Research Article  ·  Applied Network Science  ·  Published 2026

Network Analysis of Cultural Funding in Brazil: Uncovering Regional Differences and Signatures

Diogo Pacheco  ·  Gustavo Moraes  ·  Diego Pinheiro  ·  Josemar Faustino  ·  Marcos Oliveira  ·  Henilton Menezes  ·  Ronaldo Menezes

Applied Network Science  ·  2026  ·  DOI: 10.1007/s41109-026-00799-y

Transactions 106,804 sponsorship contributions from 2010–2014 under Brazil's Rouanet Law
Sponsors 39,809 individuals and companies who funded approved cultural projects
Approved Projects 12,326 cultural projects approved under PRONAC from 5,609 proponents
Cultural Segments 103 distinct segments across 7 areas defined by the Ministry of Culture

The Problem

Brazil's creative economy contributes 3.11% to national GDP, employing approximately 7.5 million people across more than 130,000 formalised enterprises. Public funding of cultural projects should preserve this diversity — but Brazil's main cultural funding mechanism, the Rouanet Law (Lei Rouanet, Law No. 8,313/1991), has long been criticised for concentrating resources in the Southeast, leaving peripheral states structurally disadvantaged in the competition for private sponsorship.

The law allows companies and individuals to redirect up to 6% and 4% of their income tax, respectively, to approved cultural projects. This hybrid public-private model means that approval by the government does not guarantee funding — projects must still attract private sponsors. As a result, regional disparities manifest not only in which projects are approved, but in which states' projects can actually secure the sponsorship needed to proceed.

Previous studies documented regional concentration through descriptive statistics but did not explore the underlying structural relationships between cultural segments — the systemic biases embedded in which types of culture tend to be co-funded together, and what this means for states with different cultural profiles. Network analysis offers a way to make these hidden structures visible.

"Some states are up to 40% less likely to secure private sponsorship for already-approved cultural projects — a disparity that correlates strongly with socioeconomic indicators but is not fully explained by them."

Although companies represent only 18% of all sponsors, they contribute 98% of total investment. This concentration means that the geographic location of corporate headquarters — overwhelmingly in São Paulo and Rio de Janeiro — shapes the entire funding landscape for Brazilian culture.

The Approach

The study analyses five years of Rouanet Law transactions to construct a "cultural space" — a network of cultural segments connected by statistically significant co-sponsorship relationships. This network reveals both the structural preferences of sponsors and the degree to which each Brazilian state participates in the national cultural ecosystem.

Key Findings

By combining sponsorship signature analysis with the cultural space framework, the study reveals structural inequalities in Brazil's cultural funding system that go well beyond what standard geographic concentration measures can detect.

Stark sponsorship disparities across states

Some states are up to 40% less likely to secure private sponsorship for already-approved projects. These disparities correlate strongly with GDP and other socioeconomic indicators — but a significant share of the variance remains unexplained, pointing to structural biases in how sponsors select projects beyond simple economic capacity.

Cultural signatures don't follow geography

Hierarchical clustering of state funding profiles (by cosine similarity) groups states that are geographically distant but economically similar. Neither corporate nor individual sponsorship profiles align with regional or neighbourhood proximity — each state has a distinct cultural signature shaped by local sponsor preferences and sector development history.

Companies dominate but concentrate narrowly

Companies (18% of sponsors) account for 98% of total investment. In São Paulo, corporations fund up to 89 distinct cultural segments; individuals in the same state fund approximately 57. In Amapá, corporate sponsorship spans 16 segments, while individual sponsorship covers only 2 — revealing how corporate breadth shapes a state's cultural diversity.

Performing arts and audiovisual form the backbone

The national cultural space has a dense giant component of 74 segments. The strongest φ-correlations link performing arts and audiovisual sectors, indicating these domains consistently attract shared sponsors. After FDR correction, the structural core shrinks to 47 segments, but eight of the ten most connected segments retain their rank — confirming the backbone's robustness.

Southeast dominates cultural space coverage

São Paulo achieves the broadest coverage (Sφ = 65.57), with a strong presence across nearly all cultural segments. Rio de Janeiro (58.73) and Minas Gerais (52.43) follow. These three states together account for roughly 70% of Brazil's GDP. Mato Grosso, at the opposite extreme, has Sφ = 7.70. When restricted to the FDR backbone, only six states (SP, RS, MG, RJ, PR, DF) retain non-zero coverage.

A self-reinforcing structural cycle

Because corporate sponsors are predominantly headquartered in the Southeast, cultural organisations in these states benefit from greater access to sponsorship networks — attracting broader cultural diversity, which in turn draws additional funding. This self-reinforcing cycle means that the structural imbalance is not simply a resource allocation problem: it is embedded in the geography of where sponsors are located.

Why It Matters

The cultural space framework gives policymakers a structural map of Brazil's cultural funding system — not just a ledger of who received how much, but a picture of which cultural domains are interconnected, which states are embedded in that network, and where targeted interventions could have lasting structural effects.

Infographics

Pessoa jurídica vs pessoa física: very different cultural footprints
Companies (pessoa jurídica, left) are few but reach far — each large node funds 8–9 distinct cultural segments, shown as the dense web of flows. Individuals (pessoa física, right) are numerous but each reaches only 1–2 segments, producing a sparse scatter. Although companies are just 18% of all sponsors, they contribute 98% of total investment and account for the structural breadth of Brazil's cultural funding.
Pessoa Jurídica — Companies 6,999 sponsors · 98% of investment · broad segment reach
vs
Pessoa Física — Individuals 32,810 sponsors · 2% of investment · narrow segment reach
Cultural space coverage across Brazilian states
Each coloured node is a cultural segment; brighter nodes are actively covered by sponsors from that state, grey nodes are largely absent. The Sφ metric summarises how broadly a state's sponsors participate in the national cultural space. Watch the network go from nearly fully lit (São Paulo) to mostly dark (Mato Grosso) as the state cycles.
Cycling through 8 states — from highest to lowest Sφ coverage state abbreviations shown at bottom

Paper & Authors

Authors

Diogo Pacheco (corresponding) Department of Computer Science & BioComplex Laboratory
University of Exeter, UK
Gustavo Moraes MDCC, Federal University of Ceará
Fortaleza, CE, Brazil
Diego Pinheiro Universidade Católica de Pernambuco
Recife, PE, Brazil
Josemar Faustino BioComplex Laboratory
University of Exeter, UK
Marcos Oliveira Department of Computer Science
University of Exeter, UK
Henilton Menezes Cultural Ministry of Brazil
Brasília, DF, Brazil
Ronaldo Menezes Department of Computer Science & BioComplex Laboratory
University of Exeter, UK

Cite This Work

@article{pacheco2026cultural, title = {Network Analysis of Cultural Funding in {Brazil}: Uncovering Regional Differences and Signatures}, author = {Pacheco, Diogo and Moraes, Gustavo and Pinheiro, Diego and Faustino, Josemar and Oliveira, Marcos and Menezes, Henilton and Menezes, Ronaldo}, journal = {Applied Network Science}, year = {2026}, doi = {10.1007/s41109-026-00799-y} }

Open access · CC BY 4.0