Why Opportunity Clusters
The local labor market is the mechanism that turns geography into a career opportunity set. This chapter asks why productive firms, skilled workers, and high-value jobs become concentrated in particular places, and why that concentration can create a thicker career ladder for some residents than for others. The aim is not to decide whether cities or rural areas are universally better. It is to trace the connections among firms, workers, infrastructure, services, housing, and job quality so that a reader can tell what a location actually offers and what it merely appears to offer from a distance.
A location does not hand someone a career. It changes the number, variety, and accessibility of routes available to them. Those routes can include entry-level work, opportunities to learn from experienced colleagues, employers willing to pay for scarce capabilities, and adjacent jobs that make a change of role possible without starting over. They can also be narrowed by high housing costs, poor transport, limited care, or a thin set of employers. The same place can therefore be highly productive in aggregate and difficult for a particular resident to use.
How Firms and Workers Make a Labor Market Thicker
A career ladder is easier to climb when it has more than one rung at each stage. In a thin local market, a worker may have only a few employers, a narrow industry mix, and little room to change roles without moving. In a thicker market, several firms may need related capabilities. A worker can gain experience in one organization, carry that experience to another, and find a better fit without abandoning the region. Employers, in turn, have more ways to recruit and compare candidates. The result is not guaranteed advancement, but a larger set of possible matches.
The OECD describes three connected agglomeration mechanisms behind this thickness. First, a concentration of firms can speed employer-worker matching. More employers searching for similar capabilities and more workers looking for related roles increase the chance that a suitable match exists. Second, proximity can support knowledge transfers between firms. Workers observe practices, carry experience across organizational boundaries, and build skills that are useful beyond their first job. Third, dense supplier and client networks can create demand for complementary businesses and occupations. These mechanisms help explain why high-productivity firms and innovation are not distributed evenly across a country.
The important word is connected. A group of firms does not create opportunity simply by occupying the same map. It creates more opportunity when workers can reach those firms, when firms can find and retain people, and when institutions supply the skills, infrastructure, and services that let the relationships persist. Agglomeration is therefore a system of interfaces: workers meet employers through hiring and job search; firms meet suppliers and customers through local networks; and institutions connect both sides through education, transport, digital infrastructure, and public services.
High-value employment is one visible expression of that system. Across the countries with available data, the OECD finds that high-value-added services account for an average of 26.6% of employment in capital-city regions, compared with 17.8% in other metropolitan regions and 12.5% in non-metropolitan regions. The comparison is about the composition of jobs, not a verdict on the worth of every job or every region. It indicates that capital regions are more likely to contain the kinds of firms and activities associated with high value added, and therefore may offer a different range of career entry points and progression routes.
The value of that concentration is not just a higher starting wage. It can change the risk of a career decision. In a narrow local market, accepting the wrong first job may leave a worker with little nearby alternative; changing employers may require a move, and a move may require abandoning a partner’s job, a care arrangement, or affordable housing. In a thicker market, a worker may still make a poor choice, but there are more chances to correct it without losing the connections that make work possible. The existence of alternatives can also change bargaining power: an employer is not the only local buyer of a capability when several firms compete for related workers. These are mechanisms that can widen option value, not guarantees that every resident receives a high-quality opportunity.
The same logic applies to employers. A productive firm benefits from a reliable pool of workers, but it also depends on suppliers, customers, transport links, and institutions that reduce the cost of coordinating complex work. A regional cluster can therefore support jobs that would be difficult to sustain in isolation. Yet concentration has a cost: when too much employment depends on one industry or a small group of firms, the local market becomes exposed to a common shock. The career ladder is thicker while the industry is growing, but its rungs can disappear together when demand changes. Opportunity is thus shaped by both density and diversity.
That difference can compound. If a place has more productive firms, those firms may pay for or attract specialized skills. A deeper pool of workers can make it easier for another productive firm to establish itself. The arrival of another firm can create further demand for services, suppliers, and experienced workers. More people may then find it worthwhile to move there, or may remain because changing employers locally is feasible. Over time, the local labor market can become more attractive precisely because it already offers a dense set of options.
The reverse is also possible. When productive firms leave a lagging region, the remaining market may contain fewer employers and less varied work. A worker who loses a job has fewer nearby alternatives, and a young person deciding where to begin may see less reason to stay. The outflow of people with more education or stronger employment prospects can reduce the skills base available to local firms. This does not mean that a region’s residents are the cause of its weakness. It means that firm location, worker movement, and job availability can reinforce one another over time.
A useful way to think about the interface is to ask what happens after a first job. Does the local market offer a second employer with related work? Is there a neighboring occupation that uses some of the same capabilities? Can a worker acquire experience that is recognized outside the original firm? Are there institutions that help translate training into a real local demand? A high employment rate answers none of these questions by itself. Job quality and option value depend on the structure around the job.
The diagram is intentionally not a one-way chain. Firms and workers influence the market, but so do the conditions that allow people to participate in it. An employer may be present without being reachable by a worker who cannot afford housing nearby, cannot travel reliably, or cannot obtain care during working hours. A skills program may exist without producing opportunity if local employers are not hiring for the capabilities it teaches. The quality of the local system is determined at these interfaces.
Why the Local System Is More Than Its Firms
A tempting explanation for regional employment gaps is that different kinds of people live in different places. Education, age, family composition, health, and migration background can all affect employment prospects. They matter, but the OECD’s comparisons show that they do not exhaust the geographic pattern. In the countries with suitable microdata, differences in observable worker characteristics explain only about one-third to one-half of the gap between high- and low-employment regions. Residents of low-employment regions remain significantly less likely to be employed after the analysis accounts for several individual and household characteristics.
This finding should be read carefully. It is not a clean estimate of a causal “place effect.” Data cannot fully capture ability, motivation, risk-taking, specific skills, or the fact that people may have selected themselves into a region. Nor does it prove that moving a person across a boundary would produce a particular result. It does establish that the local context is not reducible to the average characteristics of residents. Similar people can encounter different employer demand, networks, institutions, and constraints depending on where they live.
This matters for the interpretation of skills. A qualification is not a complete career asset independent of context; its practical value depends on which employers recognize it, how many employers need it, and whether nearby work offers a route to accumulate related experience. A worker with the same formal education may therefore face a short search and several plausible next steps in one region, but a long search and a narrow choice in another. That does not mean the qualification has changed. The surrounding market has changed the number of ways it can be used.
It also matters for who benefits from agglomeration. A region can attract highly educated workers and high-productivity firms while leaving residents with weaker labor-market attachment facing larger barriers. The headline concentration of opportunity can coexist with unequal access to it. The OECD’s finding that worker characteristics explain only part of regional gaps should not be turned into the opposite simplification—that every resident benefits equally from place. It means that analysis must examine both the supply of capabilities and the local institutions that connect people to demand.
The local job mix is one reason. A region can contain many jobs but few jobs that build toward a broad set of future roles. It can have employment concentrated in a small number of industries, leaving workers exposed when demand changes. Another region may have a larger share of high-value services or a wider variety of productive firms. The difference is not simply that one place has “more jobs.” It is that the jobs connect differently to one another and to the skills people can acquire.
Services determine whether those connections are usable. The OECD’s evidence on essential services shows that residents of lower-GDP-per-capita regions often face longer travel times to public employment services, early childhood education and care, and primary schools. In France and Germany, almost all residents of the highest-GDP-per-capita regions can reach a public employment service within 15 minutes by car, compared with fewer than half in the lowest-GDP-per-capita regions. This measures physical accessibility, not quality, capacity, or the effect of a service on employment. It still captures a practical constraint: the cost of searching for work and sustaining a job rises when basic support is farther away.
Care is a career variable because time is a labor-market resource. Transport is a career variable because the relevant set of jobs is limited by the distance a person can reliably cover. Housing is a career variable because a wage premium can be offset by the cost of entering the region. These conditions affect people unevenly. A worker with savings, a flexible schedule, and no caregiving responsibility may respond to a distant opportunity in a way that a worker with children, a disability, or weak access to credit cannot. The local market is therefore experienced through a person’s constraints, not only through its headline employment rate.
Commuting makes the picture more complicated. The OECD generally measures employment by where people live, while productivity and sectoral employment may be measured by where work occurs. A high-productivity employment center may draw workers from surrounding regions. The center receives credit for output, while the surrounding areas record the employment of residents who travel elsewhere. Treating those places as isolated can create a misleading comparison. The relevant opportunity system may be a connected commuting area even when statistical boundaries divide it into several regions.
This is also why a high-income place is not automatically an accessible place. Capital regions often have higher median incomes and greater concentrations of high-value jobs, but higher housing costs can make those advantages difficult to capture. A worker may gain access to more employers by moving closer to a major center, yet lose purchasing power or stability if housing absorbs the gain. The opportunity set includes both the opportunities on offer and the price of being able to participate in them.
The system perspective also clarifies why policy cannot rely on a single lever. Skills policy can help only when it connects to local demand and when workers can attend and complete training. Transport can enlarge a job search area, but it cannot create employers where demand is absent. Attracting a firm can raise local employment, but a single industry may increase vulnerability if it dominates the region. Improving housing supply can make a high-employment region more accessible, but it can also change who can afford to remain. Each intervention changes one interface and may expose a constraint somewhere else.
The strongest conclusion is therefore narrower than “cities win” or “place determines success.” Local labor markets create uneven combinations of firm quality, job variety, matching, learning, services, and costs. Worker characteristics help explain who is positioned to use those combinations, but they do not fully explain why the combinations differ. A location’s economic output is informative but incomplete. The practical unit of analysis is the opportunity system that a resident can actually reach and sustain.
That conclusion creates the next problem. If places offer different combinations of jobs, incomes, ownership, and services, no single statistic can represent opportunity in full. The next step is to separate the measures—and to see why a region that looks successful on an income map may look very different on a wealth or homeownership map.