A single negative assessment can prevent a bond from forming
En las universidades austriacas, un equipo de investigadores ha construido un espejo matemático que refleja una verdad antigua: los seres humanos tienden a vincularse con quienes se les parecen, y ese impulso, multiplicado por cada categoría identitaria que la sociedad añade, se convierte en arquitectura invisible de la segregación. El modelo MAPS no inventa este fenómeno, sino que lo mide con precisión suficiente para que la política pública pueda actuar sobre él. La pregunta que deja abierta no es si la homofilia existe, sino si las sociedades tienen la voluntad de diseñar los espacios donde la diferencia se vuelva encuentro.
- Cada rasgo identitario —etnia, edad, clase económica, género— actúa como un filtro independiente: basta que uno falle para que una relación no llegue a formarse.
- En 70 escuelas secundarias estadounidenses con casi 42.000 estudiantes, la etnia y el nivel académico se impusieron como las fuerzas más poderosas de segregación entre pares.
- En los matrimonios de las cincuenta ciudades más grandes de EE.UU., la clase económica mostró una lógica distinta: quienes están abajo aspiran hacia arriba, pero quienes están arriba se cierran sobre sí mismos.
- Los investigadores advierten que multiplicar categorías identitarias sin intervención deliberada no amplía la diversidad social, sino que levanta más barreras.
- La solución propuesta no es abstracta: mezcla escolar planificada, vecindarios heterogéneos y políticas que prioricen el contacto real entre grupos distintos.
Un equipo de la Universidad Tecnológica de Graz y el Complexity Science Hub de Austria ha desarrollado un modelo matemático llamado MAPS que cuantifica algo que la sociología intuía pero no podía medir con precisión: la manera en que los rasgos identitarios funcionan como filtros invisibles en la formación de amistades y matrimonios.
El hallazgo más revelador no es que la homofilia existe, sino cómo opera. El modelo descubrió que las personas no sopesan los rasgos de los demás en una escala gradual, sino con una lógica de todo o nada: una relación solo se forma si cada dimensión relevante —etnia, clase, edad, género— recibe una evaluación positiva. Un solo rechazo basta para que el vínculo no nazca.
Al aplicar MAPS a datos de casi 42.000 estudiantes de secundaria en Estados Unidos, la etnia y el nivel académico emergieron como los principales organizadores de los grupos de amigos. Un detalle esperanzador: los estudiantes de origen étnico mixto actuaban a veces como puentes entre grupos separados, insinuando que la diversidad puede crear conexiones donde de otro modo no las habría.
En los patrones matrimoniales de las cincuenta ciudades más grandes del país, la etnia, la edad y el género funcionaron como filtros primarios. La clase económica, sin embargo, mostró una asimetría: las personas de ingresos bajos y medios tendían a buscar parejas de mayor nivel económico, mientras que quienes ya ocupaban los estratos más altos se emparejaban entre sí.
La advertencia central del equipo es que cada nueva categoría identitaria que una sociedad incorpora a su vocabulario colectivo se convierte, sin intervención deliberada, en una barrera adicional. El remedio que proponen es concreto: políticas de integración que apuesten por el contacto real entre grupos —aulas mixtas, vecindarios diversos— en lugar de multiplicar las etiquetas. El modelo MAPS, concluyen, no es solo un diagnóstico; es una herramienta para saber dónde intervenir.
A team of researchers at Austria's Graz University of Technology and the Complexity Science Hub has developed a mathematical framework that explains why people cluster into homogeneous groups—and why those clusters persist. The model, called MAPS, quantifies something sociologists have long observed but struggled to measure precisely: the way identity traits like age, ethnicity, gender, and economic status act as invisible gatekeepers in friendship and marriage.
The MAPS model works by testing different mechanisms of preference aggregation against real social data. Rather than assuming people weigh traits on a sliding scale, the researchers found that the data fit best with a stricter logic: a relationship forms only when every relevant identity dimension is evaluated positively. A single negative assessment—whether on ethnicity, economic class, or academic standing—can be enough to prevent a bond from forming, even if other dimensions align. This all-or-nothing selectivity, the team discovered, is what drives social fragmentation.
When the researchers applied MAPS to school friendships across 41,800 American high school students, the pattern became clear. Grade level and ethnicity emerged as the dominant forces shaping peer groups. Gender mattered less, but the overall trend was unmistakable: students gravitated toward classmates who shared their ethnic background and academic level. The data from 70 schools confirmed this segregation was stable and reproducible. Interestingly, students of mixed ethnic background sometimes served as bridges between otherwise separated groups, suggesting that diversity itself can create pathways for connection.
Marriage patterns in America's fifty largest cities told a similar story. Age, gender, and ethnicity functioned as primary sorting mechanisms. But economic status operated differently. People in lower and middle income brackets often sought partners from higher economic tiers—an aspirational pairing. Those already at the top, by contrast, married within their own stratum. Researcher Samuel Martin-Gutierrez emphasized that this multidimensional selectivity is what perpetuates social boundaries. Each added identity category, without deliberate intervention, raises the likelihood of fragmentation.
The implications are sobering. As societies become more conscious of identity categories—adding more dimensions to how we classify ourselves and others—the mathematical model suggests we risk creating more barriers, not fewer. Each new dimension becomes another gate through which a potential connection must pass.
But the research also points toward solutions. The team argues that integration policies should focus on creating opportunities for cross-group interaction rather than proliferating identity categories. Educational mixing—deliberately placing students from different backgrounds in the same classrooms—and diverse neighborhood initiatives can foster heterogeneous bonds. The key is interaction itself. Karimi and her colleagues stress that understanding how people evaluate identity traits is not merely academic; it is essential for designing public policy that builds cohesion rather than fragmentation. The MAPS model offers a tool for identifying which dimensions carry the most weight in any given context, allowing policymakers to target interventions where they matter most.
Citações Notáveis
A single negative assessment on any identity dimension can be enough to prevent a relationship from forming, even if other dimensions align— Samuel Martin-Gutierrez, researcher on the MAPS study
Understanding how people evaluate identity traits is essential for designing public policy that builds cohesion rather than fragmentation— Fariba Karimi, lead researcher at Graz University of Technology
A Conversa do Hearth Outra perspectiva sobre a história
So the model is saying that people use a kind of veto system when choosing friends or partners—not a weighted average, but an all-or-nothing check?
Exactly. It's not that ethnicity counts for 40 percent and class for 30 percent. It's more like: if I find your ethnicity acceptable AND your class acceptable AND your age acceptable, then we can be friends. If one dimension fails, the whole thing fails.
That sounds almost mechanical. Do people really think that way?
The data suggests they do, at least in aggregate. When the researchers tested different models against actual friendship and marriage patterns, this strict evaluation mechanism fit the real world better than softer alternatives. It's not that individuals are consciously running through a checklist. It's that when you look at millions of relationship choices, this pattern emerges.
And the researchers found that adding more identity categories makes segregation worse?
Yes. Each new dimension you add becomes another hurdle. If a school starts tracking and emphasizing five identity categories instead of three, mathematically you're creating more ways for people to be sorted into separate groups. The barriers multiply.
So the solution isn't to stop seeing identity—it's to create situations where people from different groups actually interact?
That's their argument. You can't legislate away how people evaluate traits. But you can change the structure so that people from different backgrounds are in the same spaces regularly. A mixed classroom, a diverse neighborhood initiative. Interaction itself becomes the intervention.
Does the model say which trait matters most?
It varies by context. In schools, ethnicity and grade level dominate. In marriage, it's age, gender, and ethnicity. But the researchers can now measure which dimension carries the most weight in any specific setting, which means policymakers can target their efforts where they'll have the most effect.