The Federal
Shadow
联邦主义
之影
Brazil's 1988 constitution built one of the most fiscally decentralized states in the world — and then left the sub-national units without the capacity to use that decentralization. Twenty-six states and 5,570 municipalities received constitutional autonomy. Most received constitutional dependency. The states are not a background condition. They cast a shadow of their own. 巴西1988年宪法建立了世界上财政最分权的国家之一——然后却让次国家单位缺乏利用这种分权的能力。二十六个州和5,570个市获得了宪法自治权。大多数获得了宪法依赖性。各州不是背景条件。它们投下各自独特的阴影。
Federalism distributes power. It does not distribute capacity. That gap is the shadow. 联邦主义分配权力。它不分配能力。这个差距就是阴影。
Brazil's federal structure casts a shadow over national policy in the specific sense that it creates constraints no government can easily override. It does so in three directions simultaneously: it empowers São Paulo to resist redistribution (the producing state refuses to be net donor without voice); it entrenches the Northeast's political dependency (transfers produce electoral loyalty, not development); and it gives smaller states disproportionate representation in the Senate (two senators per state regardless of population) that allows them to veto policies that threaten their transfer receipts. Every president since 1988 has had to navigate this architecture — and every president has been shadowed by it. 巴西联邦结构在国家政策上投下阴影——它创造了无法轻易覆盖的约束。它同时在三个方向上发挥作用：赋权圣保罗抵制再分配（生产州拒绝成为没有话语权的净捐助方）；巩固东北部的政治依赖性（转移支付产生选举忠诚而非发展）；并给予较小的州在参议院中不成比例的代表权（每州两名参议员，不论人口多少），使它们能够否决威胁其转移收入的政策。1988年以来每位总统都不得不驾驭这一架构——并且每位总统都处于它的阴影之下。
Constitutional Design宪法设计
1988 Constitution gave states and municipalities significant own revenue (ICMS, ISS, IPTU) plus mandatory federal transfers (FPE, FPM, Fundeb). Decentralization was the reaction against 1964–85 fiscal centralism. 1988年宪法给予州和市大量自有收入（ICMS、ISS、IPTU）加上强制联邦转移（FPE、FPM、联教基金）。分权是对1964–85年财政集中主义的反应。
Capacity Asymmetry能力不对称
Spending mandates (health, education) were transferred with constitutional minimums but without matching fiscal capacity. Most states have insufficient own revenue to fund health and education mandates without federal transfers. The shadow: they must receive transfers, but receiving transfers creates political dependency. 支出任务（卫生、教育）与宪法最低标准一同转移，但没有相应的财政能力。大多数州没有足够的自有收入来资助没有联邦转移的卫生和教育任务。阴影在于：它们必须接受转移支付，但接受转移支付会产生政治依赖性。
Political Lock-In政治锁定
States with high transfer dependency have weak fiscal incentive to develop their own tax base (because development would reduce their FPE share). Governors of transfer-dependent states have strong electoral incentive to support whichever federal government guarantees transfer flows. The PT built its Northeast hegemony on this logic. 高度依赖转移支付的州对发展自己的税基缺乏财政激励（因为发展会减少其FPE份额）。依赖转移的州的州长有强烈的选举激励来支持保证转移流动的联邦政府。劳工党在这一逻辑基础上建立了其东北部霸权。
The Constituent Assembly gave the states what they wanted: money and autonomy. It did not give them what they needed: capacity. 制宪议会给了各州它们想要的：钱和自治权。它没有给它们所需要的：能力。
Brazil's 1988 Constituent Assembly was the most genuinely democratic constitutional drafting process in the country's history: 559 members, wide popular participation, and a mandate to reverse the centralization of the military regime. The fiscal architecture it produced was correspondingly redistributive and decentralizing. States and municipalities received a guaranteed share of federal tax revenues (IPI, IR, ITR) through the FPE and FPM. The ICMS (value-added tax on goods and services) was assigned entirely to states. Municipal ISS (services tax) and IPTU (property tax) went to municipalities. 巴西1988年制宪议会是该国历史上最真正民主的宪法起草过程：559名成员、广泛的民众参与，以及推翻军事政权集权化的授权。它产生的财政架构相应地具有再分配性和分权性。州和市通过FPE和FPM获得联邦税收（IPI、IR、ITR）的有保证份额。ICMS（货物和服务增值税）完全分配给各州。市级ISS（服务税）和IPTU（财产税）归市政府。
The problem: spending mandates were also decentralized, but with constitutional minimums (states must spend at least 12% of revenues on health, 25% on education) that in many cases exceed own-source revenues. The result is a federalism of mandates: the federal government sets spending floors while states bear fiscal responsibility. For states like Piauí, Maranhão, and Amazonas, the constitutional mandates cannot be funded without federal transfers. Those states' fiscal autonomy is formal, not real. 问题在于：支出任务也被分权化，但附有宪法最低标准（州必须将至少12%的收入用于卫生、25%用于教育），在许多情况下超过了自有收入。结果是任务联邦主义：联邦政府设定支出底线，而各州承担财政责任。对于皮奥伊州、马拉尼昂州和亚马逊州等州，没有联邦转移就无法资助宪法任务。这些州的财政自治权是形式上的，而非实质上的。
One state. 22% of Brazil's population. 33% of GDP. 36% of federal tax receipts. And a permanent grievance. 一个州。占巴西22%的人口。33%的GDP。36%的联邦税收。以及一个持久的不满。
São Paulo's structural position in Brazilian federalism is unique: it is large enough to be a net fiscal donor under almost any redistribution scheme, and organized enough to resist that position politically. The São Paulo state government has historically been run by the PSDB (1994–2018), which built a political identity partly around resistance to federal fiscal extraction. The slogan "São Paulo não pode parar" (São Paulo cannot stop) reflects a real structural condition: the state's ICMS receipts alone are larger than the entire budgets of many smaller states. 圣保罗在巴西联邦主义中的结构性地位是独特的：在几乎任何再分配方案下它都足够大，是净财政捐助方，且足够有组织来政治上抵制这一地位。圣保罗州政府历来由巴西社会民主党（1994–2018年）管理，其政治身份部分建立在抵制联邦财政提取上。"圣保罗不能停"（São Paulo não pode parar）这一口号反映了一个真实的结构性条件：该州仅ICMS税收就比许多较小州的整个预算还要大。
The FPE formula distributes revenue inversely to economic development: the less developed a state, the higher its FPE share coefficient. São Paulo consequently receives the minimum FPE share despite generating the most federal taxes. This redistribution is constitutionally mandated and politically non-negotiable — any government that tried to reduce Northeast transfers would lose the Senate majority that underpins governance. The shadow falls on São Paulo from the opposite direction: it generates the fiscal surplus that feeds the transfer system, and generates enough political grievance to make fiscal reform perpetually contentious. FPE公式与经济发展程度成反比分配收入：一个州越不发达，其FPE份额系数越高。圣保罗因此尽管产生最多的联邦税收，却获得最低的FPE份额。这种再分配在宪法上是强制性的，政治上是不可谈判的——任何试图减少东北部转移的政府都将失去支撑治理的参议院多数。阴影从相反方向笼罩圣保罗：它产生了养活转移制度的财政盈余，并产生了足够的政治不满，使财政改革永远充满争议。
São Paulo's fiscal size gives it both leverage and a shadow. No government can afford to ignore São Paulo's interests (capital flight, investment decisions, credit market signals). But São Paulo's interests include resisting the very transfer mechanisms that make national governance stable. The result is a permanent tension that is never resolved — it is permanently managed through patronage, federal investment decisions, and Centralão bargaining. 圣保罗的财政规模使其既有杠杆又有阴影。没有任何政府能够忽视圣保罗的利益（资本外逃、投资决策、信贷市场信号）。但圣保罗的利益包括抵制使全国治理稳定的转移机制本身。结果是一种从未解决的永久性张力——它通过庇护关系、联邦投资决策和中间派议价得到永久管理。
Nine states. 57 million people. The PT's electoral anchor. And a fiscal dependency that predates the party by 100 years. 九个州。5700万人。劳工党的选举锚。以及一个比该党早100年出现的财政依赖性。
The Northeast's position in Brazilian political economy is a product of structural underdevelopment that long predates Lula and cannot simply be attributed to PT patronage. The region was the original core of the colonial sugar economy, then progressively marginalized by coffee (São Paulo), industry (Southeast), and agribusiness (Center-West/South). Its per-capita income is consistently around 50% of the national average; its Human Development Index scores cluster in the lowest national quartile; its fiscal capacity is lowest of any macroregion. 东北部在巴西政治经济中的地位是结构性欠发展的产物，这一欠发展远早于卢拉，不能简单归因于劳工党的庇护。该地区是殖民地甘蔗经济的原始核心，随后被咖啡（圣保罗）、工业（东南部）和农业综合体（中西部/南部）逐步边缘化。其人均收入持续在全国平均水平的约50%；其人类发展指数得分聚集在全国最低四分位数；其财政能力是所有宏观区域中最低的。
The PT's Northeast hegemony — Lula won 15%–20% margins in every Northeastern state in 2006, 2022 — is partly explained by the structural resonance of Bolsa Família in cash-transfer-dependent populations, partly by Lula's personal biography (born in Garanhuns, Pernambuco), and partly by the political economy of transfers. Northeastern governors who aligned with the PT received investment, infrastructure (transposição do São Francisco, Programa de Aceleração do Crescimento), and favorable transfer treatment. Those who opposed risked fiscal isolation. 劳工党的东北部霸权——卢拉在2006年、2022年每个东北部州都赢得了15%–20%的优势——部分由家庭津贴在依赖现金转移人口中的结构性共鸣解释，部分由卢拉的个人经历（出生于伯南布哥州加拉尼温斯），部分由转移支付的政治经济学解释。与劳工党结盟的东北部州长获得了投资、基础设施（圣弗朗西斯科河调水、加速增长计划）和有利的转移待遇。反对者则面临财政孤立的风险。
Northeast Profile (9 States)东北部概况（9个州）
São Paulo Profile圣保罗概况
The Northeast collects 52% of FPE on 15% of GDP. São Paulo collects 3% on 33%. The formula is not a technicality — it is the architecture. 东北部以15%的GDP获得52%的FPE。圣保罗以33%的GDP仅获3%。公式不是技术细节——它就是这道阴影的架构。
FPE dependency rate = FPE receipts as % of total state fiscal revenue. Norte states (RR, AP, AC, AM, PA, RO, TO) average 60–90% dependency; Northeast states cluster at 50–70%; São Paulo at ~3%. Data: STN (2022), weighted by state share of regional FPE. FPE依赖率 = FPE收入占州财政总收入百分比。北部州（罗赖马、阿马帕、阿克里、亚马逊、帕拉、朗多尼亚、托坎廷斯）平均依赖率60–90%；东北部各州集中在50–70%；圣保罗约3%。数据来源：国库局（2022年），按州在区域FPE中的份额加权。
The Fundo de Participação dos Estados transfers 21.5% of IPI and IR to states. The formula is a political settlement frozen in 1976. 州参与基金将21.5%的IPI和IR转移给各州。分配公式是1976年冻结的政治协定。
The FPE (Fundo de Participação dos Estados) is the primary mechanism for fiscal equalization in Brazilian federalism. Constitutionally mandated since 1988, it distributes 21.5% of federal IPI (industrialized products tax) and IR (income tax) revenues to the 27 states and the Federal District. The distribution formula weights inversely with economic development: the less developed states receive proportionally more. The five North and nine Northeast states together receive approximately 75% of FPE despite generating well under 30% of national tax receipts. 州参与基金（FPE）是巴西联邦主义中财政均等化的主要机制。自1988年以来由宪法强制，它将21.5%的联邦IPI（工业化产品税）和IR（所得税）收入分配给27个州和联邦区。分配公式与经济发展程度成反比：欠发达州按比例获得更多。北部五州和东北九州合计约获得FPE的75%，尽管它们产生的全国税收远不足30%。
The FPE formula was defined by a 1976 Complementary Law during the military period and was essentially grandfathered into the 1988 Constitution. The STF ruled in 2010 that the formula was unconstitutional in its 1989 form and demanded revisions, which were legislated in 2013 (LC 143). The revised formula updated state GDP weights slightly but preserved the overall redistributive direction. The fundamental political economy has not changed: FPE is the mechanism through which federal governments purchase Senate support from smaller states, and through which Northeast states receive the transfers upon which their fiscal survival depends. FPE公式由军事时期1976年的一项补充法律定义，实质上被继承进了1988年宪法。最高联邦法院在2010年裁定该公式以1989年形式违宪，并要求修订，修订于2013年（LC 143）立法。修订后的公式略微更新了州GDP权重，但保留了整体再分配方向。基本政治经济学没有改变：FPE是联邦政府从较小州购买参议院支持的机制，也是东北部各州获得其财政生存所依赖的转移支付的机制。
5,570 municipalities. The largest 15 generate enough ICMS to be net national contributors. Most of the other 5,555 would dissolve without FPM. 5,570个市。最大的15个产生足够的ICMS成为净全国贡献者。其余5,555个中的大多数没有FPM将会解体。
The FPM (Fundo de Participação dos Municípios) distributes 22.5% of IPI and IR revenues to municipalities, using a formula that prioritizes smaller municipalities. A city of 10,000 receives proportionally far more per capita than São Paulo — the inverse-population weighting means that roughly half of FPM goes to municipalities of fewer than 20,000 inhabitants. Brazil has approximately 1,250 municipalities of fewer than 5,000 people, most of which exist essentially as fiscal entities: their mayors manage FPM transfers, employ the local public sector, and mobilize votes. They produce essentially no own-source revenue. 市参与基金（FPM）将22.5%的IPI和IR收入分配给各市，使用优先考虑较小市的公式。一个1万人的城市按人均比例获得的远比圣保罗多——反人口加权意味着约一半的FPM流向人口不足2万的市政府。巴西约有1,250个人口不足5,000的市，其中大多数本质上作为财政实体存在：其市长管理FPM转移支付，雇用地方公共部门，并调动选票。它们基本上不产生自有收入。
The political economy consequence is systematic: FPM creates thousands of municipal political machines dependent on federal transfer flows. These machines constitute the base of the Centrão's national bargaining power. Congressional deputies who negotiate FPM supplementary appropriations (emendas) are the connective tissue between federal fiscal transfers and local electoral machines. When Temer and Bolsonaro expanded the "emendas de relator" system to R$16B+ annually, they were feeding this exact architecture: federal money, municipally allocated, electorally mobilized. 政治经济学后果是系统性的：FPM创造了依赖联邦转移流动的数千个市政政治机器。这些机器构成了中间派全国议价权力的基础。谈判FPM补充拨款（修正案）的国会众议员是联邦财政转移与地方选举机器之间的结缔组织。当特梅尔和博索纳罗将"报告员修正案"体系扩大到每年超过160亿雷亚尔时，他们正在供养这个确切的架构：联邦资金，市级分配，选举动员。
The most rigorous analyst of Brazilian federalism shows that decentralization improved service delivery efficiency without resolving the fundamental inequality between units. 对巴西联邦主义最严格的分析者表明，分权提高了服务提供效率，但未解决单位间的根本不平等。
Marta Arretche (CEBRAP/USP) is the essential scholar for understanding Brazilian federalism as a structural constraint. Her research program, particularly Estado Federativo e Políticas Sociais (2000) and Trajetória das Desigualdades (2015), establishes the core argument: the mode of Brazilian decentralization after 1988 transferred policy implementation responsibility to sub-national governments without equalizing their capacity to implement policy effectively. The mandate-without-capacity gap is the federalism's structural shadow. 玛尔塔·阿雷切（CEBRAP/USP）是理解作为结构约束的巴西联邦主义的基本学者。她的研究项目，尤其是《联邦国家与社会政策》（2000年）和《不平等的轨迹》（2015年），建立了核心论点：1988年后巴西分权的模式将政策执行责任转移给次国家政府，而没有均等化其有效执行政策的能力。无能力任务差距就是联邦主义的结构性阴影。
Arretche's crucial empirical contribution: she shows that decentralization did reduce regional inequalities in access to services (health, education) — the reforms worked in the sense that the poorest municipalities began providing services they could not provide before. But the inequality in quality of those services increased: São Paulo's municipal health system and Piauí's are both "universal access" systems on paper, but the effective gap in quality is larger in 2026 than in 1988. Decentralization democratized access and preserved inequality of outcome. 阿雷切的关键实证贡献：她表明分权确实减少了获得服务方面的区域不平等（卫生、教育）——改革发挥了作用，使最贫困的市开始提供它们以前无法提供的服务。但这些服务质量方面的不平等增加了：圣保罗市卫生系统和皮奥伊州的在纸面上都是"普遍获取"系统，但质量上的有效差距在2026年比1988年更大。分权使获取民主化，并维持了结果不平等。
Building a governing coalition in Brazil requires managing the fiscal interests of 27 state governments. This is not a side effect of federalism. It is federalism. 在巴西建立执政联盟需要管理27个州政府的财政利益。这不是联邦主义的副作用。这就是联邦主义。
Every Brazilian president faces the same structural problem: they need a Senate majority to pass legislation, and the Senate is designed to over-represent smaller, poorer, more transfer-dependent states. Twenty-seven states have 2 senators each. São Paulo's 46 million voters elect 2 senators; Roraima's 600,000 elect 2 senators. The result: policies that reduce federal transfers, reform pension structures, or rationalize fiscal equalization face a structurally hostile upper chamber. 每位巴西总统面临同样的结构性问题：他们需要参议院多数来通过立法，而参议院的设计使较小、较穷、更依赖转移的州获得更大代表权。二十七个州各有2名参议员。圣保罗4600万选民选出2名参议员；罗赖马州60万人选出2名参议员。结果：减少联邦转移、改革养老金结构或合理化财政均等化的政策面临结构上不利的上议院。
The shadow mechanism: federal governments must buy Senate cooperation through targeted investment (obras), discretionary transfer supplements (emendas), and investment bank credit lines (BNDES operations in politically key states). This creates a secondary fiscal architecture of political pricing — the market in which Senate votes are converted into federal public goods for specific states. Cardoso, Lula, Dilma, Temer, Bolsonaro, and Lula III all operated within this market. None could escape it, because escaping it would mean electoral exposure in the Senate, which would mean legislative paralysis. 阴影机制：联邦政府必须通过定向投资（工程）、可自由裁量的转移补充（修正案）和开发银行信贷额度（特定政治关键州的BNDES操作）购买参议院合作。这创造了一个政治定价的次级财政架构——将参议院选票转化为特定州联邦公共品的市场。卡多佐、卢拉、迪尔玛、特梅尔、博索纳罗和卢拉三期都在这个市场内运作。没有人能逃离它，因为逃离它意味着在参议院的选举风险，这将意味着立法瘫痪。
Brazil has spent 35 years transferring resources to the Northeast. The Northeast remains at 50% of national per-capita income. The paradox has a structural explanation. 巴西花了35年向东北部转移资源。东北部仍处于全国人均收入的50%。这一悖论有结构性解释。
Transfers have raised living standards in the Northeast — dramatically so. Infant mortality, school enrollment, and poverty rates have all improved substantially since 1988. But the relative position of the Northeast within Brazil has not improved equivalently, for structural reasons: transfers finance consumption and public sector employment, not productive investment. They do not address infrastructure deficits, industrial base absence, or the fiscal trap that discourages own-revenue generation. 转移支付提高了东北部的生活水平——显著地如此。自1988年以来，婴儿死亡率、学校入学率和贫困率都有了实质性改善。但东北部在巴西内部的相对地位并没有同等改善，原因是结构性的：转移支付资助消费和公共部门就业，而非生产性投资。它们没有解决基础设施赤字、缺乏工业基础或抑制自有收入产生的财政陷阱。
The fiscal trap operates as follows: a highly transfer-dependent state that successfully develops its tax base (attracts industry, grows formal employment, increases own ICMS receipts) will see its FPE coefficient reduced in the next formula review. The incentive structure rewards stagnation and punishes development. This is not a bug; it is a feature of a formula designed by states that had strong interests in not being graduated out of the transfer-receipt category. SUDENE (Northeast Development Superintendency), refounded in 2007, has produced modest investment but has not resolved the structural fiscal trap. 财政陷阱的运作方式如下：一个成功发展税基（吸引工业、增加正规就业、增加自有ICMS收入）的高度依赖转移的州将在下一次公式审查中看到其FPE系数降低。激励结构奖励停滞而惩罚发展。这不是漏洞；这是公式的特性，该公式由对不被列入非转移受体类别有强烈兴趣的州设计。2007年重建的东北部发展监督局（SUDENE）产生了适度的投资，但没有解决结构性财政陷阱。
Federalism distributes power without distributing capacity. That architecture casts a shadow — over the poor states it sustains, the rich states it taxes, and any federal government that tries to govern them all. 联邦主义分配权力而不分配能力。这一架构投下阴影——笼罩着它维持的贫困州、它征税的富裕州，以及任何试图治理所有这些州的联邦政府。
Brazil's fiscal federalism has three structural consequences: First, it creates an over-represented Senate that blocks fiscal rationalization (smaller states' senators protect transfer flows). Second, it creates a political market in which every federal government purchases cooperation through discretionary public investment — the Centrão's emendas are written on this architecture. Third, it creates a fiscal trap in transfer-dependent states that rewards dependency over development. Marta Arretche showed that decentralization democratized service access. It did not democratize state capacity. The federalism shadow is the gap between those two things. 巴西财政联邦主义有三个结构性后果：第一，它创造了一个阻止财政合理化的过度代表参议院（较小州的参议员保护转移流）。第二，它创造了一个政治市场，每届联邦政府通过可自由裁量的公共投资购买合作——中间派的修正案写在这一架构之上。第三，它在依赖转移的州中创造了一个奖励依赖胜于发展的财政陷阱。玛尔塔·阿雷切表明分权使服务获取民主化了。它没有使国家能力民主化。联邦主义阴影就是这两件事之间的差距。
No government has escaped the shadow because no government can afford to. Reducing FPE would cost Senate votes. Raising the fiscal capacity standard for smaller municipalities would cost House votes. Reforming the transfer-dependency incentive structure would cost the Northeast. And losing the Northeast means losing the election. 没有任何政府能够逃离这个阴影，因为没有任何政府能够承担这样做的代价。减少FPE将损失参议院选票。提高较小市的财政能力标准将损失众议院选票。改革依赖转移的激励结构将失去东北部。而失去东北部意味着输掉选举。