How Agglomeration and Scale Economies Shape Urban Growth

Cities don’t just happen by accident. There’s a powerful economic logic behind why millions of people and thousands of businesses crowd into dense urban areas, paying premium rents and enduring traffic jams. That logic comes down to two interrelated concepts: agglomeration economies and scale economies. Together, they explain why cities form, why they grow, and ultimately, why that growth has limits. Understanding these forces is essential for anyone studying how urban environments evolve and how economic policy can shape the cities of the future.

What are agglomeration economies? 🔗

Agglomeration economies refer to the cost advantages and productivity gains that businesses and workers enjoy when they locate near one another in urban areas. The concept is straightforward: when economic activity clusters in one place, it creates benefits that wouldn’t exist if the same firms and workers were spread out across the landscape.

As Edward Glaeser of Harvard University explains, these benefits fundamentally stem from reduced transportation costs – interpreted broadly to include the movement of goods, people, and ideas. The closer firms are to each other, the easier and cheaper it becomes to exchange inputs, hire talent, and share knowledge.

This is precisely why cities exist. If there were no advantages to proximity, economic activity would be evenly distributed. Instead, we see concentrated pockets of production and employment that attract even more businesses and workers, creating a self-reinforcing cycle of growth.

Three core mechanisms behind agglomeration 🔗

Urban economists have identified three main channels through which agglomeration economies operate:

Sharing: Firms in cities can share specialised suppliers, infrastructure, and services that would not be viable in smaller, more isolated markets. A single legal firm specialising in patent law, for instance, might struggle in a small town – but in a major metro area with hundreds of tech firms, it thrives. This sharing of intermediate inputs and suppliers lowers costs for everyone involved.

Matching: Dense urban labour markets make it easier for employers to find workers with the right skills, and for workers to find jobs suited to their expertise. This improved matching efficiency means less time and money spent on recruitment, and better outcomes for both parties. Studies have shown that industries in metropolitan areas benefit significantly from access to shared labour pools, particularly in specialised production-related occupations.

Learning: Perhaps the most important channel in today’s knowledge economy is the flow of ideas. When skilled professionals work in close proximity, knowledge spills over between firms and industries. Research from the OECD shows that comparable workers are measurably more productive in larger cities, partly because of these knowledge spillovers and learning effects. A worker in a city of 6 million inhabitants can be around 15% more productive than a similar worker in a city of 120,000.

Types of scale economies: internal vs. external 🔗

To fully understand urban growth, you need to distinguish between two types of scale economies. Both contribute to why businesses cluster in cities, but they work through different mechanisms.

Internal economies of scale 🔗

Internal economies of scale occur within a single firm as it grows larger. When a factory doubles its production, it doesn’t necessarily need to double all its inputs. Fixed costs like machinery, management overhead, and research get spread over more units of output, reducing the average cost per unit.

The automobile industry is a classic example. Car manufacturing requires enormous upfront investment in assembly lines, robotics, and tooling. A small-scale car maker can’t compete on price because those fixed costs are spread over too few vehicles. This is why automotive production has historically concentrated in specific cities – Detroit in the United States, Stuttgart in Germany, and Toyota City in Japan – where manufacturers can achieve the volume needed to bring down per-unit costs.

External economies of scale 🔗

External economies of scale arise not from the growth of a single firm, but from the growth of an industry or an entire city. These are the benefits that spill over to all firms in a cluster, regardless of their individual size. External economies are further divided into two categories:

Localisation economies benefit firms within the same industry. When many tech companies concentrate in one region, they collectively support a deep pool of software engineers, specialised venture capital firms, and niche component suppliers. No single company created these resources, but every company in the cluster benefits from them.

Urbanisation economies benefit firms across all industries in a city. A large, diverse city offers shared infrastructure – airports, broadband networks, universities, hospitals – that supports businesses in finance, healthcare, manufacturing, and technology alike. The sheer diversity of a major urban economy makes it more resilient to downturns in any single sector.

Silicon Valley: the textbook case of agglomeration 🔗

No discussion of agglomeration economies is complete without Silicon Valley. The San Francisco Bay Area represents the most studied and most cited example of how clustering drives innovation and economic growth.

The story begins with Stanford University and its role in seeding technology spinoffs in the surrounding area. Early semiconductor firms took root in Santa Clara County, and from there a self-reinforcing ecosystem emerged. Venture capitalists set up nearby to fund startups. Skilled engineers migrated to the region for job opportunities. Law firms, accountants, and recruiters specialising in tech followed.

Today, the Silicon Valley ecosystem extends across the entire Bay Area and encompasses computing, electronics, software, biotechnology, and artificial intelligence. The agglomeration advantages are substantial: companies benefit from proximity to world-class research universities, access to deep venture capital networks, and a massive pool of specialised talent. As a Harvard Business School working paper on tech clusters notes, the agglomeration advantages created by early movers like Microsoft in Seattle attracted countless other technology firms to the same region, including Amazon.

But Silicon Valley also demonstrates the costs of agglomeration. Housing prices in the Bay Area are among the highest in the world. Traffic congestion is severe. Income inequality within the region is stark. Many middle-income workers have been priced out entirely, forced to commute long distances or relocate. These are the diseconomies that inevitably accompany intense urban concentration.

The automotive cluster: a different model 🔗

While Silicon Valley showcases knowledge-driven agglomeration, automotive manufacturing illustrates a more traditional form of industrial clustering based on supply chain proximity and specialised labour.

Detroit’s rise in the early 20th century was driven by localisation economies. Once Henry Ford established mass production there, suppliers of steel, glass, rubber, and components set up nearby. Workers with automotive skills concentrated in the region. Training programmes and technical schools emerged to feed the industry’s labour needs.

The pattern has repeated globally. Germany’s automotive industry clusters around Stuttgart (Mercedes-Benz, Porsche) and Munich (BMW), while Japan’s centres on Toyota City and the broader Nagoya region. In each case, the concentration of the industry in one area creates efficiencies that individual firms couldn’t achieve in isolation – shared suppliers, specialised logistics networks, and a workforce with deep domain expertise.

However, Detroit also serves as a cautionary tale. Over-reliance on a single industry made the city vulnerable when global competition and technological shifts disrupted the automotive sector. A cluster’s strength can become its weakness if it fails to diversify.

Optimal city size and the limits of growth 🔗

If agglomeration economies are so powerful, why don’t cities just keep growing forever? The answer lies in the concept of optimal city size – the point at which the benefits of further growth are exactly offset by rising costs.

A Harvard Kennedy School policy brief by Gomez-Ibañez and Ruiz Núñez explains this using a framework originally developed by economist William Alonso. As a city grows, the incremental benefits of agglomeration – better matching, more suppliers, greater knowledge flows – start to decline. At the same time, incremental costs – congestion, pollution, rising rents, strained infrastructure – begin to rise. The optimal city size sits at the intersection of these two curves.

The benefits side 🔗

Larger cities deliver clear productivity advantages. Workers earn higher wages, firms generate more output per employee, and innovation rates are higher. Research consistently shows that doubling a city’s population increases average productivity by roughly 3-8%. Larger cities also produce more patents per capita and see faster adoption of new technologies. Their economic diversity provides resilience – when one industry contracts, others can absorb displaced workers.

The costs side 🔗

But these benefits come with mounting costs. Traffic congestion is the most visible diseconomy. As more workers commute into a growing city centre, travel times increase and productivity is lost. According to research on urban public transit systems, when demand exceeds certain thresholds, the social costs of congestion can escalate rapidly, even overwhelming the economies of scale that public transport normally provides.

Housing affordability is another major cost. Successful cities attract more workers, driving up demand for limited real estate. Cities like San Francisco, London, and Mumbai have seen housing costs spiral to levels that effectively price out essential workers. Mumbai, for instance, has office rents that, when adjusted for purchasing power, are among the highest in the world – largely due to restrictive land-use policies that constrain supply.

Environmental pressures also intensify with city size. Larger cities generate more pollution, consume more resources, and produce more waste. Infrastructure – roads, utilities, schools, hospitals – comes under increasing strain, and the cost of upgrading these systems can become prohibitive.

Why “optimal” is not one-size-fits-all 🔗

The optimal city size is not a fixed number. It varies depending on industry, geography, governance, and the quality of infrastructure. A technology hub like San Francisco may sustain a larger population because the knowledge spillovers are so valuable. A manufacturing city might reach its optimal size sooner because the agglomeration benefits plateau once supply chains are established.

Policy choices play a huge role. Cities that invest in efficient public transport, allow higher-density construction, and manage environmental impacts can push their optimal size higher. Cities with poor planning, restrictive zoning, and underinvestment in infrastructure will hit their limits sooner. As research on compact cities suggests, building taller and denser city centres through urban regeneration can weaken the negative effects of population growth on congestion, effectively extending the benefits of agglomeration.

When agglomeration becomes diseconomy 🔗

The flip side of agglomeration economies – diseconomies of agglomeration – deserve attention because they determine when and why urban growth slows or reverses.

Diseconomies include not just the tangible costs of congestion and pollution but also subtler effects. Rising wages and rents increase the cost of doing business, making it harder for startups and small firms to compete. Crime may increase in overcrowded areas. Social inequality widens as high-income professionals drive up the cost of living for service workers.

These diseconomies create a natural tension that prevents cities from growing without limit. When the costs become too great, firms and workers start relocating to smaller cities or suburban areas where they can still access some regional agglomeration benefits at lower cost. This dynamic explains the emergence of polycentric urban regions – metropolitan areas with multiple centres of economic activity, each capturing different agglomeration advantages while avoiding the worst congestion of a single mega-centre.

The role of policy in shaping urban outcomes 🔗

Understanding agglomeration and scale economies isn’t just an academic exercise. These concepts have direct implications for urban policy.

Governments that recognise agglomeration forces can make smarter investments. Building transport infrastructure that connects suburban areas to city centres extends the effective labour market, boosting matching efficiency. Investing in universities and research institutions can strengthen knowledge spillovers. Relaxing overly restrictive zoning can increase housing supply and keep cities affordable enough to attract diverse talent.

Conversely, policies that ignore these forces – or that attempt to artificially disperse economic activity – often backfire. Forcing businesses to relocate to underdeveloped regions may seem equitable, but if those regions lack the critical mass of firms and workers needed for agglomeration economies, the relocated businesses may simply become less productive.

The most effective approach is to support a system of cities with different sizes and specialisations. Mega-cities serve as hubs for innovation and high-value services. Mid-sized cities can capture manufacturing and logistics advantages. Smaller cities can specialise in niche industries. Each city in the system plays a role, and smart policy helps each reach and maintain its optimal size.

What do you think? Given the rising costs of living in major cities worldwide, is the era of mega-city agglomeration reaching its limits – or will technology and better urban planning continue to push those limits further? And as remote work reshapes where people choose to live, how might the traditional advantages of agglomeration evolve in the coming decades?

References

1. https://www.nber.org/system/files/chapters/c7977/c7977.pdf
2. https://en.wikipedia.org/wiki/Economies_of_agglomeration
3. https://research.upjohn.org/cgi/viewcontent.cgi?article=1256&context=reports
4. https://ideas.repec.org/a/sls/ipmsls/v32y20179.html
5. https://carnegieendowment.org/russia-eurasia/research/2024/01/the-silicon-valley-model-and-technological-trajectories-in-context
6. https://www.hbs.edu/ris/Publication%20Files/20-063_97e5ef89-c027-4e95-a462-21238104e0c8.pdf
7. https://www.hks.harvard.edu/sites/default/files/centers/taubman/files/inefficientcities_final.pdf
8. https://www.sciencedirect.com/science/article/abs/pii/S0047272723000853
9. https://www.sciencedirect.com/science/article/abs/pii/S0264999322000748