Theme 32 of 43
Connectivity, Digital Education & Telecom Infrastructure
Schools receive, in practice, only 30 to 40% of the internet speed they contracted for — and Brazilians pay 3.5% of the minimum wage for 1 GB, more than Mexico and Chile.
SIMET directly measures, inside Brazilian schools, the actual download speed, upload speed, and latency of the contracted connection, while Anatel tracks fixed-broadband access density and the Brazilian Connectivity Index — which combines 4G and 5G coverage, fiber-optic presence, and market concentration by municipality. Cross-referencing these measurements with Enem's socioeconomic questionnaire makes it possible to link the quality of available internet directly to student academic performance. The picture that emerges is of a country with two-speed connectivity: one for capital cities and private schools, another — far slower and more expensive — for the interior and the public network.
Digital deserts coincide with infrastructure deserts
The territorial distribution of basic infrastructure in Brazil follows a pattern that repeats across virtually every infrastructure indicator: the Southeast concentrates the highest density of equipment and services, while the North and Center-West have far scarcer infrastructure per capita. It's the same geographic layout that, further on, repeats in internet coverage — basic infrastructure and connectivity move together, and both are scarce in the same regions.
| Region | Health facilities |
|---|---|
| Southeast | Highest concentration |
| North | Least infrastructure |
| Center-West | Few facilities |
The North has less basic infrastructure per capita than any other region — the same territorial pattern that, further along, also shows up in connectivity.
The telecom oligopoly is also a regional desert
The Brazilian Connectivity Index confirms what the telecommunications market already showed in other indicators: a concentration index above 2,500 points, with just three companies accounting for 80% of the national market. This oligopoly isn't spread evenly across the territory — it overlaps exactly with the regions already identified as digital deserts, the North and Center-West, worsening the scarcity of infrastructure investment there.
| Indicator | Concentration |
|---|---|
| Telecom HHI | > 2500 (oligopoly) |
| Dominant companies | 3 control 80% |
| Digital deserts | North, Center-West |
The telecom sector is oligopolized — and that oligopoly overlaps exactly with the regions that are already digital deserts.
Rural and public: the worst internet in schools
Connection speed measured in schools varies drastically by location and school network: urban schools have speeds ten times higher than rural ones, and private schools, five times higher than public ones. Only 40% of schools reach the 100 Mbps target considered adequate for intensive pedagogical use of the internet — leaving most of Brazil's public network, especially in the interior, well short of the minimum needed for classes and assessments that depend on a stable connection.
| School type | Speed |
|---|---|
| Urban | 10x higher than rural |
| Private | 5x higher than public |
| Target (100 Mbps) | Reached by 40% |
Rural and public schools have the worst connectivity in the country — and only four in ten schools reach the minimum speed considered adequate.
The poorer you are, the worse your internet
The municipal Brazilian Connectivity Index correlates inversely with the Social Vulnerability Index: more vulnerable municipalities consistently register worse internet quality. It isn't a coincidence — it's a self-feeding cycle, in which lack of connectivity limits access to education, remote work, and digital services, which in turn deepens the very vulnerability that caused the digital exclusion.
Digital exclusion isn't a side effect of poverty — it's part of the mechanism that perpetuates territorial inequality in Brazil.
Promised speed isn't delivered speed
Direct measurements taken by SIMET inside schools show a serious gap between what's contracted and what's actually delivered: across every contract tier tested, schools receive between 30% and 40% of the promised speed. This isn't an isolated failure — it's a systematic pattern that suggests insufficient oversight of compliance with the connectivity contracts signed with public authorities for education.
| Contract | Promised speed | Actual speed | % |
|---|---|---|---|
| 10 Mbps | 10 Mbps | 3 Mbps | 30% |
| 20 Mbps | 20 Mbps | 8 Mbps | 40% |
| 100 Mbps | 100 Mbps | 30 Mbps | 30% |
Brazilian schools receive, in practice, between 30% and 40% of the internet speed that was contracted — actual quality is even worse than coverage numbers suggest.
5G already reaches a third of phones — but not half of the municipalities
Anatel's monthly access records show fast adoption: in August 2022, when the network was first switched on, only 1.6% of mobile lines with data used 5G; by August 2026 the share was 31.6%, nearly 70 million lines. Expansion still follows the urban hierarchy. All 27 state capitals have 5G users, and there the technology already accounts for 41.7% of lines; in the rest of the country the share is 27.6%, and 2,740 of the 5,570 municipalities still record no 5G access at all. The Northeast lags furthest, at 21.1% of lines — against 37% in the Southeast.
| Year | 5G lines | % of lines | Municipalities with 5G |
|---|---|---|---|
| 2022 | 3.6 million | 1.6% | 77 |
| 2023 | 14.6 million | 6.9% | 316 |
| 2024 | 33.2 million | 15.3% | 879 |
| 2025 | 52.0 million | 24.1% | 1,800 |
| 2026 | 69.7 million | 31.6% | 2,830 |
| Region | % of lines on 5G | Municipalities without 5G |
|---|---|---|
| Southeast | 37.0% | 901 of 1,668 |
| Center-West | 33.5% | 276 of 467 |
| South | 33.0% | 318 of 1,191 |
| North | 27.9% | 293 of 450 |
| Northeast | 21.1% | 952 of 1,794 |
In four years 5G went from 1.6% to 31.6% of mobile lines — yet half of Brazil's municipalities still have not a single access on the new network.
Fixed broadband in the North and Northeast belongs to small providers
In fixed internet, regional concentration is the opposite of what is usually assumed. In December 2025 the country had 52.7 million fixed broadband accesses, 77% of them over fiber — in 2019 fiber was 31%. The shift was driven by small providers, which went from 31% to 63% of accesses. In the Northeast they serve 83% of connections, and in the North 77%; only in the Southeast do the large carriers still split the market roughly in half. Fiber reached the interior through thousands of local companies, not the big carriers.
| Year | Accesses | % fiber | % small providers |
|---|---|---|---|
| 2019 | 32.9 million | 30.7% | 31.1% |
| 2021 | 40.7 million | 59.3% | 47.4% |
| 2023 | 47.2 million | 71.5% | 54.1% |
| 2025 | 52.7 million | 77.0% | 63.4% |
| Region | Accesses | % small providers |
|---|---|---|
| Northeast | 8.8 million | 83.1% |
| North | 2.9 million | 77.3% |
| South | 10.5 million | 73.6% |
| Center-West | 4.2 million | 68.0% |
| Southeast | 26.4 million | 50.6% |
Eight in ten fixed connections in the Northeast come from small providers — fiber reached the interior through local companies, not the big carriers.
The pandemic accelerated connectivity — but not enough
The pandemic forced a real expansion of school connectivity: the share of schools with internet rose from 40% to 60%, and online classes went from a marginal fraction to 70% of instructional activities at the peak of the crisis. But that acceleration didn't close the gap — more than half of Brazilian students still lacked real, stable internet access, and dropout associated with the pandemic is estimated at two million students.
| Indicator | Before | After |
|---|---|---|
| Schools with internet | 40% | 60% |
| Students with access | 30% | 45% |
| Online classes | 5% | 70% |
| Dropout (COVID) | — | +2 mi |
The pandemic accelerated school connectivity, but 55% of Brazilian students remained without real internet access — and two million dropped out in the process.
Brazilians pay more for less internet
Compared to neighboring countries with similar development levels, the cost of 1 GB of mobile internet in Brazil eats up a much larger share of the minimum wage: more than double Mexico's and more than double Chile's. Only South Africa, among the countries compared, pays proportionally more than Brazil — which sits poorly alongside the already-identified low 4G and 5G coverage outside the capitals.
| Country | Cost of 1 GB (% of minimum wage) |
|---|---|
| Mexico | 1% |
| Chile | 1.5% |
| Brazil | 3.5% |
| South Africa | 4% |
Brazilians spend 3.5% of the minimum wage on 1 GB of internet — more than double Mexico or Chile, for lower-quality coverage.
Powerful cross-references
- Connectivity × Education: rural and public schools have the worst internet in the country.
- IBC × Vulnerability: poorer municipalities have less access to quality internet.
- Telecom × Oligopoly: concentration index above 2,500 points.
- Speed × Actual: schools receive between 30% and 40% of contracted speed.
- 4G × Rural: 15% rural coverage versus 95% in capitals — a 6x gap.
- 5G × Interior: 41.7% of mobile lines on 5G in state capitals versus 27.6% elsewhere; 2,740 municipalities with no 5G access at all (Aug 2026).
- Broadband × Small providers: 83% of fixed connections in the Northeast come from small providers, versus 51% in the Southeast (Dec 2025).
- Cost × Wage: 3.5% of the minimum wage for 1 GB — pricier than comparable countries.
Explanatory hypotheses
Telecom carriers don't invest in areas with low projected profitability, leaving regions with lower population density and lower average income systematically short of needed investment. Mobile telephony, even with 4G or 5G, doesn't adequately replace fiber optics for intensive educational use, which requires the stability and upload speed mobile networks don't deliver. And the even greater market concentration in the North and Northeast shows that the telecom oligopoly is most severe precisely in the peripheral regions — an even more captured market where competition should be most needed.
Policy implications
Direct public investment in fiber-optic infrastructure could correct the disparity that the market, on its own, has no incentive to fix. Requiring coverage expansion as a condition for spectrum licenses could force investment in currently neglected areas. Subsidies for low-income internet access could democratize access where the cost per GB still eats up a significant share of the minimum wage. Turning the school into a connectivity hub, sharing school internet with the surrounding community, could amplify the impact of each infrastructure investment. And price regulation could bring the cost of a gigabyte down from 3.5% toward levels closer to 1% of the minimum wage, already practiced by neighbors like Mexico.