Innovation does not emerge only where research and development spending is high. In regions where people can attain more education than their parents and pursue new careers, previously hidden talent also finds its way into laboratories and companies. A study tracking educational mobility and patents across 36 European countries shows how equality of opportunity can translate into new ideas.

* Equality of opportunity forms part of the foundation for innovation. When parental education strongly determines children’s educational and professional outcomes, substantial talent never reaches laboratories or companies. Increasing mobility is therefore more than a social policy; it is a growth strategy that helps society use its talent more efficiently.

* Educational expansion is not the same as greater mobility. Even when more people graduate from university, relative mobility may remain unchanged if children of less-educated parents continue to occupy the lower ranks of the educational distribution. Policymakers must examine who gains access to which education and occupations, not merely the average level of schooling.

* Regions with greater mobility also displayed stronger patenting activity. Across European regions, a 0.1 decline in intergenerational educational persistence was associated with an increase of approximately 4% to 23% in patent counts. The overall direction remained consistent across 420 additional estimates using different measures and samples.

* The greatest untapped innovation potential may lie in regions where opportunity has been restricted. The relationship between improved mobility and innovation was strongest in regions with low or moderate initial mobility. Removing a first major barrier can generate a larger change than adding another opportunity where access is already broad.

* The social ladder is economic infrastructure for moving talent. Schools, scholarships, regional universities, career information and open recruitment do more than support individual advancement. They create a route through which ideas hidden by family circumstances can move into education, research, entrepreneurship and patents.

***

Imagine two children with the potential to become inventors. The first child’s parents are university graduates, and the home is filled with books and computers. Through the parents’ acquaintances, the child meets engineers and researchers and attends science camps during school breaks. Advice on choosing a school and university major is readily available, and even if the child fails once, the family has the time and financial means to support another attempt.

The second child also enjoys taking machines apart and putting them back together. The child builds toys using the motor from a broken electric fan and constantly wonders how smartphones work. Yet the parents have no experience of attending university, and there are no researchers or engineers in the family’s social circle. Information about higher education is difficult to obtain, and even if the child enters university, tuition and living expenses remain formidable burdens. If the child must begin working early to support the family, a career requiring many years of education and training is likely to disappear from the list of realistic options.

Although the two children may have similar potential, they can arrive at very different places as adults. One may enter a laboratory or technology company, while the other may remain in a job that does not make full use of that potential. What disappears in this process is not merely an individual opportunity. Inventions that are never introduced, companies that are never founded and problems that remain unsolved disappear with it.

Explanations of innovation usually begin with R&D expenditure, venture capital, prestigious universities and industrial clusters. All of these factors matter. Yet even the finest research institute can use only part of society’s available talent if its doors are open exclusively to people from particular backgrounds. In regions where parental education and economic status strongly determine children’s educational attainment, family background functions as a mechanism for selecting talent more powerfully than ability does.

Economists describe this as the “misallocation of talent.” People capable of producing valuable ideas fail to obtain educational and professional opportunities, while others secure those positions partly because they were born into more advantaged families. It is unfair to individuals, a loss of talent for companies and a loss of productivity and innovation for regional economies. Inequality of opportunity is not merely a moral concern. It is a structural cost that prevents an economy from making full use of its capabilities.

The relationship was examined in depth in “Intergenerational Mobility Fosters Innovation in Europe,” published in the international journal Nature in July 2026. Sarah McNamara, Guido Neidhofer and Patrick Lehnert analysed how intergenerational educational mobility had changed across European regions and over time, and how those differences were related to innovation performance.

The study began with a simple but difficult question: how strongly does parental background hold on to a child’s future? National averages for educational attainment could not provide a sufficient answer. Even if more people were graduating from university, the underlying opportunity structure might have changed very little if children of highly educated parents continued to dominate access to better education and occupations. Conditions could also differ sharply among capitals, industrial cities, rural communities and declining regions, even when their national averages were similar.

To investigate the question, the researchers constructed the European Atlas of Spatially Disaggregated Intergenerational Mobility, or EUROPE-IGM-ATLAS. They combined information about respondents’ and parents’ education from 11 rounds of the European Social Survey conducted between 2002 and 2023. They then harmonized the information to make comparisons possible across countries and periods. Intergenerational educational mobility was measured by calculating the strength of the relationship between parents’ years of schooling and those of their children.

If parental education almost completely determines a child’s educational attainment, intergenerational persistence is high and mobility is low. If children can receive an education suited to their abilities even when their parents had little schooling, persistence is low and mobility is high. Mobility in this context does not simply mean that more people have earned university degrees. It describes the extent to which individuals can move beyond the educational circumstances of the families into which they were born.

The scope of the analysis was extensive. The researchers divided 36 European countries into 225 microregions or 108 mesoregions and constructed annual indicators covering the period from 1985 to 2025. They compared three birth cohorts: people born between 1940 and 1959, between 1960 and 1979, and between 1980 and 1999. This enabled them to track, at the regional level, how much education people born in different periods obtained relative to their parents and whether they changed their position within the educational distribution.

The researchers also carefully considered when an individual’s educational attainment would affect economic performance. If a person born in 1950 entered the labour market at around the age of 20 and remained economically active for several decades, that person’s educational opportunities would affect the regional economy over a long period rather than in a single year. The team therefore transformed the cohort-level data into annual mobility indicators by weighting when each generation entered the labour market, accumulated experience and eventually retired. In effect, they turned what had resembled a still photograph of educational mobility into a moving picture.

Next, they placed a map of innovation over this map of opportunity. Innovation performance was measured not through surveys or companies’ subjective assessments, but through regional patent applications, granted patents and citation-weighted patents. Patents do not capture every form of innovation, but they offer a representative indicator that can be compared across countries and regions over long periods.

The innovation analysis used 233,057 observations from 137 regions. Regions formerly belonging to the Warsaw Pact were followed from 1992 to 2020 to account for the effects of political and economic transition, while the remaining regions were studied from 1985 to 2020. The analysis also controlled for factors that could influence patenting, including average educational attainment, educational inequality, regional characteristics, common changes over time and country-specific trends.

The researchers tested whether the findings remained consistent when the measurement methods and samples were changed. They replaced patent applications with granted patents, used citation-weighted patents instead of simple patent counts and varied both the construction of the mobility indicators and the treatment of migrants in the sample. In total, they conducted 420 additional estimates. Around 89% showed a statistically significant relationship between lower intergenerational persistence?or higher educational mobility?and greater patenting activity.

The repetition of the same overall pattern under different analytical conditions was one of the study’s strengths. It also differed from earlier studies by examining both changes over several decades and regional variation within countries instead of treating Europe as a single unit. Nevertheless, because the analysis relied on observational data, it could not prove that greater mobility had unilaterally caused an increase in patents. Unobserved regional characteristics may have influenced both. What the study demonstrated was a strong and robust relationship in which regions with higher intergenerational educational mobility also consistently produced stronger innovation outcomes.

Educational opportunities across Europe clearly expanded over recent decades. Younger generations generally received more education than their parents, with particularly notable improvements among children from families at the lower end of the educational distribution. The indicator measuring how closely children’s years of schooling resembled those of their parents fell on average from 0.459 in the oldest cohort to 0.355 in the youngest.

At first glance, the social ladder appeared to have moved considerably. Yet the figures contained an important complication. The absolute amount of education increased, but people’s relative positions within society did not change to the same extent.

Consider a town in which most members of the parents’ generation finished only middle school, while most of their children completed high school. The overall level of education has risen. But if children from less-educated families remain near the bottom of their generation’s educational distribution, while children of university-educated parents continue to occupy the top, relative mobility has barely improved. The entire staircase has moved up one floor, but the order in which people stand on it has remained much the same.

The standardized measure reflecting changes in people’s relative positions within the educational distribution showed only a slight decline in intergenerational persistence, from 0.415 to 0.400. The change was not statistically significant. An increase in university attendance did not automatically eliminate the influence of family background.

As degrees become more common, the type of degree, the prestige of the university, the field of study, postgraduate education, internship experience and professional networks can emerge as new dividing lines. Two people may both graduate from university, but one may use information and relationships accumulated since childhood to enter a research profession, while the other learns only belatedly that such a door exists. Educational opportunity therefore cannot be assessed simply by counting diplomas. It is also necessary to examine who receives which kind of education and where that education allows them to go.

The findings challenge the optimistic belief that inequality naturally declines as average educational attainment rises. An increase in years of schooling is an important achievement, but it is not synonymous with equality of opportunity. Even when everyone has climbed higher than the previous generation, a structure in which parental background determines children’s relative positions can remain intact.

When the researchers overlaid the mobility and patent maps, a consistent pattern emerged. Patent activity was lower in regions where parents’ and children’s education remained closely linked and higher where that connection was weaker. The relationship persisted even after controlling for average education, educational inequality, regional and temporal differences and country-specific trends.

Its magnitude was not trivial. A 0.1 decline in the persistence indicator measuring the similarity between parents’ and children’s education was associated with an increase of approximately 4% to 23% in the number of patents. A change of 0.1 was close to the average shift observed between Europe’s oldest and youngest cohorts.

The scale becomes easier to understand when translated into a hypothetical region. Suppose a region had previously produced 100 patents a year and the influence of parental education declined by an amount comparable to the change actually observed across European generations. A simple application of the study’s association would place patenting somewhere between approximately 104 and 123. This should not be treated as a predictive formula that applies uniformly to every region. Universities, companies, industrial structures, R&D investment and migration also influence patent production. Nevertheless, the relationship between opportunity structures and innovation performance was large enough to be economically meaningful.

The connection between the two can be understood through the use of talent. Innovation resembles a long selection process. A child must first become interested in science, receive adequate instruction, enter higher education, acquire research experience and then join a technology company or establish a business. If family background creates inequality at every doorway along that path, the pool of people who finally reach a laboratory becomes progressively narrower.

Scholarships, public education, regional universities, vocational training and accessible career information can reduce the disadvantages associated with family background. More people are then able to participate in the process, widening the pool of candidates who reach the entrance to the innovation system. Problems arising from different lives and experiences also enter research and development. Society is not suddenly producing more geniuses. It is simply losing fewer talented people who were already there but had previously remained undiscovered.

The distinction between mobility and average educational attainment is critical. A region with many university graduates is not necessarily highly mobile. If only children of highly educated parents enter prestigious universities and research careers, average education may be high while the pool from which talent is selected remains narrow. When people can obtain educational opportunities regardless of parental education, inventors and entrepreneurs can emerge from a much wider range of backgrounds.

The ability to develop a new battery does not suddenly appear when someone enters a university laboratory. It is built over many years through childhood curiosity, opportunities to learn mathematics and science, the freedom to try again after failure, role models with similar backgrounds, and teachers or mentors who explain how to enter technical professions.

Imagine a student growing up in a region where mines or factories have closed. That student may have a remarkable intuition for energy technology and understand better than most how to improve the efficiency of an ageing boiler or reduce wasted electricity. Yet if the local school offers no advanced science courses, the nearest university is far away and the student must help cover the family’s living expenses, that insight may never become a patent. The seed of innovation existed, but there was no path along which it could grow.

The opposite is also possible. Even when neither parent attended university, a student’s range of possibilities can expand through a dedicated science teacher, an experimental programme offered by a regional university and financial support for tuition and living costs. The student may enter a technology company through a university?industry partnership and develop a new process based on a problem encountered at work. A highly mobile region turns such possibilities from matters of personal luck into institutional pathways.

Exposure also matters in innovation. For a child who has never met an inventor, inventing may feel less like a realistic career than a role in a film. If researchers live nearby or schools are connected to universities and companies, innovation becomes a visible option. Even the capacity to imagine what one might become is shaped by the surrounding environment.

Parental education produces differences not only in money but also in information and expectations. An invisible information gap separates families that know which universities to consider, where to find scholarships, how to obtain a research internship and what to do after failure from those that do not. Two students with identical grades may face completely different sets of choices: one compares multiple routes, while the other sees only the nearest available path.

Inequality of opportunity is therefore broader than a gap in examination scores. It encompasses information, expectations, relationships and the ability to absorb failure. Lowering tuition alone does not remove every barrier, just as increasing university admissions does not immediately diversify the population of researchers and inventors. The entire sequence of thresholds connecting education to employment must be examined.

The same insight applies to companies. Businesses that recruit repeatedly from the same universities, career backgrounds and personal networks may believe they are selecting proven talent. In practice, they can mistake people who received early opportunities for those with the greatest ability. Widening recruitment channels, evaluating potential and providing training after hiring are more than acts of corporate citizenship. They are innovation strategies that allow companies to draw from a larger pool of talent.

The relationship between educational mobility and patenting was not identical across every European region. In major innovation hubs that remained within the top 10% of the patent distribution over long periods, the relationship was weaker and, in some analyses, moved in the opposite direction. At first glance, this appeared to conflict with the broader finding that equality of opportunity supports innovation.

The difference becomes more understandable when the workings of an established innovation hub are considered. Such regions contain universities, research institutes, investors, skilled workers, large corporations and start-ups accumulated over many decades. Engineers move from one company to another, researchers establish businesses, and information is exchanged at conferences and through informal encounters. Trained workers continually arrive from outside, while professional services and capital markets are already well developed. Innovation is shaped not only by educational mobility within the region but also by accumulated agglomeration effects and dense knowledge networks.

Even when a low-income child born in a globally successful technology city continues to face limited opportunities, that city can maintain a strong patent record by attracting people who were educated elsewhere. Its innovation score may be impressive, while the doors of opportunity remain closed to many local residents. A high patent count and equality of opportunity do not always point in the same direction.

The exceptionally high productivity of top innovators may also matter. When a small number of inventors and scientists account for a disproportionate share of patents and research output, changes in the average talent pool may not immediately appear in patent totals. In regions that already possess powerful research organizations and leading inventors, the concentration of elite talent and the circulation of knowledge may have a greater effect on short-term outcomes than intergenerational mobility does.

This does not mean that equality of opportunity is unimportant in innovation hubs. Instead, it raises questions that patent counts alone cannot answer. Who participates in innovation? Which problems are selected for investigation? Do local residents share in the benefits of growth? A city that prospers mainly by importing trained workers may preserve high output, but when its internal social ladder breaks down, it can face rising housing costs, social separation and a fragile local talent base.

The benefits associated with improved mobility were especially pronounced in regions where initial mobility was low or moderate. Removing the first major barrier where many people have been excluded can produce a larger change than opening one additional door where opportunity is already broad. Regions with restricted educational opportunity may contain more dormant talent and therefore stand to gain more innovation when that talent is activated.

Policies intended to increase social mobility are often classified as redistributive measures. Scholarships, investment in public education, support for regional universities and career programmes for low-income students are frequently discussed as questions of how to divide wealth that has already been created. The study redirects the question. Expanding opportunity may distribute wealth more fairly, but it can also help produce new wealth and ideas.

Roads allow people and goods to move, while communications networks allow information to flow. Education and mobility are forms of infrastructure that allow talent to move. An economy can make full use of its human capital only when people with valuable ideas can travel from their family backgrounds into schools, universities, laboratories and companies. A broken social ladder does more than prevent individual advancement. It severs the route through which ideas reach productive activity.

Policy must therefore look beyond university enrolment rates. It should examine whether children have access to quality education from an early age, whether school quality differs excessively by region, whether students whose parents did not attend university can still obtain information about education and careers, and whether living costs force people to abandon their studies. After university, fair access to research experience, internships, start-up capital and professional networks also matters.

Regional growth policy cannot be completed with buildings and equipment alone. Alongside research parks and laboratories, regions need long educational pathways that enable local students to work in those facilities. Primary and secondary schools, vocational institutions, universities and companies must be connected. First-generation university students need access to information and mentors, while social protections should allow people to try again after failure. Otherwise, a new research complex can become an island separated from the surrounding community and dependent entirely on imported talent.

Companies can be both beneficiaries of and participants in social mobility. Recruitment that moves beyond narrow educational and career credentials, paid internships, partnerships with regional universities and training after employment can connect a wider range of people to the innovation process. People from different backgrounds may question assumptions that established organizations take for granted and identify the needs of customers and markets that have previously been overlooked.

The maps of opportunity and patents across Europe convey a simple but sharp message: the opposite of innovation may not be failure, but waste. Students who abandon education because of the families into which they were born, young people who never consider research careers because they lack information, and technicians who cannot test their ideas because they cannot afford to do so are never recorded statistically as inventors. Patents that never exist and companies that are never founded are invisible, making their loss easy to forget.

A society seeking an advantage in the innovation race does not need to wait for more geniuses to be born. It needs to open routes that allow talent already present in every community to move beyond parental education, income and postcode. Repairing the social ladder does more than lift an individual. It also brings that person’s unrealized ideas into the world.

Nature, July 2026, Sarah McNamara, Guido Neidhofer and Patrick Lehnert, “Intergenerational Mobility Fosters Innovation in Europe”