The School You Attended in São Paulo Depends on Where You Were Born
There is an experiment you can do without leaving the city of São Paulo.
Take two babies born on the same day, in the same municipal hospital, in the same year. Place one in Moema. Place the other in Marsilac. Don't change anything else — same city, same mayor, same public education system on paper.
Twenty years later, the chances of these two young people having the same level of education, the same access to opportunities and the same economic outlook are statistically close to zero.
Access data from 5,570 municipalities via API
GDP, HDI, safety, labor market, connectivity and more. Free plan available.
Continue reading the full article below
Official data from the São Paulo City Hall explain why.
The Number That Summarizes Everything
The IDHM-Education measures the educational dimension of human development: it combines adult schooling with school attendance of children. It ranges from 0 to 1.
In Moema, the IDHM-Education is 0.917.
In Marsilac, it is 0.566.
The difference is 0.351 points. To put it in perspective: the difference in the education component between Brazil as a whole and El Salvador — countries at completely different stages of development — is 0.102 points. The educational gap within São Paulo is three times larger than that.
Seven districts of São Paulo have IDHM-Income equal to 1.0 — the index ceiling. They are: Alto de Pinheiros, Campo Belo, Moema, Morumbi, Perdizes, Pinheiros and Vila Mariana. In these districts, income is no longer a limiting factor for anything. They concentrate 15,200 families in CadÚnico. Combined.
The 10 districts with the worst IDHM-Education concentrate 344,345 families in CadÚnico — 19.3% of all low-income families in the city, in just 10 of 96 districts.
The Two Extremes in Detail
Moema — Where Education Reached the Ceiling
IDHM-Education 0.917. IDH-Income 1.0 (the maximum possible). 681 families in CadÚnico in a universe of 76 thousand inhabitants.
In Moema, the question about access to quality education already has an answer before it's even asked. The concentration of high-level private schools, prep courses, colleges and graduate programs in the surroundings ensures that the educational trajectory of the neighborhood's children is practically determined from birth — but in a positive way.
The Moema family child who does poorly in public school finds prep courses. The one who does well finds scholarships. The one who wants to study medicine finds a support system that doesn't exist in other districts. It's not individual merit. It's private and public educational infrastructure adding up in a territory where both have quality.
Marsilac — Where Education Still Hasn't Really Arrived
IDHM-Education 0.566. The lowest in São Paulo. By a wide margin.
Marsilac is, in area, the largest district in São Paulo: 195 square kilometers of Atlantic Forest, reservoirs, dirt roads and isolated communities in the far south of the municipality. It has approximately 8,700 inhabitants spread across a territory larger than the municipality of Joinville itself.
The nearest school for many children is kilometers away on a dirt road. On rainy days, access is compromised. The internet — where it exists — is intermittent. The teacher who agrees to work in Marsilac needs motivation that goes far beyond salary.
The result shows in the IDHM-Education of 0.566: a significant portion of Marsilac's adults did not complete primary school. Not for lack of intelligence or effort — but because the system simply didn't reach them.
The Pairs That Make the Data Impossible to Ignore
The difference is not just between the absolute extremes. It appears in comparisons that reveal the map of a divided city:
Moema (edu 0.917, 681 CadÚnico families) vs Grajaú (edu 0.651, 87,582 families): Difference of 0.266 in IDHM-Education. Ratio of 128 times in the number of vulnerable families. Grajaú has 360 thousand inhabitants — it's larger than Florianópolis — but with educational development conditions incomparably worse than a neighborhood of 76 thousand fewer people less than 20 kilometers away.
Vila Mariana (edu 0.909) vs Cidade Tiradentes (edu 0.646): Both are large districts. Vila Mariana, in the privileged south quadrant, has 4,635 families in CadÚnico. Cidade Tiradentes, in the far east, has 49,535 — ten times more. The difference in IDHM-Education is 0.263 points — larger than the educational difference between Brazil and Equatorial Guinea.
Pinheiros (edu 0.865) vs Lajeado (edu 0.657): Pinheiros has 2,112 families in CadÚnico. Lajeado has 39,742 — eighteen times more. In Pinheiros, the educational problem is refinement: which college, which language course, which exchange program. In Lajeado, it's fundamental: finishing high school, getting a place in technical school.
Grajaú: The Largest and Most Unequal
Grajaú not only leads the ranking of families in CadÚnico in São Paulo — 87,582 families, the largest of all 96 districts — but also is the district with the greatest internal inequality.
Within Grajaú, the HDI ranges from 0.638 to 0.891 — a range of 0.253 points. This means there is a part of Grajaú with a development index comparable to middle-income countries, and another part in the same territory with an index comparable to the poorest regions of Brazil.
This happens because Grajaú is huge — more than 360 thousand inhabitants — and internally heterogeneous. Neighborhoods near main avenues have access to public and private facilities. The alleys on the hillsides have a completely different reality. The single number for the district erasesthis heterogeneity.
It's a real methodological problem: any indicator by district is an average. And averages lie when the distribution is very unequal.
The Sacomã Paradox
Not every map is predictable. Sacomã appears as a case that defies simple logic.
With 34,866 families in CadÚnico — a high number, an indicator of real socioeconomic vulnerability —, Sacomã has IDHM-Education of 0.827. It is one of the highest in the city, above districts like Santana, Ipiranga and Vila Guilherme, which have far fewer vulnerable families.
What explains Sacomã? A combination of industrial and working-class history that values technical training, the presence of state and municipal schools with consistent performance, and proximity to ABC Paulista, which created a culture of education as an instrument of social mobility during the decades of industrial expansion.
It is evidence that the relationship between income and education is not deterministic — and that continuous investment in educational infrastructure can create islands of quality even in territories of high economic vulnerability.
What Maximum Income Doesn't Solve Alone
A disturbing fact: the 7 districts with IDHM-Income equal to 1.0 — the absolute ceiling of the index, indicating that income is no longer the limitation — none of them have IDHM-Education above 0.917.
In other words: even where income reached the maximum measurable by the index, education has not yet.
This does not mean that Moema has an educational problem. It means that IDHM-Education measures the current adult population's schooling — and that previous generations, even in the wealthiest districts, grew up in contexts where access to higher education was not yet universal. The index captures history, not just the present.
But it captures something else too: in the poorest districts, where income is far from 1.0 and education is at 0.566, today's young generations are being trained in a system that prepares them for less. The cycle perpetuates itself.
Why This Matters Now
São Paulo has, in July 2024, 1,786,205 families registered in CadÚnico. Almost 5 million people.
These families are concentrated in districts with the worst IDHM-Education. It is not a coincidence — it is accumulated causality over decades. Low parental education reduces income. Low income limits access to children's education. Children grow up with low education. The cycle begins again.
Breaking this cycle requires more than schools — it requires that schools exist near where children live, that teachers want to be there, that transportation works, that parents have income so they don't need their children to work early.
None of these problems are new. What the data does is make the scale of the problem impossible to downplay.
It is not "a part of the city that still needs attention". It is 344 thousand families concentrated in 10 districts where the educational dimension of human development is equivalent to that of countries we consider radically less developed than Brazil.
Within the same city. Under the same mayor. In the same public education system.
Methodology
The IDHM-Education by UDH (Human Development Unit) was calculated by UNDP based on the 2010 Census and made available by the São Paulo City Hall on dados.prefeitura.sp.gov.br. Values by district were calculated as the average of the UDHs in each district. CadÚnico data is from July 2024, from the same portal. The analysis covers 85 of the 96 districts — the remaining 11 were not crossed due to name discrepancies between sources.
The comparison with countries uses the 2023 Human Development Report from UNDP, education component.
Data from all districts of São Paulo is available at scorecidades.com.br
Source: São Paulo City Hall · UNDP/Brazil Atlas · CadÚnico/MDS · Score de Cidades
Cofundador do Seu Crédito Digital e idealizador do Score de Cidades. Jornalista, bacharel em Administração de Empresas pela UFRGS e especialista em SEO e inteligência territorial. Responsável pela curadoria e metodologia dos dados de cidades, estados e bairros.