Walk through a bazaar in Osh, a textile workshop in Namangan or a dairy cooperative outside Dushanbe and the entrepreneurs you meet are, more often than not, women. Look at the resident lists of the region’s technoparks, the portfolios of its venture funds or the stages of its pitch competitions and they almost disappear. Central Asia’s women run a large share of its actual small-business economy and appear in a small share of its startup statistics — a gap that is partly measurement, partly the structure of the economy, and partly the design of the programs that were built for someone else. This article reviews the evidence on what works for women’s enterprise, the barriers that are specific to the region, and a program design that starts from where women’s businesses already are.
The idea in brief. Women’s labour-force participation ranges from high in Kazakhstan to low in Tajikistan, and women’s businesses across the region cluster in services, retail, food, textiles and agriculture, mostly informal, mostly financed by household savings and remittances, and largely outside the reach of technoparks and venture funds. The randomised evidence on women’s enterprise programs is unusually good: mentorship by experienced local entrepreneurs raises profits where classroom training does not; personal-initiative training works; mobile money moves women into business; and bundled programs that combine skills, finance and networks outperform any component alone. The barriers that matter in Central Asia — collateral and property, mobility and social norms, time and care, thin networks — are addressable by design. The program that works recruits in districts, uses women entrepreneurs as mentors, bundles credit with behavioural training through microfinance institutions, builds on digital rails, and measures survival and income at 12 and 24 months.
The numbers, and what they hide
The headline indicator — women’s labour-force participation — spans the region’s diversity. World Bank and ILO modelled estimates put it above sixty per cent in Kazakhstan, a Soviet legacy of near-universal women’s employment that survived the transition; in the mid-forties in Kyrgyzstan; around forty per cent in Uzbekistan; and around thirty per cent in Tajikistan, where male migration, large families and conservative norms in rural districts combine to keep women out of the recorded workforce.
The participation numbers understate women’s economic activity for a reason that matters for program design: much of it is informal and home-based, and much of it is enterprise. In the remittance economies of the south, as article 44 noted, the household’s business — the shop, the plot, the sewing machines, the guesthouse — is run by the woman who stayed while the men migrated. World Bank Enterprise Surveys of the region’s formal firms find women in ownership or top management at a minority of firms, with the share varying by country and measure; the informal sector, which the surveys do not cover, is where most women’s businesses are. The technopark statistics that dominate the region’s startup narrative — residents, exports, relocations — describe a sector that is overwhelmingly male, urban and formal, and that is a small fraction of the region’s private economy.
Where women’s businesses are
| Sector | Typical business | Finance | Growth constraint |
|---|---|---|---|
| Retail and bazaar trade | Market stalls, small shops, re-export trade | Household savings, supplier credit, microfinance | Working capital; formalisation costs; mobility |
| Food and hospitality | Catering, cafés, home-based food production, guesthouses | Savings, remittances | Licensing; access to premises; seasonal demand |
| Textiles and crafts | Sewing workshops, embroidery, carpets, garment sub-contracting | Savings; buyer advances; microfinance | Access to buyers beyond the district; equipment finance |
| Agriculture and processing | Dairy, fruit and vegetable processing, greenhouses, livestock | Remittances; cooperatives; donor programs | Land rights; cold chain; market linkage |
| Services | Education and tutoring, beauty, health, childcare, bookkeeping | Savings | Premises; professional licensing; scale |
| Digital and remote | Online retail, design, content, software services, remote employment | Savings; technopark grants; angels (rare) | Skills, connectivity, payments; the one sector where technopark programs reach women |
The pattern is the one Leora Klapper and Simon Parker documented across countries in the World Bank Research Observer in 2011: women’s businesses are smaller, more concentrated in services and retail, more often home-based and part-time, and more constrained by credit and by the legal and social environment than men’s — and the gap reflects the environment more than the entrepreneurs.
The barriers that are specific to the region
- Collateral and property. Bank credit in the region is collateral-based, and property — land, housing, vehicles — is disproportionately registered to men. A woman running a profitable business often cannot borrow against it because she does not own the assets it uses.
- Mobility and norms. In rural districts of Tajikistan, Uzbekistan and southern Kyrgyzstan, women’s travel to markets, training venues and government offices is constrained by norms and by care responsibilities; a program in the provincial capital is, for many, a program in another country.
- Time and care. Large families, absent migrant husbands and limited childcare make time the scarcest input; a twelve-week cohort with weekly full-day sessions excludes most of the women it is meant to reach.
- Networks. The informal networks through which men find suppliers, customers, credit and information run through spaces — teahouses, bazaar guilds, male-dominated associations — that women enter less easily.
- Formalisation costs. Registration, licensing and tax administration were simplified in the reform decade (article 41), but the residual costs fall hardest on the smallest, home-based businesses, which are disproportionately women’s.
- Visibility. Programs, funds and statistics are built around the formal, urban, technology-oriented founder; women’s businesses are neither counted nor targeted.
What the evidence says works
The randomised literature on women’s enterprise is among the strongest in development economics, and it is unusually consistent about mechanisms.
Mentors beat classrooms. Wyatt Brooks, Kevin Donovan and Terence Johnson’s 2018 experiment in Nairobi, published in the American Economic Journal: Applied Economics, assigned women microenterprise owners either to formal business classes or to mentorship by an experienced local entrepreneur in the same trade. Mentorship raised profits by around a fifth; classes did not measurably raise them at all. The mentors’ value was specific, local knowledge — suppliers, prices, customers — that no curriculum contained.
Behaviour beats content. The Togo experiment reviewed in article 40, published in Science, found that personal-initiative training raised firm profits substantially where traditional business training did not, with effects that held for women as well as men. Erica Field, Seema Jayachandran and Rohini Pande’s 2010 experiment with women in India, in the American Economic Review, found that business training’s effects depended on social constraints — it helped most among women whose mobility and decision-making were restricted, precisely because it addressed the constraint rather than the accounting.
Bundles beat components. Oriana Bandiera and colleagues’ evaluation of an adolescent-girls program in Uganda, in the American Economic Journal: Applied Economics in 2020, found that combining vocational skills with life skills in girls’ own clubs raised the likelihood of income-generating activity substantially and improved a range of other outcomes; the combination, delivered in a safe space, did what neither component had done alone. The design lesson recurs across the literature: finance without skills, skills without finance, and either without networks under-perform the bundle.
Rails matter. Tavneet Suri and William Jack’s 2016 work in Science found that mobile money lifted a substantial share of Kenyan households out of poverty, and that the effect ran disproportionately through women moving from agriculture into business — because digital rails gave them savings, payments and credit histories that the physical economy had denied them. In remittance economies where transfers already arrive digitally, this is the cheapest lever available.
Microcredit alone is modest. As article 46 noted, the synthesis of six randomised microcredit evaluations found real but modest effects; microfinance is the delivery channel that reaches districts, not the intervention.
The region’s programs
Most of what exists for women entrepreneurs in Central Asia is financed by development institutions and delivered through banks, microfinance institutions and business associations. The European Bank for Reconstruction and Development‘s Women in Business programs — active in Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan — combine credit lines through partner banks with advisory support and training; the Asian Development Bank and the International Finance Corporation fund similar bundles and gender-focused components in larger projects; the Women Entrepreneurs Finance Initiative, hosted by the World Bank, channels finance and advisory to women-led small firms in the region; and UN Women and national agencies run enterprise and skills programs, particularly for rural women. National instruments — Kazakhstan’s Damu fund has women-focused lending windows; Uzbekistan’s state programs and women’s committees have expanded credit and training for women since the reform decade — add scale. The gaps, across almost all of them, are the ones the evidence would predict: classroom training rather than mentorship, credit without the behavioural component, delivery in provincial capitals rather than districts, and outcome measurement that stops at loans disbursed and women trained.
Two starting points: Kazakhstan’s base and Uzbekistan’s push
The region’s two largest economies begin from different places. Kazakhstan inherited near-universal women’s employment and a large cohort of women in banking, telecoms, retail and the professions; its women entrepreneurs are concentrated in formal urban services and are the region’s most likely to appear in technopark, bank and venture statistics, and its constraint is less participation than growth — the transition from a profitable small business to a scaling one, which is where collateral rules and thin equity networks bite. Uzbekistan’s participation rates are lower and its rural norms more restrictive, but its reform program has made women’s enterprise an explicit priority: expanded credit lines through state banks and microfinance, training programs run through the country’s district-level women’s committees, and quotas and preferences in several enterprise schemes. The push has reached large numbers of women quickly; its weakness, as with most of the region’s programs, is that it has delivered loans and courses rather than mentors, bundles and markets, and that it measures disbursement rather than growth. The program design that follows fits both starting points, with different emphases — growth finance and networks in Kazakhstan, district reach and bundled delivery in Uzbekistan and the mountain republics.
The digital track: women founders in the technoparks
A distinct and growing group of women founders exists in the region’s digital economy — in online retail, education technology, design and content, software services and, increasingly, product companies — and they are the women the technoparks and venture funds can actually reach. Their constraints are different: not collateral or mobility but the thinness of women’s networks in a male-dominated technology sector, the absence of women among mentors, investors and judges, and the design of programs around full-time, residential, evening-heavy formats that assume no care responsibilities. The fixes are inexpensive: women mentors and investors recruited onto benches and committees with targets rather than hopes, cohort schedules and formats that fit care, parity targets in selection that are reported, and connection to the region’s women’s business associations and to the corridor’s women-in-tech networks in the Gulf and India. This track deserves to exist; it should not be mistaken for the whole strategy.
Measuring what matters
Women’s enterprise programs are, more than most, judged by inputs — women trained, loans disbursed, events held — because those are the numbers funders ask for and the numbers that are easy to produce. The measurement architecture set out in article 32 applies with one addition: because women’s businesses are more often informal and home-based, the outcome survey must be designed to see them. Survival and income should be measured for the business the participant actually runs, formal or not; formalisation should be tracked as an outcome rather than assumed as a precondition; employment should count family and part-time workers; and control over income — the share of business income the woman herself decides how to use — should be asked, because the literature is clear that it is where much of the welfare gain from women’s enterprise sits. Where a program is oversubscribed, the qualified applicants who did not get a place are the comparison group, and the difference between them and participants at 24 months is the only number that will ever settle the argument about whether the program worked.
A program design that starts where women’s businesses are
- Recruit in districts, around existing businesses. The founder to recruit is the woman already running the household’s shop, workshop or plot — not the woman who can travel to a pitch competition. Microfinance institutions, women’s associations and cooperatives are the recruiting channels; the district, not the capital, is the venue.
- Mentors from the same trade. Experienced women entrepreneurs in the same sector and district as mentors, with obligations — on the Nairobi model — rather than visiting speakers from the technopark.
- Behaviour-based curriculum, time-designed. Personal-initiative and customer-discovery content delivered in short sessions at times and places that fit care responsibilities, in Uzbek, Kyrgyz, Tajik, Kazakh and Russian; childcare provided where the program meets.
- Finance bundled and unblocked. Credit or matched savings delivered by the host microfinance institution alongside the training, with collateral substitutes — group guarantees, cash-flow lending, movable-asset registries where they exist — so that property registered to a husband is not the gate.
- Digital rails as infrastructure. Mobile-money accounts, digital bookkeeping and online marketplaces as program components, because they are the route to savings, credit histories and customers beyond the district.
- Market linkage. Buyers — textile clusters, retailers, exporters, tourism operators, corridor traders — engaged as program partners with purchase commitments, on the demand-side model this series uses everywhere.
- Measurement that donors and ministries will fund. Business survival, income, employees and formalisation at 12 and 24 months, by cohort, with a comparison group from oversubscription (article 32).
The design deliberately does not centre on the technopark. Women founders in the digital economy exist and deserve programs, and the technoparks’ regional branches are the right hosts for them; but a women’s enterprise strategy that starts with software reaches a few hundred women and misses a few million.
A composite case: the loan window that became a program
A composite from several regional engagements; details altered.
A microfinance institution in a southern district had run a donor-funded women’s lending window for several years: thousands of small loans, excellent repayment, and — when a funder asked — no evidence that the borrowers’ businesses had grown. Borrowers used the loans as working capital and repaid from the same businesses at the same size.
The redesign kept the window and wrapped a program around it. Borrowers with a year’s history were invited into district cohorts of twenty-five, meeting in short sessions at the institution’s branches with childcare provided. The curriculum was personal initiative and customer discovery, not accounting. Each cohort was mentored by three established women entrepreneurs from the district’s main trades — garments, food, retail — paid for their hours. A second, larger loan tranche was released to members who reached a milestone, against cash flow rather than collateral; a mobile-money and bookkeeping module gave every member a digital record; and a buyers’ day connected the garment producers to a regional textile cluster and the food producers to two retail chains. At 24 months, the cohort’s businesses were larger, more of them were formal, and the institution’s own data showed the difference against borrowers who had not joined. The funder financed a second district, and the ministry — for the first time — asked for the cohort’s numbers by name.
What could go wrong
- Programs designed for the capital. Antidote: district delivery through microfinance institutions and associations.
- Classroom training as the intervention. Antidote: mentors from the same trade; behavioural curriculum.
- Credit gated by collateral women do not own. Antidote: cash-flow lending, group guarantees, matched savings.
- Time-blind schedules. Antidote: short sessions, local venues, childcare.
- Technopark-first design. Antidote: sector realism; digital programs as one track, not the whole.
- Loans disbursed counted as impact. Antidote: survival, income and employment at 12 and 24 months.
Questions ministers and donors actually ask
“Why do our women’s programs show high participation and little growth?” Because they deliver classes and loans, and the evidence says growth comes from mentors, behaviour, bundles and markets.
“Where are the women founders for our technopark?” A few hundred are in the digital economy and can be recruited through regional branches and universities. The million-plus are in the districts, running businesses the technopark does not count.
“What does collateral reform have to do with it?” Everything. Cash-flow lending, movable-asset registries and matched savings unblock the finance that property rules have gated.
“What is the role of the corridor?” Buyers and models: Gulf and Indian retailers, textile and food importers as demand-side partners, and the South Asian and Gulf women’s enterprise programs — self-help groups, women’s business councils — as design references.
“What is the one number?” Income and survival of participants’ businesses at 24 months against a comparison group. Loans disbursed and women trained are inputs.
Methodology & data notes
Labour-force participation figures are approximate ILO modelled estimates published by the World Bank for recent years and are sensitive to how informal and home-based work is counted; they understate women’s economic activity for the reasons the text explains. Enterprise Survey ownership shares vary by country, year and definition and are cited only as orders of magnitude. Findings from the randomised literature are summarised approximately; readers should consult the papers for effect sizes and settings, and should note that most were conducted outside Central Asia. Program descriptions reflect public information as of 2025. The composite case combines several engagements with details altered. Companion articles cover remittance economies (44), finance (46), founder readiness (40) and program design (50).
References & further reading
- Brooks, W., Donovan, K. & Johnson, T. (2018) — Mentors or Teachers? Microenterprise Training in Kenya, American Economic Journal: Applied Economics
- Field, E., Jayachandran, S. & Pande, R. (2010) — Do Traditional Institutions Constrain Female Entrepreneurship? A Field Experiment on Business Training in India, American Economic Review
- Bandiera, O. et al. (2020) — Women’s Empowerment in Action: Evidence from a Randomized Control Trial in Africa, American Economic Journal: Applied Economics
- Klapper, L. & Parker, S. (2011) — Gender and the Business Environment for New Firm Creation, World Bank Research Observer
- Campos, F. et al. (2017) — Teaching personal initiative beats traditional training in boosting small business in West Africa, Science
- Suri, T. & Jack, W. (2016) — The long-run poverty and gender impacts of mobile money, Science
- World Bank Open Data — labour-force participation rate, female (ILO modelled estimate)
- World Bank Enterprise Surveys — gender indicators for Central Asian firms
- EBRD — Women in Business programs in Central Asia
- Women Entrepreneurs Finance Initiative and UN Women — women’s enterprise finance and programs
- Asian Development Bank and IFC — gender-focused SME programs in the region
HexGn designs women’s enterprise programs that start where the businesses are — district cohorts, mentors from the same trade, finance bundled with behavioural training, digital rails and measured outcomes — for governments and development partners across Central Asia.