In the last week of September 2022, the border crossings between Russia and Kazakhstan saw queues that stretched for days. Russia had announced a military mobilisation, and hundreds of thousands of its citizens — disproportionately young, urban, educated and employed in technology — left in the space of weeks, on the heels of a first wave that had departed after the invasion of Ukraine in February. Most went to the countries they could enter without a visa: Kazakhstan, Kyrgyzstan, Uzbekistan, Armenia, Georgia, Turkey. For Central Asia’s young ecosystems, it was the largest inflow of technical talent since independence, arriving unplanned and all at once. This article examines what happened, what the research on migrant talent predicts, what the region actually gained, and how governments can design for the retention that turns a passing wave into a permanent asset.
The idea in brief. The 2022 migration delivered, within months, a population of experienced engineers, product managers and founders that Central Asia’s programs could not have produced in a decade. The research on migration and innovation — from the German émigrés of the 1930s to the Soviet mathematicians of the 1990s to Silicon Valley’s transnational engineers — is unambiguous that such inflows raise patenting, company formation and knowledge transfer where the newcomers are able to work and stay. The region’s governments responded quickly with visas, tax statuses and relocation programs, and the gains were real: company registrations, technopark residents, salaries and rents all rose. But most of the migrants worked remotely for Russian employers, many moved on within a year or two, and integration into local ecosystems was thin. The design lesson is that a relocation program is not a retention program: residency pathways, company formation, banking, schools, and — above all — connection to local founders, customers and capital are what convert arrivals into ecosystem.
What happened
Two waves left Russia in 2022. The first, after February, was a departure of those with the means and the motive to leave quickly — technology workers whose companies relocated or whose clients were abroad, journalists, professionals with international ties. The second, after the mobilisation order of September, was larger, more sudden and more male, and it went overwhelmingly to the neighbours that Russians could reach by land or without visas. Kazakhstan’s interior ministry reported on the order of two hundred thousand Russian citizens entering in the weeks after the announcement, of whom a large majority moved on — to Kyrgyzstan, Uzbekistan, Turkey, the Gulf, Serbia, Armenia and Georgia — or later returned; Kyrgyzstan and Uzbekistan each received tens of thousands. Russia’s own digital ministry estimated that around a hundred thousand information-technology specialists had left the country during 2022, and that most continued to work for Russian employers remotely.
The composition mattered as much as the numbers. The migrants were young, educated and employed; a large share worked in software, product and design; many had savings and a professional network that ran to Berlin, Tel Aviv and San Francisco as well as Moscow. For ecosystems that had spent a decade training junior coders, the arrival of senior engineers and product managers in the thousands was a shock to the top of the talent pyramid.
What the research says migrant talent does
The economics of migration and innovation has an unusually clear evidence base, built on natural experiments that let researchers see what happens when skilled people arrive somewhere suddenly.
Petra Moser, Alessandra Voena and Fabian Waldinger’s 2014 study in the American Economic Review examined the German Jewish scientists who fled to the United States in the 1930s and found that patenting by American inventors rose substantially in the research fields the émigrés brought with them — not merely because the émigrés themselves patented, but because they trained and collaborated with locals who then patented more. George Borjas and Kirk Doran’s 2012 paper in the Quarterly Journal of Economics studied the arrival of Soviet mathematicians in the United States after 1992 and found a more complicated picture: the newcomers were extraordinarily productive, and in the sub-fields where they clustered, American mathematicians published less and moved to other topics — competition as well as spillover. Jennifer Hunt and Marjolaine Gauthier-Loiselle’s 2010 work in the American Economic Journal: Macroeconomics estimated that a one-percentage-point rise in the share of immigrant college graduates raised patents per capita by around ten per cent, largely because immigrants were disproportionately scientists and engineers.
Closer to the entrepreneurship question, Sari Pekkala Kerr and William Kerr’s 2020 analysis of American business-owner surveys in Research Policy found that immigrants founded firms at rates well above their population share, that immigrant-founded firms were more likely to export, and that they created jobs at comparable rates to native-founded firms. And AnnaLee Saxenian’s work on transnational engineers, discussed in article 36, showed that migrants who stay connected to both places become the channels through which capital, knowledge and practice flow in both directions.
The evidence, then, points one way with a caveat. Skilled inflows raise innovation and company formation — through the newcomers’ own work, through what they teach locals, and through the networks they carry. The caveat, from Borjas and Doran and from the practical experience of every host country, is that the gains depend on integration: newcomers who work alongside locals produce spillovers; newcomers who work remotely for employers elsewhere, in an enclave, produce rent increases.
Who the arrivals were: three profiles
Treating the 2022 wave as a single population obscures the design problem, because three quite different groups arrived together and needed different things. The remote employee — the largest group by far — kept a job with a Russian or international technology company and needed a place to live, a bank account that could receive a salary, and a tax status that did not double-tax it; this group’s economic footprint was consumption and rent, its ecosystem footprint minimal, and its stay contingent on the employer’s tolerance for remote work and the host’s tolerance for its presence. The relocating company — hundreds of firms, from small studios to units of large groups — needed an entity, an office, a tax regime and, above all, banking that could handle international clients; this group registered with the technoparks, hired locally in modest numbers, and stayed where the banking worked. The founder in transit — the smallest group and the most valuable — arrived with an idea, savings and a network, and needed what founders everywhere need: co-founders, customers, capital and a residency long enough to build. The programs of 2022 served the first two groups adequately and the third almost not at all, because they were built as arrival services and the third group’s needs are integration services. A retention design starts by asking which of the three it is for.
The employer’s view: hiring the arrivals
For the region’s own companies and capability centres, the wave was a hiring opportunity of a kind that had not existed before — senior engineers and product leaders available locally, at salaries that were high by regional standards and low by Moscow’s. The employers that benefited did three things: they hired quickly, before the arrivals’ networks pulled them onward; they paired arrivals with local teams rather than building Russian-speaking enclaves inside the company; and they treated the arrivals’ remote-work expectations as a design parameter, offering hybrid arrangements that the local market had not previously needed to offer. Employers that waited for the wave to settle found that it had moved on. The lesson generalises to the region’s capability-centre and technology employers for any future inflow: the window is measured in months, and the integration that produces spillover has to be built into the job, not left to the city.
Armenia and Georgia as comparators
Two non-Central-Asian neighbours offer a useful comparison. Armenia, with a small, dense and internationally connected technology sector, absorbed a large inflow of relocating firms and specialists relative to its size, registered thousands of companies, and saw measurable effects on its technology exports — helped by a shared language of business, an active diaspora and a government that moved fast on registration and banking. Georgia received a comparable inflow and a stronger political backlash, and retained less. Neither offered residency pathways of the kind that the Gulf and Serbia later used to attract the same people, and both saw significant onward movement by 2024. The comparison suggests that density and connectivity help a host retain arrivals, that political tolerance is a precondition, and that no host in the region has yet built the residency-and-integration machinery that would have kept the wave.
What the region did
| Country | Response | Observed effect |
|---|---|---|
| Kazakhstan | Visa-free entry retained; a digital-nomad visa introduced; the technopark offered relocation support and its tax status to relocating companies; company registration by foreign citizens simplified | Thousands of company registrations by Russian citizens; a surge in technopark residents; sharp rises in Almaty rents and technology salaries; a wave of remote workers serving Russian and international employers |
| Uzbekistan | The technopark’s dedicated relocation program offering fast-track residency, tax status, office space and assistance with banking and housing | Hundreds of relocated firms and thousands of specialists registered; Tashkent’s ecosystem gained senior talent it had never had; the program became a regional case study |
| Kyrgyzstan | Digital-nomad status; light-touch registration; the technology park’s regime open to relocating firms | Bishkek became a low-cost base for remote workers and small teams; rents and demand for co-working rose; the inflow was large relative to the city |
| Armenia, Georgia, Turkey, Serbia, the Gulf | Varied: from Armenia’s rapid absorption of relocating firms to the UAE’s residency-by-investment routes | Competing destinations that absorbed a share of the second and third moves as migrants sought long-term residency, banking and schools |
The responses were fast and, by the standards of migration policy anywhere, generous. They were also, almost everywhere, relocation programs rather than retention programs: they solved the problems of arrival — a visa, a tax status, an office — and left the problems of staying — residency that leads somewhere, bank accounts that work, schools for children, a local professional network, a reason to build here rather than remotely for there — to the migrants themselves.
What the region gained, and what it did not
Three gains were real. Senior talent arrived at the top of a pyramid that had been all base: product managers, architects, data scientists and founders with years of experience in Russia’s large, sophisticated technology industry. Company formation rose sharply, both from relocating firms and from new ventures founded by arrivals, and the technoparks’ resident and export figures jumped. Connectivity improved: the arrivals brought networks into Europe, Israel and the United States, and a familiarity with global product practice that local ecosystems had lacked.
Three gains were smaller than hoped. Spillover to locals was limited, because most arrivals worked remotely for Russian or international employers and lived in Russian-speaking enclaves; the Borjas–Doran pattern of competition without collaboration was visible in some sectors, where local engineers found themselves priced out of apartments and, occasionally, out of jobs. Local company-building was less common than remote work: a specialist earning a Moscow salary from Almaty had little incentive to found a Kazakh company. And permanence was low: by 2023 and 2024 a large share of the second wave had moved on to Serbia, Cyprus, the Gulf or further west in search of residency and banking, or had returned to Russia as the immediate threat receded. The region kept a valuable minority; it did not keep the wave.
The chart is an illustrative model of the retention problem: a relocation program that provides a visa and a tax status keeps a small share of arrivals beyond two years, while a program that adds residency pathways, company-formation support, banking, schooling and integration into local cohorts keeps materially more. The shape is drawn from the experience of host ecosystems, not from a single dataset, and the point is the gap between the two curves.
Designing for retention
The next inflow — from Russia or elsewhere — will come, and the design for it is clear from what happened in 2022.
- Residency that leads somewhere. A nomad visa is a stay; a pathway from work permit to long-term residency is a life. Arrivals who can see a five-year horizon form companies, buy apartments and enrol children; arrivals on annual stamps keep a bag packed.
- Company formation and banking as one service. The most common silent failure for relocating founders was the bank account. A named banking partner with a service standard, integrated with company registration, is worth more than any tax holiday.
- Integration into local cohorts. Relocated founders and specialists should enter the same founder programs, accelerators and mentor benches as locals — as participants, mentors and co-founders — rather than parallel programs in Russian. Co-founder matching between arrivals with product experience and locals with market knowledge is the single highest-leverage design in this field.
- Local customers. Procurement access, pilots with state enterprises and banks, and introductions to the region’s corporates give relocating firms a reason to build for the market they live in rather than serve the one they left.
- Schools, housing and family. The decisions that determine whether a family stays are made at the school gate. International and bilingual school capacity, and housing policy that anticipates inflows rather than absorbing them through rent spikes, are ecosystem policy.
- Compliance clarity. Relocating firms and their bankers need to know, in writing, how sanctions and counter-sanctions apply to them. Ambiguity is the surest way to lose a firm to a jurisdiction that has answered the question.
- Measure retention, not arrivals. Report arrivals registered, companies formed, and — at 12 and 24 months — companies still operating, specialists still resident, local hires made. The last three are the program.
The wider pattern: Central Asia as a destination
The 2022 wave was the largest but not the first sign that the region’s migration story is reversing at the top. Kazakhstan has for years attracted managers and specialists from Russia, Ukraine, Turkey, Korea and the wider region to its resource, banking and technology sectors; Uzbekistan’s opening drew returnees from its own diaspora and, increasingly, professionals from across Eurasia; the Gulf–Central Asia corridor discussed in article 49 brings Gulf capital with Gulf-based managers; and Indian firms building for Central Asian markets bring their own. A region that has spent three decades exporting labour is learning to import talent, and the institutions it builds for the next inflow — residency, banking, integration, measurement — are the same ones that will let it keep its own graduates.
A composite case: the relocation program that became an integration program
A composite from several regional engagements; details altered and no single program described.
A technopark had run a relocation program through 2022 that was, by every input measure, a success: hundreds of firms and thousands of specialists registered, offices filled, export figures up. A review in 2024 found that a majority of the relocated specialists had left the country, that most of those who remained worked remotely for employers elsewhere, and that the number of relocated founders who had built companies with local co-founders, local customers or local investors could be counted on two hands.
The redesign kept the arrival services and added the staying services. Residency pathways were negotiated with the migration authority; a banking partner was signed with a service standard; the park’s accelerator opened its cohorts to relocated and local founders together, with a co-founder matching track that paired product experience with market knowledge; a procurement channel with two state enterprises gave the mixed teams a first customer; and a school-places arrangement was made with two international schools. The program’s report a year later led with retention at 12 months, companies formed with local co-founders, and local hires — numbers that were small, real, and rising, and that the park had never previously been able to produce.
What could go wrong
- Arrivals counted as outcomes. Antidote: retention and local company formation as the reported metrics.
- Enclave economics. Remote work for foreign employers, Russian-speaking bubbles, rent spikes without spillover. Antidote: integration into local cohorts and co-founder matching.
- Banking failure. Antidote: a named partner with a service standard as part of the program.
- Compliance ambiguity. Antidote: written guidance for firms and banks.
- Local resentment. Housing and wage pressure blamed on arrivals. Antidote: housing policy that anticipates inflows; programs that visibly benefit locals through jobs and co-founding.
- Waiting for the next wave. Antidote: build the retention machinery now; it serves returnees and the region’s own graduates too.
Questions ministers actually ask
“Did the 2022 migration help us?” Yes, and less than it could have. The senior talent, the company registrations and the networks were real; the retention and the spillover to locals were thin because the programs were built for arrival rather than staying.
“Should we compete with the Gulf and Serbia for these people?” On residency, banking and integration, yes — those are what the migrants were seeking when they moved on. On tax holidays, no; they already had those, and left anyway.
“How do we make sure locals benefit?” Put arrivals and locals in the same cohorts, match co-founders across the two groups, and give mixed teams the first customers. Spillover follows collaboration, not proximity.
“What is the risk?” Sanctions ambiguity, housing pressure and political sensitivity about Russian populations. All three are manageable with written guidance, anticipatory housing policy and programs that visibly serve locals.
“What should we build first?” Residency pathways and banking, because they are what the region lacked in 2022; then integration into founder programs, because it is what turns talent into ecosystem.
Methodology & data notes
Migration figures are approximate and drawn from statements by the Kazakh interior ministry, Russia’s digital ministry and press reporting at the time; they describe orders of magnitude, and the number who settled permanently is unknown and smaller than the number who entered. The retention chart is an illustrative model, not a measurement. Findings from the migration-and-innovation literature are summarised approximately; readers should consult the papers for samples, methods and the important caveats in Borjas and Doran. Descriptions of government responses reflect public information as of 2025. The composite case combines several engagements with details altered. Companion articles cover the regional map (41), Kazakhstan (42), Uzbekistan (43), the corridor (49) and program design (50).
References & further reading
- Moser, P., Voena, A. & Waldinger, F. (2014) — German Jewish Émigrés and US Invention, American Economic Review
- Borjas, G. & Doran, K. (2012) — The Collapse of the Soviet Union and the Productivity of American Mathematicians, Quarterly Journal of Economics
- Hunt, J. & Gauthier-Loiselle, M. (2010) — How Much Does Immigration Boost Innovation?, American Economic Journal: Macroeconomics
- Kerr, S.P. & Kerr, W. (2020) — Immigrant entrepreneurship in America: Evidence from the survey of business owners 2007 & 2012, Research Policy
- Docquier, F. & Rapoport, H. (2012) — Globalization, Brain Drain, and Development, Journal of Economic Literature
- Saxenian, A. (2006) — The New Argonauts, Harvard University Press
- Astana Hub and IT Park Uzbekistan — relocation and resident programs
- International Organization for Migration — Central Asia migration data and reports
- EBRD — Transition Report analysis of the 2022 migration’s economic effects
- World Bank Open Data — regional economic indicators
HexGn designs retention-first talent and founder programs for governments and technoparks — residency, banking and integration built into the cohort — so that the next inflow of talent stays and builds.