PE-I-02: Peruvian Decentralisation — Regional Governments and the Territorial State (2002–2026)

Status: DRAFTWords: 9,682

1. Key Takeaways

  • The 2002 decentralisation was the post-Fujimori transition's largest structural reform and its most ambivalent legacy. Ley 27680 (March 2002) rewrote the constitution's territorial chapter; Ley 27783 (July 2002, the Ley de Bases de la Descentralización), Ley 27867 (November 2002, the Ley Orgánica de Gobiernos Regionales, LOGR), and Ley 27972 (May 2003, the Ley Orgánica de Municipalidades, LOM) built the operating framework; and the November 2002 elections seated twenty-five elected regional governments — one per department, plus the constitutional province of Callao, with Metropolitan Lima under a special regime — in January 2003. The reform's political logic was double: to dismantle the Fujimori recentralisation (which had abolished García's incipient regions in 1992 and governed the territory through appointed Consejos Transitorios de Administración Regional, CTARs), and to answer the provincial mobilisation — above all the June 2002 arequipazo against the EGASA/EGESUR privatisations — that had shown the Toledo government how combustible the periphery remained.

  • The design contained its own defeat: departments were converted into "regions" without consolidation. The framers knew that Peru's twenty-four departments were too small, too poor, and too administratively hollow to function as meso-level governments, and the framework therefore treated departmental regionalisation as transitional — Ley 28274 (2004) offered fiscal incentives for departments to merge into macro-regions, with the mergers to be approved by referendum. The 30 October 2005 referendum, covering proposals to fuse sixteen departments into five macro-regions, failed comprehensively: [TBD-VERIFY: the NO vote prevailed in all five proposed macro-regions, with Arequipa commonly cited as the only department where YES won — confirm department-level results against ONPE records]. No merger referendum has been attempted since. Peru has consequently spent a quarter-century governed by twenty-five meso-governments drawn on nineteenth-century departmental lines that nobody designed for the purpose and nobody has been able to redraw — the stillborn macro-regions are the original sin from which most of the record below descends.

  • The fiscal architecture decentralised windfalls, not taxing power. Regional governments were given expenditure responsibilities (health, education, agriculture, transport functions transferred in waves through the 2000s) but essentially no own-source revenue: their budgets are overwhelmingly central transfers — recursos ordinarios negotiated annually with the Ministerio de Economía y Finanzas (MEF), plus the resource-revenue shares (canon minero, canon gasífero, regalías, FOCAM) and, for municipalities, FONCOMUN. The intended second stage of fiscal decentralisation (Decreto Legislativo 955 of 2004) conditioned real tax-sharing on the formation of macro-regions; when the 2005 referendum failed, fiscal decentralisation died with it. The result is a meso-government tier that spends without taxing — administratively dependent on Lima, fiscally unaccountable to its own electorate, and exposed to the full volatility of mineral prices.

  • The canon concentrated the boom in producing regions and converted geological accident into fiscal inequality. Under the canon framework (Ley 27506 of 2001; distribution rules and record in PE-G-02 §13), half of mining income tax flows to the producing region's subnational governments. The 2004–2014 supercycle consequently delivered extraordinary windfalls to Áncash, Cusco, Moquegua, Tacna, Arequipa, and Cajamarca [TBD-VERIFY: aggregate canon minero transfers peaked at roughly S/ 5 billion in the 2007 and 2012 fiscal years, with Áncash alone exceeding S/ 1 billion in 2007 — confirm against MEF Transparencia Económica series] while non-producing departments — Huancavelica adjacent to mining wealth but outside the formulas' main flows, the Amazonian departments, the northern sierra — received fractions of the per-capita amounts. The canon simultaneously generated the spending-capacity paradox: subnational governments chronically failed to execute their investment budgets [TBD-VERIFY: regional investment execution rates persistently in the 60–70 per cent range through the boom years], accumulating unspent balances in the billions of soles while their populations lacked water, paved roads, and staffed clinics.

  • The governance record is dominated by the corruption harvest. The roll of imprisoned, prosecuted, or fugitive regional presidents — César Álvarez of Áncash (the La Centralita network, linked to the 2014 assassination of regional councillor Ezequiel Nolasco), Félix Moreno of Callao (Odebrecht's Costa Verde works), Jorge Acurio of Cusco (Odebrecht's Vía de Evitamiento), Gerardo Viñas of Tumbes (fled to Ecuador), Gregorio Santos of Cajamarca (held in pre-trial detention while in office, ran for president from prison in 2016), Waldo Ríos of Áncash (elected in 2014 while promising S/ 500 to every citizen, later convicted) — produced by mid-decade the narrative of the presidentes regionales as a criminal class: [TBD-VERIFY: the Procuraduría Anticorrupción reported in 2014 that 22 of 25 sitting regional presidents were under investigation; later Contraloría and Procuraduría counts covering ex-governors investigated since 2006 reach the large majority of all who have held the office — confirm counts and convictions]. The narrative is substantially earned and partially unfair: the audit architecture that should have caught corruption early was structurally weak (ex-post control, órganos de control institucional funded by the entities they audited until the 2017 reform), prosecutions concentrated where canon money concentrated, and a minority of competently governed regions (Moquegua under Martín Vizcarra and his successors is the standard exhibit, PE-H-PRES-08) disappears in the aggregate indictment.

  • Decentralisation transferred functions faster than capacity, and the service-delivery record shows it. Health and education — the two largest transferred functions — exhibit the persistent execution gaps, the regional hospital projects abandoned mid-construction (Cusco's Hospital Antonio Lorena, paralysed for years after the Odebrecht-linked contractor's exit, is the genre's exemplar), and the white-elephant genre of canon-funded stadiums, monumental plazas, and oversized municipal palaces in districts without sewerage. The COVID-19 pandemic was the system's terminal stress test: the regional health directorates (DIRESAs) collapsed under the first wave, coordination between MINSA and the regions failed on oxygen, ICU beds, and data, and Peru ended with the world's highest recorded per-capita COVID mortality — a catastrophe with many causes (PE-G-03's informality among them) in which the broken intergovernmental health architecture was a major contributor. The capacity diagnosis is structural: SERVIR (created 2008) and the Ley del Servicio Civil (2013) never penetrated the regional tier at scale, and each change of governor rotates the technical staff, resetting institutional memory to zero every four years.

  • Politically, the regional tier became the territorial expression of the party-system collapse rather than a school for national politics. Regional movements — one-region electoral vehicles, frequently built around a single caudillo and dissolved or renamed between cycles — displaced national parties from subnational power: by the 2014 and 2022 regional elections, movements held the large majority of governorships and national parties had nearly vanished from the territory (the national-party hollowing is analysed in PE-O-01 §2). The governorship has been a poor presidential springboard — César Acuña's repeated attempts failed; Gregorio Santos's prison candidacy was marginal; the exception, Vizcarra, reached the presidency through the vice-presidency, not through a regional machine — so the regional tier accumulates neither party-building capacity nor national leadership. In the mining conflicts, governors have oscillated between mediator (Vizcarra at Quellaveco) and insurgent leader (Santos against Conga), revealing the regional government's structural position: caught between a Lima that holds the permits and the money and communities that hold the roads (the conflict architecture is PE-G-02's subject).

  • Since 2016 the centre has been quietly taking back what 2002 gave, through bypass rather than repeal. The Autoridad para la Reconstrucción con Cambios (ARCC, created by Ley 30556 after the 2017 El Niño costero) centralised reconstruction spending over the heads of regions and municipalities; its mixed record led not to devolution but to deeper centralisation — the government-to-government contracting model (UK G2G for reconstruction, the Lima 2019 Pan American Games "Legado" precedent) and, in 2023, the creation of the Autoridad Nacional de Infraestructura (ANIN) as a permanent central delivery agency [TBD-VERIFY: ANIN's statutory basis and the formal status of ARCC's absorption]. The Plan Nacional de Infraestructura and the proliferating proyectos especiales de inversión pública run major works through ministerial special units. No government has proposed abolishing regional governments; every recent government has built machinery to route around them — recentralisation by bypass, without the political cost of constitutional reform.

  • The 2022–2023 crisis read the decentralisation's failure back to Lima as an indictment. The protest geography after Castillo's removal — Puno, Cusco, Apurímac, Ayacucho, the southern corridor that absorbed nearly all of the 49+ deaths (PE-D-03) — maps precisely onto the regions where two decades of decentralisation delivered offices and canon flows but not state presence, and where trust in every Lima institution runs at single digits. The Lima–provinces settlement question that this raises for the 2030s is treated as a megatrend in PE-O-01 §6; this document supplies its institutional history. The bicameralism restoration (Ley 31988 and Ley 32101, PE-A-04) carries a territorial-representation promise — senators with regional constituencies giving the departments a voice in national lawmaking for the first time since 1992 [TBD-VERIFY: the final allocation between the national-district and regional-district senate seats under Ley 32101] — but a senate is representation, not capacity, and none of the 2026-era reform debates has yet produced a second-generation decentralisation design.

  • The comparative verdict — Peru as the Latin American decentralisation wave's cautionary variant — supports two honest readings. The dominant reading, "decentralisation without capacity," holds that Peru transferred functions and windfalls to a meso-tier that lacked parties, bureaucracies, and audit institutions, and harvested corruption, under-execution, and conflict. The revisionist reading holds that Lima never actually decentralised the things that matter — taxing power, permitting authority over mines, the MEF's budget veto — so the regions were set up to fail and then blamed for failing; on this reading (associated with Vergara's La danza hostil and the resource-federalism literature on the canon as natural experiment), the disappointment is evidence about the design and the centre, not about the provinces. Both readings agree on the diagnosis's gravity and disagree on the remedy — recentralisation versus deepening — which is precisely the fork the 2030s must choose among (deepening, recentralisation by stealth continued, or asymmetric reform that treats Moquegua and Huancavelica differently).

2. The 2002 Decentralisation Bargain

2.1 The Antecedents: García's Regions and Fujimori's Recentralisation

Peru entered the twenty-first century as one of Latin America's most centralised states, but not for want of trying. The 1979 Constitution mandated regionalisation; the first García government finally executed it at the decade's end, creating twelve regions (1989–1990) through departmental groupings, with indirectly constituted regional assemblies. The experiment was brief, chaotic, and politically tainted by its association with the late-García hyperinflationary collapse. Fujimori's 5 April 1992 autogolpe swept it away: the regional governments were dissolved and replaced by Consejos Transitorios de Administración Regional (CTARs) — appointed bodies, one per department, answering to the Ministerio de la Presidencia, which became the regime's territorial patronage machine. For the rest of the decade, "transitory" administration was the permanent condition: the 1993 Constitution retained a decentralisation chapter on paper while the regime governed the territory directly, distributing works through FONCODES and the presidential ministry in a pattern that electoral studies of the period associate with the cultivation of Fujimori's provincial vote.

The 1990s therefore bequeathed two contradictory legacies. Institutionally, the territory was a blank slate: no elected meso-government, no regional bureaucracies beyond the CTAR shells, no subnational fiscal system beyond municipal minima. Politically, centralism was now indelibly associated with the fujimorato — so the transition's democratisers treated decentralisation as a component of redemocratisation itself, not merely an administrative preference. Valentín Paniagua's transitional government (2000–2001) commissioned the groundwork; every serious candidate in 2001 ran on decentralisation; and Alejandro Toledo (PE-H-PRES-03), the cholo president from Cabana who had campaigned against the centralist regime, made it his signature structural reform.

2.2 The Arequipa Demonstration and the Political Logic

The reform's urgency was supplied by the street. In June 2002, Toledo's attempt to privatise the Arequipa electricity generators EGASA and EGESUR — in apparent breach of a campaign promise — produced the arequipazo: days of riots, a state of emergency, the resignation of the interior and prime ministers in the ensuing reshuffle, and the government's capitulation on the sale. The episode taught the Toledo government two lessons that shaped the decentralisation's design and its timing. First, the provinces' tolerance for Lima-decided economic policy was exhausted; a constitutional channel for regional voice was cheaper than recurrent insurrection. Second, the government's own weakness — Toledo's approval was already collapsing — meant the reform had to be delivered fast, before the political capital ran out. The decentralisation was thus simultaneously a democratic conviction, an appeasement of provincial demands, and a pre-emptive concession by a weak government to a mobilised periphery. The combination explains both the reform's speed and its shortcuts.

The framework was assembled in eighteen months. Ley 27680 (6 March 2002) replaced the constitution's Chapter XIV of Title IV, defining decentralisation as "a permanent policy of the State, of obligatory character," establishing regional governments with political, economic, and administrative autonomy in matters of their competence, and creating directly elected regional presidencies and councils. Ley 27783, the Ley de Bases de la Descentralización (LBD, 17 July 2002), set the architecture: the principles (subsidiarity, gradualism, fiscal neutrality), the three-tier structure (national, regional, local), the accreditation system through which functions would be transferred to regions as they demonstrated capacity, and the conduction of the process by a Consejo Nacional de Descentralización (CND). Ley 27867, the Ley Orgánica de Gobiernos Regionales (LOGR, 16 November 2002, amended within weeks by Ley 27902 to strengthen the regional councils and add civil-society participation), defined the regional governments' organs — president (from 2015, governor), vice-president, regional council, and the participatory Consejo de Coordinación Regional — and enumerated competences across agriculture, health, education, energy and mines (promotional functions only), transport, and a dozen other sectors. Ley 27972, the Ley Orgánica de Municipalidades (LOM, 27 May 2003), rewrote the municipal framework for the 196 provincial and over 1,600 district municipalities (the municipal universe has since grown beyond 1,800 districts), which had been the only elected subnational tier to survive the 1990s.

Elections for the new regional governments were held on 17 November 2002 simultaneously with municipal elections; the first regional presidents took office on 1 January 2003. The results were an early warning the political class chose not to read: APRA won twelve regional presidencies — the best subnational result any national party would ever record under the new system — while Toledo's Perú Posible, the reform's author, won one, and independent regional lists took the rest. The decentralisation's first electoral act already displayed the pattern that Section 5 examines: the regional arena would not be an extension of national party competition but a separate ecosystem, and the national parties' 2002 strength was a residue that would evaporate within two cycles.

2.4 The Design Flaw: Departments as "Regions" Without Consolidation

The framers understood that the departmental map was unfit for purpose. Peru's departments — boundaries substantially inherited from colonial intendencias and consolidated in the nineteenth century — range from Lima's ten million people to Madre de Dios's barely 150,000; most lacked any administrative apparatus beyond the CTAR shell; and the LBD itself defined a "region" as requiring economic integration and scale that single departments mostly lacked. But creating genuine macro-regions before electing governments would have delayed the reform for years and ignited boundary disputes; the arequipazo logic demanded speed. The compromise was to regionalise on departmental boundaries transitionally: each department (plus Callao) became a "regional government," with the constitution and the LBD providing that departments would later merge into true regions via referendum, sweetened by Ley 28274 (2004), the Ley de Incentivos para la Integración y Conformación de Regiones, which promised merged regions enhanced fiscal benefits and a greater share of nationally collected taxes.

The 30 October 2005 referendum was the mechanism's first and only serious test. Five macro-regions were proposed, covering sixteen departments — among them a Cusco–Apurímac pairing, a northern grouping around Lambayeque and Piura, and an Ayacucho–Huancavelica–Ica combination [TBD-VERIFY: the precise composition of the five proposed macro-regions]. The proposals were defeated comprehensively: [TBD-VERIFY: NO prevailed in every proposed macro-region, with the YES vote winning in Arequipa alone among the participating departments — confirm department-level results against ONPE records]. The causes were structural, not accidental: departmental identities are old and real; every departmental capital feared becoming a merged region's periphery (would Abancay accept government from Cusco?); the canon-rich departments feared sharing their windfalls with poor neighbours; the campaign was rushed and the incentives abstract; and the national government, by 2005 deeply unpopular, was the proposal's least persuasive advocate. No subsequent government has re-attempted consolidation. The CND itself was abolished in 2007 and folded into the Presidencia del Consejo de Ministros as a secretariat — a demotion that signalled the process's loss of political sponsorship.

The consequence is the system Peru still operates in 2026: twenty-five "regional" governments on departmental lines, a tier that is too small to plan territorially, too large to be local, and too entrenched to redraw. Every pathology documented below — the fiscal inequality between canon neighbours, the capacity deficit of micro-departments, the one-region political vehicles — runs through this unconsolidated map. The stillborn macro-regions are the decentralisation's founding counterfactual: the reform that was designed as a first stage became, by referendum, the permanent settlement.

3. The Fiscal Architecture

3.1 Spending Without Taxing: The Transfer-Dependence Design

Peruvian fiscal decentralisation decentralised expenditure and retained revenue. Regional governments have essentially no taxing power: the constitution and the LBD assign them fees and contributions for services they directly provide, which in practice generate a negligible share of their budgets. Municipalities fare slightly better — the impuesto predial (property tax), vehicle and alcabala (transfer) taxes — but collection is weak outside metropolitan Lima and a handful of cities, and the property cadastres on which the predial depends are incomplete or decades old in most of the country. The overwhelming bulk of subnational finance is therefore transfers, in four main channels: recursos ordinarios, the discretionary central allocations negotiated annually through the MEF's budget process; the resource-revenue shares (canon in its several varieties, regalías mineras, FOCAM); FONCOMUN, the municipal compensation fund financed principally by the two-percentage-point Impuesto de Promoción Municipal embedded in the IGV, distributed by formula with an equalising intent; and earmarked programme transfers from sector ministries.

The design was supposed to be transitional. Decreto Legislativo 955 (February 2004), the fiscal-decentralisation decree, established a two-stage scheme: in the first stage, regions would live on transfers; in the second, merged regions would receive an assigned share of nationally collected taxes (50 per cent of the IGV net of the municipal point, the ISC, and personal income tax collected in their territory was the framework's headline figure [TBD-VERIFY: the precise tax-assignment percentages in DL 955 and its regulations]) plus incentive top-ups. The second stage was conditioned on the macro-regions that the 2005 referendum killed. Fiscal decentralisation therefore never reached stage two — a fact insufficiently appreciated in the standard indictment of regional performance: the regions' fiscal infantilisation was not an oversight but an unexecuted half of the original design. The MEF, for its part, has never shown enthusiasm for executing it. The ministry's control of the annual budget, of investment-project approval through the SNIP (replaced in 2017 by Invierte.pe), and of the public-debt window gives Lima a permanent veto over regional priorities, exercised technocratically and largely invisibly. Successive ANGR (Asamblea Nacional de Gobiernos Regionales, the governors' coordination body created in 2007) platforms have demanded the completion of fiscal decentralisation; no MEF team under any of the eleven presidents of the period has proposed it.

3.2 The Canon System: Windfalls by Geological Accident

Onto this transfer-dependent base, the commodity supercycle poured the canon. The framework (Ley 27506 of 2001, with the distribution rules, subsequent modifications, and the full mining-political-economy context covered in PE-G-02 §13) channels 50 per cent of the income tax paid by mining operators to the subnational governments of the producing department, distributed by formula among the producing district, the producing province, the department's municipalities, the regional government, and the public universities. Parallel instruments extend the logic to other extractives: the canon gasífero (which made Cusco, host of the Camisea fields, one of the richest subnational treasuries in the country), the canon y sobrecanon petrolero for the producing Amazonian and northern departments, FOCAM for the departments along the Camisea pipeline, and the regalías mineras established by the 2004 royalty law and restructured in the 2011 Humala-era fiscal package (PE-G-01).

Three features of the system define its political economy. First, concentration: the canon flows to where the minerals are, so the boom's fiscal dividend concentrated in a half-dozen departments — Áncash (Antamina), Cusco (Camisea, later Las Bambas's neighbours), Moquegua and Tacna (Southern Peru Copper, Quellaveco), Arequipa (Cerro Verde), Cajamarca (Yanacocha) — while departments without producing operations received little or nothing through this channel. Per-capita transfer disparities of an order of magnitude opened between adjacent departments [TBD-VERIFY: per-capita canon comparisons, e.g., Moquegua versus Huancavelica, from MEF Transparencia Económica data]. The inequality is doubly perverse: it rewards geological accident rather than need or performance, and it is procyclical, delivering the most money in the years when construction costs and political temptation peak. Second, magnitude: during the 2004–2014 supercycle the canon minero alone grew from under S/ 500 million at the cycle's start to peaks measured in the billions of soles [TBD-VERIFY: commonly cited peaks of roughly S/ 5 billion in 2007 and again in 2012; Áncash's 2007 allocation alone exceeding S/ 1 billion — confirm against MEF series], with the post-2021 copper-price recovery producing a second wave of historically large transfers in 2022–2023 [TBD-VERIFY: 2022–2023 canon totals]. Third, earmarking: canon funds are legally restricted to investment projects (with later flexibilisations for maintenance and, episodically, emergency uses), a rule intended to prevent their consumption in current spending and which in practice channelled them into the project pipeline that the next subsection shows the subnational tier could not manage.

3.3 The Spending-Capacity Paradox: The Unspent-Balances Chronicle

The defining fiscal datum of the decentralisation's first decade was not scarcity but unspendable abundance. Year after year through the boom, regional and municipal governments closed the fiscal year with large fractions of their investment budgets unexecuted — execution rates persistently in the 60–70 per cent range for the regional tier, with canon-rich municipalities often lower [TBD-VERIFY: execution-rate series from MEF Consulta Amigable; the precise figures vary by year and by whether the denominator is the opening budget (PIA) or the modified budget (PIM)] — while unspent balances accumulated in subnational accounts. The annual end-of-year ritual became a national genre: the MEF publishing execution league tables, the press reporting billions of soles "returned" or idle while the owning districts lacked basic services, and ministers lamenting subnational capacity.

The paradox's causes are well diagnosed. The national investment system (SNIP, then Invierte.pe) imposed feasibility-study requirements that small municipalities could not staff; the procurement framework (and the fear of Contraloría findings, which can end careers over procedural faults) made officials slow; the project pipeline was thin because formulating good projects requires precisely the technical cadres the subnational tier lacks; and the four-year political cycle, combined (after the 2015 re-election ban, Section 5) with guaranteed turnover, rewarded fast, visible, low-complexity works — plazas, stadiums, municipal palaces — over slow, invisible, high-value ones — water systems, cadastres, maintenance. The paradox also has a defence rarely heard in Lima: spending money well is genuinely hard, the rules are written by the centre, and the comparison that matters is not regional execution versus an ideal but regional execution versus the national ministries' own record, which in several sectors and years has been no better [TBD-VERIFY: comparative execution rates, national versus regional tier, from MEF data]. Both things are true: the subnational capacity deficit is real, and the system the centre built made it worse.

3.4 FONCOMUN, Equalisation, and the Budget-Negotiation Politics

Against the canon's concentration, the system's equalising instruments are weak. FONCOMUN distributes by a formula weighting population and unsatisfied basic needs, but its volume is modest relative to the canon flows in producing areas; the canon itself contains no horizontal-equalisation component; and proposals to pool or redistribute canon revenues across departments have been politically untouchable, defended to the death by the producing regions' congressional delegations and governors (the reform debate is catalogued in PE-G-02 §13). What remains is the annual budget negotiation: regions and municipalities petition the MEF and the Congress for recursos ordinarios, project financing, and the mid-year supplementary credits, in a process that rewards lobbying presence in Lima, congressional brokerage, and political alignment rather than formula or need. The gobernadores' annual pilgrimage to the MEF, the ANGR's negotiating platforms, and the congressional budget committee's December amendments — frequently inserting hundreds of small district-level works traded for votes — constitute the system's real fiscal constitution: discretionary, relational, and centralised. Peru's decentralisation, in fiscal terms, is an annual negotiation between a sovereign treasury and twenty-five supplicants holding windfall accounts they struggle to spend.

4. The Governance Record

4.1 The Corruption Harvest: The Governors-in-Prison Roll

The decentralisation's darkest ledger entry is the criminalisation of its leading offices. The emblematic cases span the map and the period. Áncash, the canon minero's richest department, produced the system's defining scandal: regional president César Álvarez (2007–2014) built the La Centralita apparatus — a parallel operations centre in Chimbote running procurement kickbacks, press manipulation, judicial capture, and, prosecutors charged, violence against opponents — and was arrested in May 2014 after the assassination of regional councillor and whistleblower Ezequiel Nolasco; he was subsequently convicted in multiple proceedings, including [TBD-VERIFY: the 2017 corruption conviction over canon-funded procurement and the later sentence in the Nolasco case — confirm charges, dates, and sentence lengths against judicial records]. His successor-but-one Waldo Ríos, elected in 2014 on a promise to give every Ancashino S/ 500, was himself later convicted and disqualified [TBD-VERIFY: conviction details]. Callao's Félix Moreno was convicted in the Odebrecht-linked Costa Verde Callao case after a period as a fugitive. Cusco's Jorge Acurio fell in the Odebrecht Vía de Evitamiento case. Tumbes' Gerardo Viñas fled to Ecuador while in office and was captured there. Cajamarca's Gregorio Santos — the anti-Conga rondero governor (Section 5.3) — spent 2014–2016 in pre-trial detention on corruption charges while still the elected regional president, ran for the presidency from prison in 2016, and was later [TBD-VERIFY: acquitted at first instance in 2017, with the case continuing through retrial — confirm final disposition]. Pasco, Ucayali, Tacna, Cerro de Pasco's municipalities, and others contributed further entries.

By mid-decade the aggregate had hardened into a narrative: the presidentes regionales as a criminal class. The commonly cited statistics supported it: [TBD-VERIFY: the Procuraduría Anticorrupción's 2014 statement that 22 of 25 sitting regional presidents were under investigation; later counts — Contraloría and press tallies through 2018–2022 — finding the large majority of all ex-governors since 2003 under investigation, prosecution, or conviction; confirm the specific counts, which vary by source and by whether "investigation" includes preliminary inquiries]. The narrative did real political work: it justified the 2015 reform package (Ley 30305, banning the immediate re-election of governors and mayors and renaming regional presidents "governors" — a change intended to deflate the office's presidential pretensions) and it fed the recentralisation currents of Section 6.

The narrative's fairness requires careful handling. Three qualifications matter. First, selection on money: investigations concentrated where the canon concentrated; the correlation between windfall scale and prosecution is strong, which implicates the fiscal architecture as much as the office-holders' character. An influential strand of the literature (Arellano-Yanguas on the canon; Muñoz and Dargent on subnational politics) argues the system was criminogenic by design — huge, earmarked, procyclical investment budgets dropped onto governments without bureaucracies, auditors, or parties to discipline them. Second, the threshold problem: "under investigation" in the Peruvian prosecutorial system is a low bar that has also swept in the competent and the eventually acquitted; conviction counts are far smaller than investigation counts. Third, the invisible numerator: the minority of well-governed regions — Moquegua's sequence from Vizcarra (2011–2014, PE-H-PRES-08) onward is the standard exhibit, with San Martín's reformist period under César Villanueva frequently cited alongside it — demonstrates that the same architecture could be made to work, and disappears entirely in the criminal-class framing. The honest summary: the corruption harvest was real, large, and concentrated; it was also the predictable product of an architecture that paired windfalls with institutional vacuum, and the moral indictment of the provinces that Lima built from it conveniently omitted the centre's authorship of the architecture — and the centre's own simultaneous corruption at presidential scale (PE-J-01, where four presidents and the Lava Jato record reside; PE-I-01 for the prosecutorial architecture).

4.2 The Audit Architecture's Weakness

The control system failed structurally before it failed in any particular case. Through the boom years, the Contraloría General de la República operated an overwhelmingly ex-post model: audits arrived years after the works were contracted, by which time the money was gone and the governor frequently re-elected or departed. The órganos de control institucional (OCIs) — the internal audit offices embedded in each regional government and municipality — were, until the 2017 reform [TBD-VERIFY: Ley 30742 of 2018 strengthening the Contraloría and the transfer of OCI personnel to the Contraloría's payroll — confirm statute and dates], staffed and paid by the very entities they audited, an arrangement whose results were as expected. Coverage was thin: hundreds of municipalities went years without any audit. And the system's incentives were perverse in both directions — too weak to deter the corrupt, yet procedurally fearsome enough (personal liability for administrative faults) to paralyse the honest, contributing to the under-execution of Section 3.3. The post-2017 reforms expanded concurrent control (real-time accompaniment of major works, applied to the reconstruction and the Pan American Games) with genuinely better results in the covered projects, but concurrent control covers a small fraction of subnational investment, and the fundamental asymmetry — twenty-five regions, 1,800-plus municipalities, one under-resourced national auditor — persists.

4.3 The Service-Delivery Record: Execution Gaps, White Elephants, and the COVID Collapse

Health and education, the two largest transferred functions, are the record's substantive test. The transfers of the 2000s moved the regional health directorates (DIRESAs) and education directorates (DREs), thousands of facilities, and hundreds of thousands of personnel onto regional payrolls — without moving the rectoral functions (policy, standards, the largest programmes) from MINSA and MINEDU, and without a capacity-building programme remotely scaled to the task. The results: chronic execution gaps in health and education investment; staffing distributions that left highland and Amazonian facilities unfilled while payrolls grew; and an intergovernmental tangle in which responsibility for any given failure could always be assigned to the other tier. The second García government's "municipalisation of education" pilot (2007–2011), which attempted to push school management further down to district municipalities, collapsed and was cancelled by the Humala government — the clearest single demonstration that transferring functions without capacity transfers failure downward. The infrastructure record contributed its own genre: the canon-funded white elephants — Cusco's Hospital Antonio Lorena, left paralysed for years mid-construction after the Odebrecht-linked contract's collapse; monumental stadiums and plazas in under-serviced districts; the oversized municipal palaces of the mining districts — which became the standing visual indictment of the system.

COVID-19 was the terminal stress test, and the intergovernmental health system failed it. The pandemic found the DIRESAs under-equipped and under-staffed; coordination between MINSA and the regions broke down on testing data, oxygen procurement, and ICU allocation; regional hospitals were overwhelmed in successive waves (Iquitos and the Loreto oxygen catastrophe of April–May 2020 became the emblem); and the eventual recount placed Peru's COVID mortality at the highest recorded per-capita level in the world — more than 217,000 deaths by the 2021–2022 reckonings [TBD-VERIFY: final cumulative toll and the per-million figure]. The catastrophe was overdetermined — informality (PE-G-03), urban density, a fragmented health system divided among MINSA, EsSalud, and the regions, comorbidities — but every serious post-mortem assigned a major role to the broken centre–region health architecture: two decades after the transfer of health functions, neither tier could execute a coordinated emergency response through the other. The pandemic also previewed the recentralising reflex of Section 6: the emergency was managed by central commands, military logistics, and presidential decree, with the regional tier reduced to a spectator in the largest health operation in the country's history.

4.4 The Capacity Diagnosis: SERVIR and the Rotation Pathology

Beneath every failure above lies the same diagnosis: the subnational tier never acquired a professional bureaucracy. SERVIR, the national civil-service authority created in 2008 (Decreto Legislativo 1023), and the Ley del Servicio Civil (Ley 30057, 2013) were designed to build one; the Gerentes Públicos programme seconded professionally selected managers into regional and municipal posts with some documented success. But the civil-service reform's implementation stalled nationally (the transition of entities into the new regime proceeded at a pace measured in decades) and barely touched the regional tier, where the workforce remains divided among career remnants, the precarious CAS contract regime, and politically appointed funcionarios de confianza. The decisive pathology is rotation: each incoming governor — and, after the 2015 re-election ban guaranteed turnover, every government is an incoming government — replaces the management tier wholesale, from regional managers to sub-directors, with campaign loyalists. Institutional memory resets to zero every four years; project pipelines die with their sponsors; the same feasibility studies are commissioned repeatedly. The re-election ban, adopted in 2015 as an anti-corruption measure, thus carries a standing indictment: by guaranteeing turnover it deepened the very capacity deficit that enables corruption, a trade-off the 2026-era reform debates have begun to revisit [TBD-VERIFY: status of congressional proposals to restore subnational re-election].

5. The Political Dimension

5.1 The Regional Movements: The Party Collapse's Territorial Expression

The decentralisation created twenty-five new political arenas at the precise historical moment the national party system was disintegrating, and the arenas were filled by what the electoral law calls movimientos regionales: parties registered in, and confined to, a single region. The trajectory is stark. In 2002, national parties still won most regional presidencies (APRA's twelve led the field). By 2006 the movements had drawn level; by 2010 and 2014 they dominated; in the 2014 and 2022 regional elections, regional movements won the large majority of governorships, with national parties reduced to a handful [TBD-VERIFY: exact governorship counts by party type per cycle — commonly cited figures give national parties single-digit wins in 2014, 2018, and 2022]. The movements are, with exceptions, one-election vehicles: built around a single caudillo or family, financed opaquely (mining-service contractors, construction interests, and in documented cases illegal economies), renamed or dissolved between cycles, and organisationally empty between campaigns. Cusco's, Áncash's, and Puno's electoral histories read as successions of such vehicles. The phenomenon is the territorial face of the national party collapse analysed in PE-O-01 §2: the same collapse that produces single-use presidential vehicles in Lima produces single-use gubernatorial vehicles in the regions, and the two layers barely interact — national parties cannot win governorships, and regional movements are legally barred from contesting national office, a wall the electoral framework maintains deliberately [TBD-VERIFY: the post-2019 electoral reforms' treatment of regional movements, including registration and cancellation rules].

The consequences run in both directions. Downward, the movements give regional government its characteristic instability: no programmatic continuity, no bench of cadres, no accountability beyond the caudillo's person. Upward, they foreclose the classic federalising path by which subnational politics seeds national parties — Brazil's or Argentina's governors-as-party-barons dynamic has no Peruvian equivalent, because the Peruvian governor commands no transferable machine. The one partial exception proves the rule: César Acuña's Alianza para el Progreso (APP), built from a La Libertad base (and the Universidad César Vallejo's resources) into a registered national party, is the only durable national organisation of the period with subnational roots — and it functions less as a programmatic party than as a franchise network for local notables.

5.2 The Governorship as Failed Springboard

For a system desperate for national leadership renovation, the governorship has produced almost none. Acuña's presidential candidacies (2016, disqualified mid-campaign; 2021) failed. Gregorio Santos's 2016 candidacy from pre-trial detention was marginal. The various governors' attempts to scale southern or northern regional success nationally have foundered on the movements' non-transferability. The single governor-to-president trajectory of the period — Martín Vizcarra, Moquegua's 2011–2014 governor, who reached the presidency in March 2018 — ran through appointment (PPK's first vice-presidency) rather than electoral scaling, and Vizcarra's subsequent national popularity (and his 2021 election to Congress with the period's highest individual vote before his disqualification) suggests the demand for province-credentialed national leadership exists even though the supply channel is broken (PE-H-PRES-08). The blockage matters constitutionally: a political system whose meso-tier cannot produce national leaders loses both a screening mechanism (governors arrive at national office pre-tested by executive responsibility) and a integrative one (national coalitions built from territorial blocks). Peru's presidents of the period instead arrived from outsider trajectories — a World Bank economist, a military comandante, a rural schoolteacher — with no executive record to examine and no territorial coalition to govern with.

5.3 The Conflict-Mediation Role: Between Lima and the Communities

In the mining conflicts that punctuate the period (the full architecture is PE-G-02's subject; the Baguazo and consulta previa thread is PE-G-02 §10), the regional government occupies a structurally impossible position: the permits, the environmental certifications, and the fiscal terms are decided in Lima; the costs land on communities; and the governor — elected by those communities but dependent on the canon the projects generate — stands between them with responsibility and no authority. The period's two emblematic governors marked the role's two poles. Vizcarra at Quellaveco (Moquegua, 2011–2012) chaired the mesa de diálogo that converted a looming Tía María-style confrontation into a negotiated settlement — water-use guarantees, a development fund, community monitoring — that allowed the Anglo American project to proceed and stands as the system's best case of regional mediation. Santos at Conga (Cajamarca, 2011–2012) took the opposite road: the governor as movement leader, placing the regional government at the head of the anti-project mobilisation, issuing a regional ordinance against the project that the Constitutional Tribunal struck down as exceeding regional competence [TBD-VERIFY: the TC ruling on Cajamarca's Conga ordinance — confirm case number and disposition], and prevailing in substance when the project was indefinitely suspended. Both poles illustrate the same institutional fact: the regional tier has no formal role in the decisions that most affect its territory — mining permitting is the centre's exclusive competence — so governors choose between brokering without power and mobilising without responsibility.

5.4 The Southern Corridor: Decentralisation's Indictment

The December 2022 – 2023 protest cycle (PE-D-03) constituted, among other things, the decentralisation's harshest audit. The protest geography — Puno, Cusco, Apurímac, Ayacucho, Arequipa's highlands; the southern Andean corridor that absorbed nearly all of the 49+ deaths, including Juliaca's seventeen on 9 January 2023 — maps onto the regions where the 2002 settlement delivered elected offices and, in several cases, large canon flows, yet where state-presence indicators (paved-road density, health staffing, water access) remain the country's lowest and trust in Lima institutions runs at single digits. Two decades of formal self-government had given the south neither effective services nor an effective voice in national decisions, and when its presidential choice was removed in circumstances it read as a Lima coup against the provinces, the response bypassed every regional institution: the mobilisation was organised through communal structures, rondas, and ad-hoc committees, not through governorships or regional councils, which were largely irrelevant or hostile bystanders. The demand structure — Boluarte's resignation, new elections, a constituent assembly — carried the constitutional question on a regional-grievance chassis. The settlement question this opens for the 2030s — accommodation, secession of attention, or rupture — is analysed as a megatrend in PE-O-01 §6; what this document adds is the institutional reading: the 2022–2023 eruption is what it looks like when a decentralisation distributes offices without power for twenty years and the periphery concludes that the offices were the point.

6. The Recentralisation Currents (2016–2026)

6.1 Recentralisation by Bypass: The ARCC Model

No Peruvian government has proposed repealing the 2002 settlement — the constitutional text declaring decentralisation "permanent and obligatory" stands untouched, and the political cost of formally stripping elected governments would be prohibitive. What the centre has built instead, accelerating after 2016, is machinery to route around the regional tier in everything that matters operationally. The template was the Autoridad para la Reconstrucción con Cambios (ARCC), created by Ley 30556 (April 2017) after the El Niño costero floods devastated the northern coast. The ARCC concentrated a multi-billion-dollar reconstruction portfolio — initially around S/ 25 billion [TBD-VERIFY: the reconstruction plan's headline amount and its subsequent revisions] — in a central authority attached to the PCM, with regional and local governments reduced to executing units for the smaller works and bypassed entirely for the larger ones. When execution lagged through 2017–2019 under successive directors, the response was not devolution but deeper centralisation and externalisation: the June 2020 government-to-government (G2G) agreement with the United Kingdom, importing British project-management offices to deliver schools, hospitals, and the integral river-basin works. The G2G model itself had been validated by the Lima 2019 Pan American Games: a special project, UK-partnered, delivered on time after years of delay, which Lima's political class read as proof that the ordinary public-investment system — national and subnational alike — could be circumvented rather than fixed.

The Boluarte government completed the institutionalisation: the Autoridad Nacional de Infraestructura (ANIN), created in 2023 [TBD-VERIFY: ANIN's enabling statute and the timeline of the ARCC's wind-down and absorption], converts the emergency-reconstruction authority into a permanent central delivery agency for large infrastructure, explicitly modelled on the G2G/special-project experience. In parallel, the Plan Nacional de Infraestructura para la Competitividad (2019, updated 2022) and the proliferating proyectos especiales de inversión pública run flagship works — hospitals, the Lima metro lines, the new Chinchero airport, the escuelas bicentenario — through ministerial special units with their own procurement regimes. Each instrument has a defensible delivery rationale, and several have delivered. Collectively they constitute a quiet constitutional revision: the investment functions that the 2002 framework assigned to the regional tier are migrating back to the centre, project by project, without a single amendment — recentralisation by bypass, purchased with the corruption harvest and the execution record of Section 4 as its standing justification.

6.2 The Sectoral and Fiscal Clawbacks

The bypass machinery sits atop a broader pattern. Social programmes were consolidated under MIDIS (created 2011) as centrally administered transfers (Juntos, Pensión 65, Qali Warma and its troubled successor), with municipalities as registration points rather than managers. In health, the pandemic's aftermath produced central commands and national procurement rather than DIRESA strengthening. In security — the 2025–2026 extortion-crisis years — states of emergency placed successive provinces under police-military command from Lima (PE-D-05, PE-D-07). Fiscally, the MEF's instruments (Invierte.pe gatekeeping, performance-conditioned transfer schemes, the obras por impuestos mechanism that lets corporate taxpayers build works directly against tax credits) each individually reasonable, cumulatively deepen central control of the subnational investment agenda. The ANGR and the municipal associations have documented and protested the trend through successive informes anuales del proceso de descentralización — the PCM's own annual reports to Congress concede the process has been stalled for over a decade [TBD-VERIFY: the characterisation in recent PCM/Secretaría de Descentralización annual reports] — but the protests collide with the governors' weak national standing: a tier whose public image is the Section 4.1 prison roll cannot win the argument for more power.

6.3 The Bicameral Restoration's Territorial Dimension

The one structural reform moving the other way is parliamentary. The bicameralism restoration (Ley 31988 of March 2024 and Ley 32101 of November 2024, full treatment in PE-A-04 and PE-I-01) creates, from 28 July 2026, a 60-member Senate elected alongside the 130-member Chamber of Deputies — and the Senate's design carries a territorial-representation promise: a substantial portion of senators elected from departmental constituencies, giving each region a guaranteed voice in national lawmaking for the first time since the 1992 abolition of the old Senate [TBD-VERIFY: the final allocation under Ley 32101 between senators elected in a single national district and those elected by departmental/regional districts, and the per-region seat counts]. Advocates argue the regional senators will supply what the decentralisation never had — a national-level institutional lobby for the territorial agenda, capable of forcing the fiscal-decentralisation second stage and disciplining the bypass machinery. Sceptics note that departmental senators will be elected through the same collapsed party system and the same single-use vehicles as everyone else, and that a chamber is representation, not capacity: no senate builds a DIRESA. The first bicameral Congress's treatment of the decentralisation file — alongside the 2026-era debates on restoring subnational re-election, canon reform, and the recentralisation agencies' permanence [TBD-VERIFY: the post-election legislative agenda on decentralisation against PE-D-08's account of the 2026 second round and transition] — will show which reading is right.

7. Comparative Perspective

7.1 The Latin American Wave and the Peruvian Variant

Peru's 2002 reform was a late entry in the Latin American decentralisation wave, and the comparison with its neighbours defines its specificity. Colombia decentralised earlier and deeper at the municipal level: elected mayors from 1988, the 1991 Constitution's entrenched transfer system (the situado fiscal and its successors) guaranteeing subnational shares of national revenue by constitutional formula, and municipalities as the system's load-bearing tier. Colombia's record is hardly unblemished — the transfers fed clientelism and, in conflict zones, armed-group capture — but its subnational tier acquired a fiscal constitution that Peru's never did: Colombian transfers are a constitutional right, Peruvian recursos ordinarios an annual negotiation. Bolivia constitutionalised territorial pluralism: the 2009 Constitution's autonomías — departmental, municipal, and indigenous-originary-peasant — created an asymmetric, identity-recognising framework born of the 2000s' near-civil confrontation between the media luna departments and the central state. Vergara's La danza hostil (2015), the comparative study built precisely on the Bolivia–Peru contrast, argues the divergence reflects the strength of subnational elites: Bolivia's Santa Cruz elite could force a territorial bargain on the centre; Peru's fragmented regional elites could not, so Peru got decentralisation as a top-down grant rather than a negotiated settlement — and what the centre grants, the centre can bypass. Chile, the wave's laggard, elected its regional governors for the first time only in 2021, having run a unitary-appointed model through its entire democratic restoration; its caution — sequencing capacity before election — is the explicit counter-model that Peruvian recentralisers cite, though Chile's own governors complain they received offices without powers, a complaint Peruvian governors would recognise.

7.2 The Canon as Natural Experiment: The Resource-Federalism Literature

Peru's canon has become one of the resource-federalism literature's favourite natural experiments, because the formula-driven, geologically determined allocation of windfalls across otherwise similar subnational units permits credible identification of money's effects. The findings triangulate the record of Sections 3–4. Arellano-Yanguas's influential work (¿Minería sin fronteras? and subsequent articles) found that the canon, far from buying social peace, increased local conflict in producing areas — by raising the stakes of municipal office, generating boundary and distribution disputes between districts, and focusing grievance on visible local windfalls that failed to translate into services. Econometric studies of canon-receiving municipalities find modest positive effects on some local outcomes alongside null or perverse effects on others, with the consistent moderating variable being pre-existing local state capacity [TBD-VERIFY: the specific findings of the Loayza–Rigolini and Maldonado lines of research on mining windfalls and local welfare/corruption — confirm citations before attributing results]. The Peruvian case's contribution to the general literature is the demonstration that revenue decentralisation without revenue-raising decentralisation reproduces the resource curse at subnational scale: governments that spend windfalls they did not tax face no fiscal accountability from their citizens, who correctly perceive the money as Lima's (or the mine's) rather than their own — the "rentier region" as the rentier state's miniature.

7.3 The Verdict and Its Critics: Two Readings

The mainstream verdict on Peru 2002–2026 — standard in the World Bank and IDB diagnostics, the Contraloría's own studies, and much of the Lima commentariat — is "decentralisation without capacity": a reform that transferred functions and money to governments that lacked bureaucracies, parties, and controls, and that predictably harvested corruption, under-execution, and service failure; the implied remedies are capacity-building, stronger control, and, in the stronger versions, the selective recentralisation that Section 6 shows is already underway. The critical reading — associated with the IEP tradition, Vergara, and the decentralist current in Peruvian public law — answers that the verdict mistakes the design for the provinces: Lima never transferred the powers that constitute real self-government (taxation, mining permitting, civil-service control, the budget veto), the macro-region stage that would have created viable units was abandoned after one failed referendum, the fiscal second stage was never executed, and the capacity-building that the framework promised (the accreditation system, SERVIR's regional deployment) was never resourced — so the regions were set up to fail, failed, and are now cited as evidence against themselves. On this reading the remedy is completion, not retreat: fiscal decentralisation, consolidation incentives revisited, a real subnational civil service, and a formal regional role in extractive decisions.

Both readings are empirically grounded, and the honest synthesis is that they describe a trap: the regions cannot be given more power until they demonstrate capacity, and they cannot develop capacity under an architecture that denies them power, rotates their staff, and bypasses them for every project that matters. A quarter-century in, the trap has produced a stable low-level equilibrium — elected regional governments that are too legitimate to abolish and too weak to govern — punctuated by the eruptions, of which 2022–2023 was the largest, that occur when the periphery audits the bargain.

7.4 The 2030s Fork

Three paths out of the equilibrium are visible from 2026, mirroring the scenario structure of PE-O-01 §6. Deepening: a government — most plausibly one electorally indebted to the south — executes the unfinished agenda: fiscal decentralisation's second stage, canon reform with an equalisation component, a regional civil-service programme at scale, restored subnational re-election with strengthened concurrent control, and a regional voice in extractive permitting; the new Senate's departmental bloc is the agenda's natural parliamentary carrier. Continued recentralisation by stealth: the ANIN/special-project model expands, the regional tier is progressively reduced to a ceremonial-and-residual layer, and the formal architecture is left standing as a façade — the path of least resistance, already underway, whose long-run cost is the further delegitimation of the only elected institutions the periphery has. Asymmetric reform: the legally hardest but analytically most honest path — abandoning uniformity to treat Moquegua and Huancavelica, Lima and Loreto differently, with differentiated competences earned by demonstrated capacity (the original accreditation idea, this time enforced) and special regimes for the Amazonian departments and metropolitan Lima. Which path is taken will be decided less by decentralisation policy than by the variables this corpus tracks elsewhere: the 2026 government's electoral geography (PE-D-08), the copper cycle's fiscal generosity (PE-G-01, PE-G-02), and whether the southern settlement question (PE-O-01 §6) is answered by accommodation or by another round of the 2022–2023 audit.

8. Conclusion

The 2002 decentralisation was conceived as the territorial completion of Peru's redemocratisation: if the Fujimori state had governed the provinces through appointed councils and clientelist works, the restored democracy would govern them through elected regional governments with their own budgets and competences. A quarter-century later, the reform's formal achievements are real and probably irreversible — twenty-five elected governments, six uninterrupted electoral cycles, a constitutional norm of territorial self-government that no political force proposes to repeal — and its substantive promise is largely unredeemed. The macro-regions died in a single referendum and took fiscal decentralisation with them; the canon converted the supercycle into concentrated, procyclical windfalls that the receiving tier could neither fully spend nor adequately safeguard; the governorship became, in the national imagination, a synonym for prosecution; the party system's collapse filled the new arenas with single-use vehicles; and the centre, armed with the failure record, has spent the last decade building the machinery to govern around the structure it cannot abolish.

Yet the indictment cannot be read as a brief for recentralisation without confronting what the record equally shows: that the centre authored the architecture whose failures it now cites; that the powers withheld — taxation, permitting, the budget — are precisely those whose absence makes regional government hollow; that the best regional administrations demonstrated the model can work; and that the alternative to legitimate-but-weak regional government, as 2022–2023 demonstrated, is not orderly central administration but an unmediated confrontation between Lima and a periphery that no longer believes its offices matter. The decentralisation distributed offices, not power; the 2030s question — deepening, stealth recentralisation, or asymmetric redesign — is whether Peru finally decides which of the two it intends the provinces to have. Within this corpus, this document supplies the institutional history beneath the mining-conflict architecture (PE-G-02), the southern protest geography (PE-D-03), the megatrend settlement question (PE-O-01 §6), and the bicameral restoration's territorial promise (PE-A-04, PE-I-01): the territorial state is where Peru's institutional weakness is least visible to Lima and most consequential for everyone else.

Primary Sources Consulted:

  1. Ley 27680 (Ley de Reforma Constitucional del Capítulo XIV del Título IV, 2002); Ley 27783 (Ley de Bases de la Descentralización, 2002); Ley 27867 (Ley Orgánica de Gobiernos Regionales, 2002) and Ley 27902 (2003); Ley 27972 (Ley Orgánica de Municipalidades, 2003); Ley 28274 (Ley de Incentivos para la Integración y Conformación de Regiones, 2004); Decreto Legislativo 955 (Descentralización Fiscal, 2004); Ley 30305 (2015, subnational re-election ban); Ley 30556 (2017, ARCC).
  2. Constitución Política del Perú (1993), Chapter XIV of Title IV as reformed in 2002, with Ley 31988 (2024) and Ley 32101 (2024) on the bicameral restoration.
  3. Oficina Nacional de Procesos Electorales (ONPE), results of the 30 October 2005 region-conformation referendum and the regional-municipal elections of 2002, 2006, 2010, 2014, 2018, and 2022, https://www.onpe.gob.pe.
  4. Ministerio de Economía y Finanzas, Transparencia Económica / Consulta Amigable series: canon minero, canon gasífero, regalías, FOCAM, and FONCOMUN transfers by department and municipality; subnational budget-execution data (PIA/PIM/devengado), 2004–2026, https://www.mef.gob.pe.
  5. Presidencia del Consejo de Ministros, Secretaría de Descentralización, Informes Anuales del Proceso de Descentralización presented to the Congress (2007–2025), and the records of the Consejo Nacional de Descentralización (2002–2007).
  6. Contraloría General de la República, Estudio del proceso de descentralización en el Perú (2014), audit reports on regional governments, and the concurrent-control reports on the reconstruction and Lima 2019 portfolios.
  7. Asamblea Nacional de Gobiernos Regionales (ANGR), annual reports and fiscal-decentralisation platforms (2007–2025); Asociación de Municipalidades del Perú (AMPE) and REMURPE position documents.
  8. Vergara, Alberto, La danza hostil: poderes subnacionales y Estado central en Bolivia y Perú (Lima: IEP, 2015), and Ciudadanos sin República (Lima: Planeta, 2018, expanded 2022).
  9. Arellano-Yanguas, Javier, ¿Minería sin fronteras? Conflicto y desarrollo en regiones mineras del Perú (Lima: IEP/PUCP, 2011), and subsequent articles on the canon and local conflict.
  10. Muñoz, Paula, Buying Audiences: Clientelism and Electoral Campaigns When Parties Are Weak (Cambridge University Press, 2019), and articles with Eduardo Dargent on subnational politics and state capacity, including their work on regional-government corruption.
  11. Dargent, Eduardo, and Paula Muñoz, "Perú 2016–2023" country-review articles in Revista de Ciencia Política, and Dargent, El Estado en el Perú: una agenda de investigación (PUCP, 2012).
  12. Tanaka, Martín, Democracia sin partidos: Perú 2000–2005 (Lima: IEP, 2005), and essays on regional movements and the subnational party vacuum.
  13. Zavaleta, Mauricio, Coaliciones de independientes: las reglas no escritas de la política electoral (Lima: IEP, 2014), on regional movements as electoral coalitions of independents.
  14. Eaton, Kent, articles on recentralisation in the Andes, including work on Peru's territorial politics and the limits of the 2002 reform, in Comparative Politics and edited volumes (2004–2017).
  15. World Bank, Peru: The Decentralization Process and its Links with Public Expenditure Efficiency and subsequent subnational public-expenditure reviews; Inter-American Development Bank diagnostics on Peruvian fiscal decentralisation.
  16. ProDescentralización (USAID), annual Informe Anual sobre el Estado del Proceso de Descentralización series (2005–2017).
  17. Procuraduría Pública Especializada en Delitos de Corrupción, statements and caseload reports on investigations of regional and municipal authorities (2014–2024); judicial records in the Álvarez (Áncash), Moreno (Callao), Acurio (Cusco), Viñas (Tumbes), Ríos (Áncash), and Santos (Cajamarca) cases.
  18. IDL-Reporteros, OjoPúblico, and Convoca, investigative series on La Centralita, canon-funded procurement, regional-movement financing, and the Odebrecht subnational cases, https://www.idl-reporteros.pe, https://ojo-publico.com.
  19. Autoridad para la Reconstrucción con Cambios, plan and execution reports (2017–2023); the Peru–UK government-to-government agreement documentation (2020); Autoridad Nacional de Infraestructura founding documents (2023) [TBD-VERIFY: statute citation].
  20. SERVIR (Autoridad Nacional del Servicio Civil), reports on the Gerentes Públicos programme and the implementation of Ley 30057 in subnational governments.
  21. Defensoría del Pueblo, Reportes de Conflictos Sociales (monthly, 2004–2026) and Informes Defensoriales on decentralisation, regional health services, and the 2022–2023 protest deaths.
  22. Loayza, Norman, and Jamele Rigolini, and Maldonado, Stanislao, working papers on mining windfalls, local welfare, and corruption in Peru [TBD-VERIFY: precise citations and findings]; McNulty, Stephanie, Voice and Vote: Decentralization and Participation in Post-Fujimori Peru (Stanford University Press, 2011).

Related Documents:

  • PE-I-01: The Peruvian Institutional Architecture — Congress, Constitutional Tribunal, JNE, and JNJ (1993–2026)

  • PE-A-04: Peru's Congressional Architecture and the Restoration of Bicameralism (1993–2026)

  • PE-D-03: December 2022 – 2023 Boluarte-Era Protests

  • PE-G-01: The Peruvian Economic Model — Mining Boom, Macro Orthodoxy, Informality (1990–2026)

  • PE-G-02: Peruvian Mining Political Economy (1990–2026)

  • PE-G-03: The Informality Political Economy — The Other Path and the Seventy-Percent Economy (1980–2026)

  • PE-N-01: Peru in International Perceptions — Fragile Democracy, Resource State (2000–2026)

  • PE-O-01: Peru Megatrends — The 2030s Questions

  • PE-H-PRES-03: Alejandro Toledo — A Biography

  • PE-H-PRES-08: Martín Vizcarra — A Biography

  • PE-M-01: Fujimorismo as Political Idea and Movement

  • PE-J-02: Three Accounts — The 2009 Bagua Massacre (El Baguazo)

ArchiveSourcesChat