High-skill migration as a matching problem
Modified
Input
High-skill migration succeeds when a person's capability, an employer's task, and the host institution's productive system fit one another. Entry rules can select candidates, but retention depends on professional opportunity, institutional predictability, knowledge access, and household conditions. Migration policy should evaluate additional productive capacity and integration cost rather than rank people by credentials or nationality alone.
More Than Moving a Worker
A vacancy exists. A capable person lives elsewhere. It is tempting to describe the solution as moving the person across a border.
The economic problem is more demanding.
The worker must prefer the destination to available alternatives. The employer must be able to use the worker's capability. The host system must recognize qualifications, permit productive work, and provide enough stability for the worker and household to remain. Colleagues must share a workable professional language and set of expectations.
Migration is therefore a matching and institutional-integration problem.
The Worker's Decision
Let worker $i$ choose destination $r$ by comparing expected lifetime utility:
where $w$ is compensation, $\ell$ is learning and career development, $a$ is amenities and household benefit, $c$ is living cost, $q$ is institutional and professional risk, and $F$ is the fixed cost of migration.
The worker migrates when
where $h$ is the home location and the remaining terms are alternative destinations.
A policy that asks only why talent should enter a country omits the comparison. Skilled people usually have options.
The Employer's Match
Let the productivity of worker $i$ in firm $f$ be
where $H_i$ is the worker's capability, $I_f$ is the firm's knowledge base, $P_f$ is its technological and financial capital, and $m_{if}$ is a match term.
The same person can produce very different output across firms because $I_f$, $P_f$, and $m_{if}$ differ.
Table 1. Components of a successful high-skill migration match
| Worker-firm condition | Likely outcome |
|---|---|
| Strong worker, strong knowledge system, strong match | High productivity and learning |
| Strong worker, weak knowledge system | Frustration or compensating effort |
| Strong worker, unsuitable task | Underuse and early exit |
| Moderate worker, strong complementary team | Development and useful contribution |
| Strong credentials, weak verified capability | Expensive hiring error |
Migration does not repair a weak organizational system. It can expose it.
Transferability of Human Capital
Some human capital transfers readily across borders. Some is location specific.
Let total capability be
where $H_i^G$ is general capability and $H_{ir}^S$ is destination-specific capability.
Mathematical reasoning may transfer broadly. Professional licensing, language, regulation, customer behavior, or institutional conventions may not. The effective capability immediately after migration is
The parameter $\kappa_{ir}$ measures how much location-specific knowledge transfers.
Over time, the worker accumulates destination-specific capital:
where $E_{ir}$ is integration effort and $\lambda_{ir}$ is the effectiveness of the learning environment.
Integration Cost Is Jointly Produced
The worker is not solely responsible for integration.
Integration cost depends on language support, documentation, management quality, credential recognition, community access, and the distance between professional systems.
We can write
where $d(\cdot)$ is institutional distance, $S_f$ is firm support, and $N_r$ is access to relevant local networks.
The employer lowers the cost by making workflows explicit, assigning meaningful work, providing feedback, and recognizing prior expertise. The host institution lowers it through predictable rules and accessible professional infrastructure.
Read as a diagnostic rather than a second scorecard, the framework becomes clear: Professional language — Possible institutional response: Technical communication support and shared documentation. Unclear authority — Possible institutional response: Explicit decision rights and role design. Credential uncertainty — Possible institutional response: Capability-based assessment. Weak local network — Possible institutional response: Mentoring, seminars, and professional associations. Household instability — Possible institutional response: Predictable residence and family conditions. Knowledge-system gap — Possible institutional response: Structured access to data, methods, and colleagues.
Integration is not cultural assimilation in the abstract. It is the construction of a workable productive relationship.
Selection Rules and Prediction
Migration systems often use observable indicators such as education, salary, occupation, experience, age, and language.
These are proxies for expected contribution. Let expected net contribution be
where $\mathbf z_i$ contains the observable indicators.
The prediction will contain error. A degree may not reveal capability. A high offered salary may reflect a shortage, market power, or an internal transfer. An occupation list can become obsolete. Language scores may poorly measure professional communication.
A credible selection system should therefore be tested against outcomes and revised. It should not confuse administrative simplicity with predictive validity.
“High Skill” Is Task Dependent
The label high skill can become an unexamined status category.
A worker creates high value when scarce capability fits a productive task. The same capability may be abundant in one location and scarce in another. A role with modest formal education may be essential when substitution through technology or capital is difficult.
A shortage index can be written
where $D_{jr}$ is demand for task $j$ in region $r$ and $L_{jr}$ is effective local supply.
Policy should consider the shortage, the social value of the output, the feasibility of training, and the cost of integration. Credential level alone is an incomplete rule.
Attraction Without Retention
A destination may admit capable migrants and still fail to accumulate human capital.
Let the human-capital stock evolve as
where $\delta$ is depreciation, $M^{in}$ is effective capability entering, $M^{out}$ is capability leaving, and $T_t$ is domestically developed capability.
Gross inflow can look impressive while net accumulation remains small.
Retention depends on:
- the ability to perform meaningful work;
- advancement across more than one employer;
- research and professional community;
- predictable rules;
- recognition of contribution;
- household opportunity; and
- a credible long-term path.
The same factors that retain locally trained specialists tend to retain migrants. A system that loses its own capable people will struggle to compensate through recruitment alone.
Retention data should also distinguish a productive departure from a failed match. A researcher may leave after creating a laboratory, training colleagues, or connecting the host institution to an international network. Another may remain while performing work far below the person's capability. Duration alone is therefore an incomplete outcome.
A better evaluation combines length of stay with responsibility, output, knowledge transfer, career progression, and the worker's stated reason for the next move. The goal is not to prevent mobility. It is to make each period of participation productive for both the individual and the host system.
Spillovers and Distribution
High-skill migration can create benefits beyond the worker's direct output: knowledge transfer, entrepreneurship, new trade links, mentoring, and stronger professional networks.
Let social benefit be
where $\sigma_i$ is the spillover and $C_i$ is the full public and integration cost.
The distribution of benefits matters. A firm may capture most direct output while a locality pays infrastructure costs. Existing workers may gain through complementary demand or face competition in a narrow occupation. Consumers may benefit from more services.
Policy analysis should identify these channels rather than assume a single national effect.
Universities as Integration Institutions
Universities can connect education, research, and migration.
They offer a setting in which international students and researchers accumulate destination-specific knowledge before entering the wider labor market. They create professional networks, verify capability, and connect employers to emerging specialists.
Their effectiveness depends on whether the transition continues beyond admission and graduation. A program that recruits internationally but teaches in an isolated track may generate weak integration. A program that connects students to research, cases, firms, and public professional work can raise $\lambda_{ir}$, the rate of destination-specific learning.
A Better Policy Sequence
A productive migration strategy asks:
- Which outputs and tasks are constrained by scarce capability?
- Can domestic training, technology, or organizational redesign relax the constraint?
- Which capabilities are internationally transferable?
- Why would suitable workers choose this destination?
- Which employers can use the capability productively?
- What integration costs will arise, and who will bear them?
- What conditions support retention and spillovers?
- How will outcomes revise the selection model?
This sequence treats migration as one instrument in a human-capital system.
Matching surplus is also affected by factors outside the job description. Licensing rules can prevent a professional from using prior expertise; visa uncertainty can shorten the horizon over which a firm is willing to invest; family constraints can make an apparently attractive offer infeasible; and language can change access to clients or informal knowledge. These costs are jointly produced by the worker, employer, and host institution. Treating them as evidence of weak individual quality creates a selection error. A better system identifies which parts of human capital transfer immediately, which require local adaptation, and which organizational changes would allow the match to become productive. Retention then becomes an outcome of match quality, not merely a second recruitment campaign. The same logic improves evaluation after arrival. Early wages may understate eventual contribution when local adaptation is still occurring, while an impressive initial title may overstate a match that provides little authority or knowledge access. Longitudinal evidence is therefore essential. For employers, this means monitoring task progression, collaboration, and decision responsibility rather than treating retention alone as proof of success. For governments, it means linking immigration policy to credential recognition, family stability, and institutions that reduce the cost of entering professional networks. These complements determine whether selection at the border becomes productivity after arrival. A high-quality match should become more productive over time as local knowledge grows; stagnation is a signal that one of the complements remains blocked. Policy quality is revealed by that trajectory, not by the number of visas issued in the first period.
The adjustment problem is visible in current AI labor analysis. The AI Trade Shock Will Reward Countries That Move Workers, Not Just Data argues that national gains depend on whether workers can move into more productive activities. AI Exposure Geography shows that local capacity conditions this movement, and The Training Gap identifies adaptation as an institutional bottleneck. Migration succeeds under the same logic: movement creates potential, but matching, recognition, networks, and organizational authority determine whether that potential becomes output.
Conclusion
High-skill migration is not a contest to collect impressive résumés. It is a matching problem among people, tasks, organizations, and institutions.
Selection can increase the probability of a useful match, but it cannot create the surrounding productive system. Attraction can initiate a move, but it cannot guarantee retention. Credentials can provide evidence, but they cannot replace capability-based assessment.
The durable objective is net productive capacity: people able to contribute, learn, and remain inside institutions that can use what they know.
References
George J. Borjas, “The Economic Analysis of Immigration”03009-6), in Handbook of Labor Economics, 1999.
Sari Pekkala Kerr, William Kerr, Çağlar Özden, and Christopher Parsons, “High-Skilled Migration and Agglomeration”, Annual Review of Economics, 2017.
AnnaLee Saxenian, The New Argonauts: Regional Advantage in a Global Economy, Harvard University Press, 2006.
The Economy Editorial Board (2026) ‘The AI Trade Shock Will Reward Countries That Move Workers, Not Just Data’, The Economy Review, 2 June.
The Economy Editorial Board (2026) ‘AI Exposure Geography Is a Capacity Problem, Not a Red-Blue Problem’, The Economy Review, 11 June.
The Economy Editorial Board (2026) ‘The Training Gap Behind the Rise of SuperHuman Labor’, The Economy Review, 28 July.
Similar Post
