Every ministry in Central Asia has been shown the same slides: Israel’s fund-of-funds, Estonia’s digital state, the Gulf’s speed, India’s scale. Each is offered as the model to copy, and each has been copied somewhere in the region — usually the visible artefact rather than the machinery behind it. This closing article of the series does the harder thing: it sorts the models by what actually transfers to economies of eight million or thirty-seven million people with young populations, strong states, thin capital and a decade of reform behind them, sets out the sequence the evidence recommends, and lays out a thirty-six-month plan that a ministry in Astana, Tashkent, Bishkek or Dushanbe could commission this year.
The idea in brief. What transfers from the exemplars is governance and sequence, not artefacts: Israel’s lesson is to build demand for capital before supplying it and to design public instruments for their own exit; Estonia’s is that a frictionless digital state is entrepreneurship policy; the Gulf’s is speed and coordination through a few well-mandated agencies; India’s is scale through replication and registries, and the campus and skills machinery that feeds a pipeline. What does not transfer is the assumption of density — the capital, exits, mentors and customers that exemplars accumulated over decades. Central Asia’s design must therefore build the funnel before the fund: cheap entry (largely done), founder programs with regional cohorts and mid-career tracks, pre-seed and angel rungs, then fund-of-funds and corridor capital — run by a small, peripheral, well-governed agency and measured by cohort from the first intake.
What transfers, and what does not
| Source | Instrument or lesson | Transfers? | Why, and how |
|---|---|---|---|
| Israel | Yozma fund-of-funds with private buy-out terms | Yes, later | Only once a pipeline exists to absorb it; sequence grants and programs first, as Israel itself did (article 39) |
| Israel | Conditional R&D grants repaid from royalties | Partly | Works for research-based ventures; the region’s early pipeline is services and platforms, so a milestone-based pre-seed grant fits better |
| Israel | A small, autonomous innovation authority | Yes | The governance lesson transfers directly; the region’s agencies suffer mandate accretion and political selection |
| Estonia | Digital government as entry-cost policy | Yes, largely done | Kazakhstan and Uzbekistan have built it; the remaining gap is banking and cross-border services |
| Estonia | E-residency and a global founder brand | Partly | A digital-nomad and remote-founder regime fits Kyrgyzstan and Kazakhstan; the brand takes a decade |
| The Gulf | Speed: a full instrument set in five years | Yes, with caution | The region can build instruments quickly; the Gulf’s own lesson is that instruments outrun pipeline |
| The Gulf | Labour funds that pay for hiring and training nationals | Partly | Wage subsidies fit the mountain republics’ employment goals; must be tied to outcomes |
| The Gulf | Coordination across a few agencies; a single front door | Yes | Directly applicable to Bahrain-sized Kyrgyzstan and to Uzbekistan’s multiplying agencies |
| India | Registry and recognition at national scale | Yes, cheaply | Reduces friction and creates the data spine; must not be counted as an outcome |
| India | Campus and incubation networks (institution councils, tinkering labs, incubators) | Yes | Fits Uzbekistan’s campus boom (article 48); needs the faculty enablement India also under-invested in |
| India | Skills missions at scale | Yes, redesigned | Uzbekistan’s coders program is already the Indian model; add employer anchoring and placement measurement (article 38) |
| India | Replication: many cohorts, not one large program | Yes | The scaling logic for a region of eighty million across five states |
| All | Density of capital, exits, mentors, customers | No | Cannot be imported; must be grown — which is what the sequence below is for |
The pattern in the table is the series’ recurring finding. Governance, sequence and program design transfer; density does not. A ministry that copies Yozma before it has founders, or Estonia’s brand before it has banking, or the Gulf’s instrument set before it has a pipeline, buys the announcement economy described in article 31. A ministry that copies the sequence gets the outcomes a decade later, which is how long the exemplars took.
Where the region stands
The Global Innovation Index is a national-systems measure and not a verdict on programs (article 32), but the region’s ranks are a useful reality check. Kazakhstan sits in the high seventies to low eighties, strong on institutions and infrastructure and weak on outputs; Uzbekistan has climbed into the low eighties on the strength of its reforms and is the region’s mover; Kyrgyzstan and Tajikistan sit further back, with the mountain republics’ thin capital and small markets showing in every pillar.
The diagnostic reading is consistent with everything this series has found: the inputs the governments control are improving, the outputs that only a working pipeline produces are not yet visible, and the gap between the two is the founder funnel.
The sequence
The order matters more than any single instrument, and the region’s own experience — Kazakhstan’s institutions ahead of pipeline, Uzbekistan’s supply side ahead of capital — shows what happens when steps are skipped.
- Entry cost and digital state. Largely done in Kazakhstan and Uzbekistan; the residual work is banking, cross-border payments and licensing that fits services businesses. Kyrgyzstan and Tajikistan have further to go, and the work is cheap.
- Founder programs with regional cohorts and mid-career tracks. The conversion machinery: selection on evidence with a trial task (article 40), behaviour-based curriculum, cohorts in six to ten cities across the country in local languages, tracks for mid-career professionals from the state, resource and services sectors, and for returnees and relocated specialists (article 45). This is the step every country in the region has under-built.
- Pre-seed and angel rungs. Milestone-released grants attached to the programs; angel networks built from the region’s successful entrepreneurs and diaspora with co-investment matching; matched-savings instruments in the remittance economies (articles 44 and 46).
- Demand-side connection. Procurement set-asides, state-enterprise pilots and sandboxes, and corporate partners with budgets — the state as first customer in economies where it is the largest buyer.
- Campus and skills feeders. Faculty enablement and action-learning on the first-wave campuses (article 48); employer-anchored skills missions with placement measurement (article 38); women’s enterprise programs starting where the businesses are (article 47).
- Fund-of-funds and corridor capital. Once cohorts are producing ventures that clear pre-seed: public capital as a limited partner in private regional managers, domiciled in the financial centre, with Gulf and Asian co-investors and Yozma-style exit terms (articles 39 and 46).
- Agency governance and measurement throughout. A small agency with a narrow mandate, a board from the market, cohort accounting from the first intake, oversubscription used to create comparison groups, and an annual report with a cost per outcome on the cover (articles 32 and 39).
The chart is an illustrative timeline of the sequence for a ministry starting this year: a diagnostic in the first quarter, a tender-ready design by mid-year, first cohorts running before the year ends, pre-seed and angel instruments live in the second year, first 12-month outcomes reported in the second year, fund-of-funds commitments at two years, and an external evaluation at three. Faster is possible; skipping is not.
Designing for the region’s geography and languages
Four design features distinguish a Central Asian program from its templates. Language: Russian reaches the educated urban population across all five states; Uzbek, Kazakh, Kyrgyz and Tajik reach the regions where most founders live; English reaches the branch-campus graduates and the corridor. A national program runs in at least two languages and its materials in three. Geography: the founder population is distributed across cities separated by hundreds of kilometres of steppe and mountain, and the technoparks’ regional branches, universities and microfinance institutions are the only infrastructure that exists in most of them; cohorts run where the hosts are, with remote mentoring from the capitals and the corridor. Sector realism: software is a welcome minority; textiles, agri-processing, logistics, construction services, tourism, education and health are where most ventures will be, and the curriculum and mentor bench must fit them. The state: the largest employer, customer, lender and, often, competitor; programs work with it as customer and recruiting ground, and design around it as competitor.
The diaspora, the arrivals and the corridor as accelerants
Three populations can compress the region’s timeline. The diaspora — millions of Central Asians in Russia, and growing professional communities in Turkey, the Gulf, Korea, Europe and North America — supplies mentors, angels, returnees and remittance capital, and is organised, where it is organised at all, by migrant associations rather than by ecosystem programs; structured roles change that. The 2022 arrivals — the relocated engineers, product managers and founders who stayed — are the senior talent the pipeline lacked, and their integration into local cohorts as co-founders and mentors is the highest-leverage retention design available (article 45). And the corridor — Gulf capital and free zones, Indian program operators and students, the Middle Corridor’s investors — supplies the customers, capital and density that the region cannot generate alone in the time available; reciprocal cohorts, corridor funds and student-founder networks are the instruments (articles 36 and 49).
The agency: mandate, board, evaluation
The sequence needs an owner, and the region’s experience — mandate accretion in Kazakhstan’s and Uzbekistan’s multiplying agencies, donor-run programs in the mountain republics that end with the grant — argues for a specific institutional form. The owner is a small agency, or a ring-fenced unit inside an existing one, with a narrow mandate: the founder pipeline and the instruments attached to it, not investment promotion, not SME lending, not research funding. Its board draws on the market — founders, investors, the technoparks’ most credible residents, the diaspora — with the ministry holding strategy and accountability rather than operational decisions. Its staff are program operators, analysts and deal-makers, paid enough to hire them and rotated so the agency does not become a career refuge. It holds an evaluation unit with a mandate to retire programs and a duty to publish a portfolio report that includes failures. And it is, in the Breznitz–Ornston sense reviewed in article 39, deliberately peripheral: small enough to experiment, close enough to the ministry to be protected, and far enough from the centre of government to be spared the announcements. This is the institutional design that let Finland’s and Israel’s agencies innovate, and it is available to a ministry in Tashkent or Astana at a fraction of the cost of another technopark.
Budgeting the sequence
A founder pipeline is cheap relative to the artefacts the region has already bought. The largest cost lines are people — program directors, cohort leads, coaches and the mentor bench’s paid hours — followed by the pre-seed instrument, the regional delivery costs, and the measurement system; buildings do not appear, because the technopark branches, universities and microfinance institutions already exist as hosts. A useful rule of thumb from program design across the corridor is that a national program of several regional cohorts a year, with a pre-seed instrument attached and measurement built in, costs less over three years than a single technopark building — and, unlike the building, produces the data that justifies its own continuation. The budget’s most important feature is not its size but its structure: multi-year, with operating money protected from the capital budget’s cycles, a performance element tied to 24-month outcomes, and a co-funding path — corporate partners, development banks, corridor investors — that reduces dependence on any single ministry’s annual allocation.
The small-state variant
Kyrgyzstan and Tajikistan cannot run the full sequence at Kazakh or Uzbek scale, and should not try. The small-state variant, drawn from article 44, starts with district cohorts hosted by microfinance institutions and universities in the second cities as well as the capitals; uses matched savings and remittance instruments as the capital layer rather than a fund-of-funds that would have nothing to invest in; builds sector-realistic programs for agriculture, food, tourism, construction services and remote work; connects to the larger neighbours’ and the corridor’s programs for the density a small state cannot generate; and treats the diaspora and returning migrants as the primary founder recruiting ground. The agency form is the same — small, peripheral, market-boarded, evaluated — and the measurement is the same. What changes is the order: finance and district reach before technoparks, employment outcomes alongside venture outcomes, and regional cohorts with Uzbekistan and Kazakhstan as the route to scale.
A thirty-six-month plan for a ministry
| Months | Build | Measure |
|---|---|---|
| 0–3 | Diagnostic: founder population by city and sector, entry and banking frictions, existing programs and their data; agency mandate and board; data dictionary and cohort-accounting design | Baseline; definitions fixed |
| 3–6 | Program design and tender: selection rubric and trial task, curriculum in two languages, mentor bench with obligations, regional hosts, pre-seed instrument design, procurement channel with two state enterprises | Tender specifies outcomes and horizons |
| 6–12 | First cohorts in four to six cities, including a mid-career and a returnee track; angel network launched with co-investment matching; faculty-enablement track on first-wave campuses; women’s district cohorts through microfinance hosts | Leading indicators: qualified demand, completion with evidence, customer validation |
| 12–18 | Second intake, oversubscribed and transparently allocated; pre-seed grants released against milestones; first reciprocal corridor cohort; skills mission redesign with employer anchoring | First 12-month outcomes; comparison group established |
| 18–24 | Fund-of-funds mandate and domicile; first commitments to private regional managers with corridor co-investors; expansion to further cities; retention services for relocated founders | Cost per operating venture; private capital crowded in |
| 24–36 | Replication across regions; instrument retirement where outcomes are weak; agency’s first portfolio report with failures included; external evaluation commissioned | 24-month outcomes by cohort; additionality; the number on the cover |
What success looks like at five years
The sequence’s end state is easy to describe and rare to observe. A founder program with a known selection rate, cohorts in most regions, and 24-month outcomes by intake published annually. A pre-seed instrument whose recipients raise angel and venture money at a measured rate. An angel network with deal flow, syndication and co-investment. Private early-stage managers, seeded by a fund-of-funds and backed by corridor investors, with a portfolio drawn from the program’s graduates. Campuses feeding teams to technopark branches that incubate them. A small agency that has retired at least two programs and published at least one report with failures in it. And — the number on the cover — a cost per operating venture that a finance ministry can compare with the alternatives. No country in the region has all of this today. Any of them could have most of it in five years, and the order in which they build it will decide whether they do.
A composite case: the ministry that ran the sequence
A composite from several regional engagements; details altered and no single country described.
A ministry in one of the region’s reforming states had, by the time it commissioned a new strategy, most of the artefacts: a technopark, a venture vehicle, a coding program, a registry, university entrepreneurship centres and a decree announcing an ecosystem. It also had, when asked, no number for ventures operating at 24 months from any of them. The strategy it commissioned did not add an artefact. It sequenced what existed.
A founder program with regional cohorts and a mid-career track was tendered against outcomes; the venture vehicle’s mandate was rewritten to fund a pre-seed instrument attached to the program and to seed an angel network; the coding program was re-anchored on employers with placement measurement; twenty campuses entered a faculty-enablement track; the technopark’s regional branches became cohort hosts; a procurement channel with two state enterprises was opened; and a small agency with a market board and an evaluation unit was given the whole portfolio. In the second year a reciprocal cohort with a Gulf free zone ran, and the vehicle made its first fund-of-funds commitment to a regional manager with Gulf co-investors. At thirty-six months the ministry published the region’s first portfolio report by cohort, with a cost per operating venture on the cover and two retired programs in the appendix. The numbers were modest. They were the first honest ones in the region, and every neighbouring ministry asked for the template.
What could go wrong
- Copying artefacts. Antidote: copy governance and sequence.
- Instruments before pipeline. Antidote: founder programs first; funds when cohorts produce ventures.
- Capital-city programs. Antidote: regional cohorts through technopark branches, universities and microfinance hosts.
- Coders counted as founders. Antidote: separate services metrics from venture metrics.
- Mandate accretion. Antidote: a small agency, narrow mandate, market board.
- Donor cycles. Antidote: national ownership and co-funding paths from day one.
- No data. Antidote: cohort accounting from the first intake; oversubscription as evaluation.
Questions ministers actually ask
“Which country’s model should we copy?” None whole. Israel’s sequence and governance, Estonia’s entry policy, the Gulf’s speed and coordination, India’s scale and campus machinery — and your own geography, languages and state.
“How long until results?” Leading indicators within a year; 12-month outcomes in the second year; 24-month outcomes and a defensible cost per venture in the third. Anyone promising faster is promising inputs.
“We already have a technopark, a fund and a registry. What is missing?” The funnel between them: founder programs with regional cohorts, pre-seed and angel rungs, demand-side connection, and measurement.
“What is the corridor’s role?” Density on loan: capital, customers, operators and networks from the Gulf and India that the region cannot generate alone in the time available.
“What is the first thing to commission?” A diagnostic and a founder program tender specified by outcomes — and a data dictionary, because the report in year three is written in month one.
Methodology & data notes
Global Innovation Index ranks are as published in each annual edition and vary with methodology; the timeline chart is an illustrative plan, not a measurement. Descriptions of exemplar institutions reflect public information and the literature cited across this series. The thirty-six-month plan is a design proposal that must be adapted to each country’s mandate, budget and institutions. The composite case combines several engagements with details altered. This article closes the Central Asia series; companion articles cover the regional map (41), Kazakhstan (42), Uzbekistan (43), the mountain republics (44), the 2022 migration (45), finance (46), women entrepreneurs (47), campuses (48) and the corridor (49), and the Government & Ecosystems series (31–40) supplies the evidence base it draws on.
References & further reading
- Israel Innovation Authority — Yozma history and current instruments
- Avnimelech, G. & Teubal, M. (2006) — Creating venture capital industries that co-evolve with high tech, Research Policy
- Breznitz, D. & Ornston, D. (2013) — The Revolutionary Power of Peripheral Agencies, Comparative Political Studies
- Estonia e-Residency and Startup Estonia — Estonia’s founder and digital-state programs
- Startup India and Atal Innovation Mission — India’s registry, incubation and campus networks
- Bahrain Economic Development Board and Tamkeen — the Gulf’s coordinated agency model
- WIPO — Global Innovation Index country profiles for Central Asia
- González-Uribe, J. & Leatherbee, M. (2018) — The Effects of Business Accelerators on Venture Performance: Evidence from Start-Up Chile, Review of Financial Studies
- McKenzie, D. (2017) — Identifying and Spurring High-Growth Entrepreneurship: Experimental Evidence from a Business Plan Competition, American Economic Review
- Campos, F. et al. (2017) — Teaching personal initiative beats traditional training in boosting small business in West Africa, Science
- Lerner, J. (2009) — Boulevard of Broken Dreams, Princeton University Press
- EBRD and OECD — Transition Report and Eurasia Competitiveness Programme analysis of Central Asia
HexGn designs and runs national founder programs for governments — regional cohorts, mid-career tracks, pre-seed and angel instruments, corridor cohorts and cohort measurement — across Central Asia, the Gulf and India.