement as a policy instrument. Policy Studies. Vol. 37, Nº 06, pp. 508-520.
• Lijphart, Arend (1971), Comparative politics and the comparative method. American Political Science Review. Vol. 65, Nº 03, pp. 682-693.
• Lindblom, Charles E. (2009), Muddling through 2: a ubiqüidade da decisão incremental. In: Políticas públicas e desenvolvimento: bases epistemológicas e modelos de análise. Edited by HEIDEMMAN, Francisco G. and SALM, José Francisco. Brasília: Editora da UnB. pp. 181-202.
• LGA - Local Government Association (2020), L Local government funding: Moving the conversation on ˂https://www.local.gov.uk/moving-the-conversation-on/fundingocal˃ Government Association. Available at <https://www.local.gov.uk/> Accessed on November, 05, 2019.
  » https://www.local.gov.uk/moving-the-conversation-on/fundingocal˃» https://www.local.gov.uk/>
• MaYor OF LONDON (2018), Sadiq Khan to kickstart major council homebuilding comeback in London. Mayor of London. May, 16, 2018. Available at ˂https://www.london.gov.uk/press-releases/mayoral/mayor-to-help-councils-start-10000-new-homes˃ Accessed on November, 05, 2019.
  » https://www.london.gov.uk/press-releases/mayoral/mayor-to-help-councils-start-10000-new-homes˃
• MARQUES, Eduardo (2014), Estructura social y segregación en São Paulo: transformaciones en la Década de 2000. Dados. Vol. 57, Nº 03, pp. 675-710.
• Marques, Eduardo (2013), Government, political actors and governance in urban policies in Brazil and São Paulo: concepts for a future research agenda. Brazilian Political Science Review. Vol. 07, Nº 03, pp. 08-35.
• Marques, Eduardo and Hoyler, Telma (2018), Política e coalizões de governo. In: As políticas do urbano em São Paulo. Edited by MARQUES, Eduardo. São Paulo: Editora UNESP. pp. 45-78.
• MENDES, Áquilas and FUNCIA, Francisco Rózsa (2016), O SUS e seu financiamento. In: Sistema de saúde no Brasil: organização e financiamento. Edited by MARQUES, Rosa Maria; PIOLA, Sérgio Francisco, and ROA, Alejandra Carrillo. Rio de Janeiro: ABrES/ Brasília, DF: Ministério da Saúde, Departamento de Economia da Saúde, Investimentos e Desenvolvimento, OPAS/OMS no Brasil. pp. 139-168.
• Paim, Debora Gambetta (2019), A instrumentalização da política urbana no município de São Paulo: uma análise do Fundo de Desenvolvimento Urbano. Master’s dissertation. Escola de Artes, Ciências e Humanidades. Universidade de São Paulo.
• PERES, Ursula Dias (2021), The politics of budget governance. In: The politics of incremental progressivism: governments, governances and urban policy changes in São Paulo. Edited by MARQUES, Eduardo Cesar Leão. São Paulo:Editora Wiley. pp. 91-115.
• PERES, Ursula Dias (2018), Análise da governança do orçamento público. In: As políticas do urbano em São Paulo. Edited by MARQUES, Eduardo. São Paulo: Editora Unesp. pp. 111-140.
• PERES, Ursula Dias and ALLÉ, Camille (2021), Comparando estruturas e processos orçamentários: os casos de Paris e São Paulo. Revista Brasileira de Ciências Sociais. Vol. 36, Nº 107, pp. 01-26.
• Pettinger, Tejvan (2016), Types of Tax in UK. EconomicsHelp. Available at ˂https://www.economicshelp.org/˃ Accessed on November, 02, 2019.
  » https://www.economicshelp.org/˃
• PINTO, José Marcelino de Rezende (2018), O financiamento da educação na Constituição Federal de 1988: 30 anos de mobilização social. Educação & Sociedade. Vol. 39, Nº 145, pp. 846-869.
• Prefeitura de São Paulo (1998/2021), Prestação de Contas Públicas - Balanço Anual. Available at https://www.prefeitura.sp.gov.br/cidade/secretarias/fazenda/contaspublicas/index.php?p=3212/ Accessed on November, 05, 2019.
  » https://www.prefeitura.sp.gov.br/cidade/secretarias/fazenda/contaspublicas/index.php?p=3212/
• REZENDE, Fernando (2015), A Política e a economia da despesa pública: escolhas orçamentárias, ajuste fiscal e gestão pública - elementos para o debate da reforma do processo orçamentário. Rio de Janeiro: Editora FGV. 168 pp..
• SANTOS, Milton(1994), São Paulo, metrópole internacional do terceiro mundo. Revista do Departamento de Geografia. Vol. 07, pp. 07-24.
• Sandford, Mark (2019), Reviewing and reforming local government finance [Briefing Paper]. Number 07538, House of Commons Library, London. Available at ˂https://commonslibrary.parliament.uk/research-briefings/cbp-7538˃ Accessed on October, 25, 2019.
  » https://commonslibrary.parliament.uk/research-briefings/cbp-7538˃
• Schick, Allen (2009), Evolutions in budgetary practice. Paris: OECD. 490 pp..
• Schick, Allen (1976), O PPB e o orçamento incremental. Revista de Administração Pública. Vol. 10, Nº 02, pp. 65-84.
• Schneider, Aaron (2005), Conflito político e instituições orçamentárias: aprofundando a democracia no Brasil. Revista de Sociologia e Política. Vol. 24, pp. 87-103.
• Slack, Enid (2017), How much local fiscal autonomy do cities have? A comparison of eight cities around the world. IMFG Perspectives. Vol. 19, pp. 01-15.
• Slack, Enid and Bird, Richard M. (2014), The political economy of property tax reform. OECD Working Papers on Fiscal Federalism, 18, OECD Publishing. Available at <https://doi.org/10.1787/22265848/> Accessed on September, 20, 2019.
  » https://doi.org/10.1787/22265848/>
• Travers, Tony (2008), The greater London Authority 2000 to 2008. Available at ˂https://www.lse.ac.uk/geography-and-environment/research/lse-london/documents/Archive/HEIF-2/Tony-T.pdf˃. Accessed on September, 17, 2019.
  » https://www.lse.ac.uk/geography-and-environment/research/lse-london/documents/Archive/HEIF-2/Tony-T.pdf
• Travers, Tony (2004), The politics of London: governing an ungovernable city. London: Macmillan International Higher Education. 256 pp..
• Vacchelli, Elena; Kathrecha, Preeti, and GYTE, Natalie (2015), Is it really just the cuts? Neo-liberal tales from the women's voluntary and community sector in London. Feminist Review. Vol. 109, Nº 01, pp. 180-189.
• Wildavsky, Aaron (1975), Budgeting: a comparative theory of budgetary processes. London: Little, Brown [and] Company. 425 pp..
• Wildavsky, Aaron (1969), Rescuing policy analysis from PPBS. Public Administration Review. Vol. 29, Nº 02, pp. 189-202.
• Wildavsky, Aaron and CAIDEN, Naomi (2004), The new politics of the budgetary process. New York: Pearson/Longman. 288 pp..
• Wilson, David and Game, Chris (2011), Local government in the United Kingdom. London: Macmillan International Higher Education. 464 pp..

• 1
  Ken Livingstone was the Leader of the Greater London Council (GLC) and a member of the Labour Party. He was elected the first GLA mayor as an independent candidate ( TRAVERS, 2008 ).
• 2
  See Pinto (2018) and Mendes and Funcia (2016) .
• 3
  Based on an interview with Prof. Mark Kleinman on November, 01, 2019, and also on an interview with a civil servant of the London Legislative Assembly.
• 4
  Additionally, according to budget advisor of the London Legislative Assembly, GLA’s and boroughs’ taxes are calculated after defining the Central Grants, which enables determining the taxing needs to balance budget expenditures. Thus, budget expenditures determine the revenues.
• 5
  In the UK, local districts are required to adopt one of three types of ‘executive arrangements’ to govern and make decisions within their councils. These arrangements are based on the ‘Local Government Act 2000 ’ and were later updated by the ‘Localism Act 2011 ’ . There are two main forms of executive arrangement: ‘leader and cabinet’ and ‘elected mayor and cabinet’ models. In the former, the mayor is elected directly by residents of the borough; in the latter, the majority party or coalition in the council appoints the leader and the cabinet. Since 2012, principal authorities have been allowed to adopt the ‘committee system’ form of governance. Under this model, the power is exercised by the Council – which elects the leader – and by committees made of councilors in proportions of their parties' representation in the Council.
• 6
  Before the GLA, there was a metropolitan authority in London called the GLC (Great London Council, London Government Act 1963). The GLC was abolished by Margaret Thatcher in 1986 ( TRAVERS, 2004) .
• 7
  Mark Sandford, House of Commons Library consultant, interview on November, 18, 2019.
• 8
  It is worth mentioning that the central government imposes a limit on annual council tax and establishes a banding system to be used by local authorities ( CIPFA, 2019) .
• 9
  As already explained, in Brazil, municipalities have to allocate 15% of taxes to health policies and 25% to education. However, in São Paulo, the Local Constitution has determined that this percentage should be increased to 31% of local taxes.
• 10
  Urban Operations (OU) are instruments of public intervention defined by the City Statute (Federal Act 10.257 /2001) and provided for in the Strategic Master Plan – PDE (Municipal Act 16.050/14). These laws establish specific urbanistic rules, such as those stimulating the construction of housing close to public transportation and encouraging the urban redevelopment of a given area in the city.
• For data replication, see: https://doi.org/10.7910/DVN/2CDWA9
• *
  Acknowledgments: The author would like to thank professors Anthony Pereira (Brazil Institute/KCL) and Patrick Le Galès (Sciences Po) for supporting this research and for all the suggestions to improve this paper. Thank are also due to all the colleagues from Center for Metropolitan Studies (CEM/USP) and the BPSR reviewers.

• Funding information: Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp). Grant-2019/09161-3

Edited by

Revised by Karin Blikstad

Publication Dates

• Publication in this collection
  10 June 2022
• Date of issue
  2022

History

• Received
  18 Dec 2020
• Accepted
  9 Aug 2021