Digitalization reinforces, not reduces, urban inequality

For more than two decades, the digital economy has often been described as weightless, borderless, and detached from physical space. Popular thinking assumes that digital firms can operate from anywhere due to low communication costs, high connectivity, and virtual workflows. Under this view, digitalization should weaken the historic pull of big cities, soften regional divides, and spread economic opportunity across more places.

A new academic investigation published in EPB: Urban Analytics and City Science challenges one of the most common assumptions of the modern era: that the digital economy operates outside the boundaries of geography. The authors argue that even the most intangible parts of the economy remain shaped by physical urban structures, infrastructure, and the spatial pull of city ecosystems.

Their study, titled Not so “placeless” after all: understanding the spatial implications of the digital economy, presents clear evidence of how digital firms actually locate themselves within a major global city. By building a new dataset of London-based companies using website text and natural language processing, the researchers uncover clear patterns of spatial clustering. These patterns challenge the idea that digitalization dissolves the importance of place.

How the digital economy has been misunderstood

For more than two decades, the digital economy has often been described as weightless, borderless, and detached from physical space. Popular thinking assumes that digital firms can operate from anywhere due to low communication costs, high connectivity, and virtual workflows. Under this view, digitalization should weaken the historic pull of big cities, soften regional divides, and spread economic opportunity across more places.

The researchers argue that this expectation has persisted partly because digital firms are not well captured by traditional data sources. Official economic classifications struggle to identify new tech-based activities, leaving many firms hidden in broad industrial categories. Without clear data, arguments for a placeless digital economy have been difficult to test.

This study introduces a novel way to identify digital firms by analyzing the language used on their websites. Instead of relying on classification codes that lag behind emerging sectors, the approach uses natural language processing to detect digital characteristics within website text. The authors apply this method to firms located across Greater London, creating the most detailed map yet of the city’s digital sector.

This dataset allows the researchers to measure the spatial footprint of digital firms with greater accuracy than previous studies. It also allows them to compare digital firms with non-digital firms while accounting for the urban environment in which they operate. The result is a clearer picture of whether digital activity truly escapes geography or remains tied to it.

How the authors measured the geography of the digital economy

After identifying digital and non-digital firms from website text, the authors analyzed how these firms are spread across London. Their method treats firms as points in space and models them as outcomes of a spatial point process. This approach allows them to break down the factors that influence where firms locate.

They focus on two categories of spatial drivers. The first set includes background characteristics of the urban environment. These factors include transport links, accessibility, land use patterns, and neighbourhood-level features that affect the attractiveness of a location. If digital firms cluster only because these areas are generally appealing for business, then their spatial pattern could be explained without referring to any special digital dynamics.

The second set includes firm-to-firm interaction forces. These forces are called endogenous factors. They reflect benefits that firms gain from being near one another, such as access to talent, knowledge exchange, and face-to-face interaction. If digital firms cluster more strongly than the background environment predicts, the effect must come from these internal agglomeration forces.

To distinguish these effects, the authors use an inhomogeneous K-function and an interaction-based extension. This model tests whether digital firms cluster beyond what would be expected from urban infrastructure and land characteristics alone. The approach provides a rigorous way to determine whether the digital sector is shaped by spatial proximity in the same way as traditional industries.

This methodological contribution is one of the most important aspects of the study. It opens the door to future research that can use website text to identify digital firms in other regions, improving the quality of digital economy data and offering policymakers better tools to track technological sectors.

What the findings reveal about the “placeless” digital economy

The results show that digital firms in London are highly clustered. The clustering remains even after adjusting for background urban features, meaning that the spatial pattern cannot be explained simply by the general attractiveness of certain neighborhoods. Instead, the study finds that digital firms show significant excess clustering at distances larger than a single street or block. These patterns match what is known about traditional agglomeration forces, where firms benefit from being located close to others in the same sector.

The authors argue that this contradicts the claim that digitalization weakens the role of place. Instead, the digital economy depends on many of the same spatial forces that shape older industries. Firms still seek locations where they can recruit skilled workers, share knowledge informally, and take advantage of dense networks of related businesses. Proximity still matters for innovation, collaboration, and access to clients.

In London, this clustering supports the idea of a layered urban economy. Digital firms do not spread evenly but instead concentrate in areas where existing businesses, strong infrastructure, and abundant talent already exist. These findings challenge the assumption that digital work will naturally decentralize economic activity or reduce dependence on urban centers.

The study suggests that rather than creating a frictionless global space, digital tools may actually increase the advantages of already successful urban nodes. Cities with strong digital infrastructure, skilled labor markets, and well-established innovation ecosystems attract even more firms, while areas without these advantages risk being left behind.

This dynamic reinforces the importance of understanding digital clustering patterns. Without a realistic view of how digital industries behave, policymakers may misjudge the scale of regional disparities or the potential impact of economic development programs.

What the study means for policymakers and regional planners

Many governments promote digitalization as a way to boost lagging regions. However, the study cautions that digital activity does not automatically disperse across space. Instead, it often gravitates toward places with strong networks, good infrastructure, and established innovation clusters.

The authors argue that expecting digitalization to balance regional inequality is unrealistic without targeted intervention. Effective policy must account for the spatial context in which digital firms operate. This requires investment in skills, connectivity, and infrastructure in lagging areas, as well as support for local innovation ecosystems that can compete with established centers.

The study also calls for better data. Because the digital economy evolves faster than official classifications, many regions lack accurate information about their digital sectors. The authors show that website text analysis can help close this gap, offering more precise insights into where digital firms are located and how they grow.

Regional planners may also need to rethink assumptions about remote work and digital mobility. The clustering documented in the study suggests that face-to-face interaction remains vital for digital activity. This means that cities will continue to play a central role in technological innovation, even as remote tools expand.

Overall, the study calls for more realistic expectations about the spatial effects of digitalization. Instead of assuming digital activity will spread evenly, policymakers must consider how physical infrastructure and spatial networks shape the digital economy. Regions that fail to adapt risk falling further behind as global competition intensifies.