Do start-ups follow their industry? Sectoral bases and start-up location in southern Brazil
This study of 483 start-ups in southern Brazil finds that new ventures are significantly more likely to locate in municipalities with a strong existing productive base for their specific sector, whereas the local knowledge base (university enrolment) shows no independent association with start-up formation.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
New businesses do not appear randomly across a map. They tend to cluster in specific places, often following a familiar pattern: they gather in large, wealthy cities where people have money, education, and connections. However, beneath this general trend of big cities attracting everything, there is a more specific geography at work. Entrepreneurs often start companies in the same industries where they previously worked, near the factories or offices that trained them, and close to the customers they already know. This idea forms the backbone of modern innovation policy. Governments and regional planners often try to boost their local economies by identifying what their area is already good at—whether that is farming, medicine, or technology—and then pouring resources into those specific sectors. The assumption is that new ventures will naturally spring up where the existing tools, workers, and knowledge for that industry are strongest. But whether this assumption holds true in the real world, especially in developing economies, has rarely been tested directly.
A researcher at a university in southern Brazil set out to test this assumption by looking at the actual locations of new companies. The study focused on the state of Rio Grande do Sul, a region with a diverse economy ranging from vast agricultural lands to a major metropolitan hub. The researcher examined 483 registered start-ups across three distinct fields: health technology, agricultural technology, and information and communication technology. These companies were spread across 497 different towns and cities in the state. The goal was to see if these new businesses chose their locations based on two different types of local assets. The first asset is the "productive base," which refers to the physical and commercial infrastructure of an industry, such as the number of hospital beds, the amount of money generated by local farms, or the count of existing technology firms. The second asset is the "knowledge base," which refers to the presence of universities and the number of students enrolled in degree programs related to that specific field.
To find the answer, the researcher compared towns that were similar in size and wealth but had different strengths in these areas. For example, the study looked at whether a town with many hospitals but few tech companies had more health start-ups than tech start-ups, and vice versa for a town with many tech firms but few hospitals. By using a statistical method that accounted for general factors like population and income, the study isolated the effect of the specific industry assets. The results were clear and consistent. The study found that a town with a stronger productive base for a specific sector had significantly more start-ups in that same sector. Specifically, for every standard increase in the strength of a local industry's productive base, the number of start-ups in that field rose by 42 percent. This pattern held true even when the researchers removed the state's largest city from the data and even when they looked at broader definitions of these industries.
In contrast, the presence of universities and the number of students studying in a specific field did not show a similar independent effect. While universities are often seen as engines of innovation, the data suggested that simply having a campus with students in health, agriculture, or technology did not, on its own, predict where new start-ups would locate. The connection between a local university program and the birth of a new company in that field was not strong enough to be detected once the physical and commercial assets of the industry were taken into account. This finding challenges the common belief that building a university program is the primary way to spark a local tech or medical industry. Instead, the evidence suggests that new ventures are more closely tied to the existing industries they serve—the hospitals, the farms, and the established companies—than to the academic institutions that train their founders.
The study also revealed that these local patterns are deeply intertwined with the broader regions. When the researchers looked at the data on a larger scale, grouping towns into the state's eight regional innovation ecosystems, the specific link between a town's assets and its start-ups became less precise. This suggests that while a specific town might have a strong local industry, the decision to start a business there is also influenced by the wider regional specialization. For instance, agricultural start-ups were found to be the most spread out across the state, mirroring the wide distribution of farms, while health start-ups were heavily concentrated in the capital city, where the largest hospital complexes are located. This indicates that new businesses tend to follow the physical landscape of their industry, seeking out the users and customers they need to survive.
Ultimately, the research provides a clear guide for how places can support new businesses. It suggests that policies aimed at helping start-ups are most effective when they build upon existing industrial strengths rather than trying to create new ones from scratch. Connecting new ventures to the established hospitals, farms, and technology firms in a region appears to be a more reliable path to growth than focusing solely on academic enrollment. The findings indicate that in southern Brazil, and likely in similar emerging economies, the geography of innovation is driven more by the tangible presence of an industry and its customers than by the proximity of a university campus. This insight helps explain why new companies often appear where the work is already being done, following the path of least resistance to the resources and networks they need to succeed.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.