Bangladesh
Women in Freelancing · Bangladesh
We trained a thousand women.
We trained a thousand women.
Then we proved it worked.
Most training programmes report their graduates. In Dhaka, Learnio did something harder. We handed the shortlist to BRAC University's research institute, which randomly assigned 900 women to the programme or to a control group that received nothing. Two years later the trial reported 54% higher income, 184% higher freelancing employment, and an effect that grew with every class attended.
Trained
1,000 women
Monthly income
+54% vs control
Funded by
Independently evaluated by BIGD, BRAC University
Independently evaluated by BIGD, BRAC University
1,000
Young women trained free of charge, across four campuses in Dhaka
900
Randomly assigned to the programme or to a control group by BRAC University
54%
Higher monthly income for trained women than for the control group
42%
Freelancing among women who attended half the classes, against 12% in the control
Bangladesh is the second largest supplier of online labour in the world, behind only India. By 2017 it held roughly a sixth of the global market, with some 650,000 registered freelancers. Women were nine per cent of them.
The gap was never about ambition. Commuting, security, workplace harassment, and the unpaid load of household and caregiving work keep educated Bangladeshi women out of offices. The country ranks 50th of 153 on the overall gender gap, and 141st on economic participation. Online freelancing pays from home on flexible hours, which ought to make it the obvious way in. Yet the women who most needed it had never been shown the door, could not afford the training, and had no one to sit with them through the first bid.
So a Netherlands-based philanthropy commissioned Learnio, then operating as CodersTrust Bangladesh, to build Women's Skills Development for Freelancing Marketplaces (WSDFM): free ICT training and post-training mentorship for a thousand underprivileged women aged 18 to 35, aligned with SDG target 8.5 and the government's Digital Bangladesh agenda. Then it did the unusual thing. Rather than let the implementer grade its own work, it commissioned the BRAC Institute of Governance and Development to wrap a randomised controlled trial around the programme.
The trial's conclusion, in the evaluators' own framing, was that steady participation in freelancing training can strongly advance the economic empowerment of educated, underprivileged women, and that the model is ready to be scaled in both size and scope. BIGD, BRAC University · WSDFM Impact Evaluation · May 2022
Voices from the cohort
Behind every coefficient,
Behind every coefficient,
a household that changed.
The evaluation counted contracts, hours and taka. The graduates measured it differently: in the first gig, the first payment, and the afternoon their family stopped calling it a waste of time.
From the first cohort to ninety projects delivered on Fiverr.
Samina came into the very first WSDFM batch from a middle-income family that had not expected her to keep studying. She began offering graphic design on Fiverr while still in class, and has since completed more than ninety projects. She now works as a freelancer full time and mentors younger women in her community.
Samina Tasneem Chowdhury
WSDFM graduate · Graphic design track
Two companies, one social media manager, no commute.
Rafat trained on the digital marketing track and now runs social media for two separate companies from home. Learnio filmed her story for the project's own record. Her film is the one at the top of this page.
Rafat Islam Nimme
WSDFM graduate · Digital marketing track
“They are pleased that I am earning.”
A 23-year-old married trainee, followed through the evaluation's qualitative case studies, now earns between ৳10,000 and ৳15,000 a month from local marketplace clients. Her mother-in-law took over the household chores so she could concentrate on the work. Family support, the evaluators found, was the single clearest dividing line between the women who kept freelancing and the women who stopped.
Anonymised, by the evaluators
BIGD case study · Graphic design track
Why this one is different
We didn't just report graduates.
We didn't just report graduates.
We reported a control group.
1,000 women trained · 900 randomised · Evaluated and published by BRAC University
Project · Women's Skills Development for Freelancing Marketplaces
How it worked
Teach the skill.
Then sit with her while she bids.
Two hundred hours across four months, three four-hour classes a week, delivered at four Learnio campuses in Dhaka. Every cohort had three mentors rather than one: a skill mentor for the craft, a buyer communication mentor for English and client handling, and a career advising mentor for the marketplaces themselves. Classroom work was followed by three months of profile building on the major platforms, portfolio review, and hands-on guidance through the first paid assignments.
2019
Design
A Dutch philanthropic funder commissions Learnio to develop a thousand underprivileged women into freelancers. Targeting is set at women aged 18 to 35 with at least a Secondary School Certificate. BIGD is engaged to evaluate by randomised trial.
1,000
Fully funded training places
2019–2021
Recruitment
Facebook and YouTube campaigns, roadshows, and campus drives run alongside partnerships with 20 educational institutions. Three-stage selection: a written test, a computer literacy test, and an interview. Half the shortlist is sent to BIGD to be randomised.
1,450+
Candidates shortlisted
4 months
Training
Fifty lectures: computer fundamentals, then the skill track in Graphic Design or Digital Marketing, eight buyer-communication classes in English, and a dedicated freelancing curriculum on gigs, portfolios, pricing and bidding.
200
Hours per trainee
Mar–Apr 2020
The COVID pivot
Bangladesh locks down mid-rollout. Within a month the whole curriculum is rebuilt for online delivery, every class is recorded for replay, and a 24-hour support channel and alumni group are stood up. Later cohorts are recruited nationally rather than only in Dhaka, widening the programme's reach beyond the capital.
1 month
To take the curriculum fully online
2021–2022
Mentorship, then the verdict
Three months of career advising, profile monitoring and one-to-one guidance close out each cohort. BIGD's follow-up survey reaches 599 of the original 833 respondents, attrition tests clean, and the impact evaluation is published in May 2022 and disseminated at a national workshop in Dhaka.
599
Women surveyed at follow-up
Outcomes
Large effects.
Large effects.
And a dose that explains them.
Because half the shortlist never received the programme, every figure below is a difference against a comparable group of women, not a before-and-after on the same people. These are treatment-on-the-treated estimates from the follow-up survey, expressed against the control group's mean.
Headline metric
54%
Higher monthly income for trained women than for the randomly assigned control group.
Control group
৳5,715/mo
Trained women
৳8,068/mo
Employed after the programme
62% vs 53%
184%
Higher freelancing employment
The programme's core objective. Hours worked in freelancing rose 168% and freelancing income 169%.
43%
Higher household spending
Food spending up 48% and non-food up 41%. More women held savings, and fewer took on debt.
38%
Spillover into other income
Skills learned for the marketplace raised what women earned everywhere else. Use of the internet for income generation rose 96%.
The finding that matters most: attendance.
BIGD split the trained women by how many of the 48 classes they actually turned up to. Nothing else in the study predicted outcomes so cleanly. Not age, not marital status, not the degree they walked in with. Engagement decided who ended up earning, which means the design question for any funder is not who to select but how to keep her in the room.
Share earning from freelancing at follow-up
By classes attended, of 48
Control groupNever offered the programme
৳778Monthly
1 to 10 classesEnrolled, barely attended
৳643Monthly
11 to 24 classesAttended up to half
৳2,083Monthly
25 classes and aboveAttended more than half
৳2,500Monthly
Women who attended more than half the course were three and a half times as likely to be earning from
freelancing as the control group, and earned more than three times as much from it. Women who enrolled and then drifted
away did no better than women who were never offered a place at all. Just over half of trainees cleared the halfway
mark; a quarter attended every class.
26%
Higher overall employment among trained women than in the control group.
28%
More trained women now name freelancing as the occupation that suits them best.
76%
More trained women reported not one day of financial crisis in the previous month.
60%
Gain in multimedia software skill, the largest of seven measured software competencies.
What the trial did not find
An honest reading of the evaluation has to include the nulls. The trial found no significant effect on self-reported confidence or aspiration, and only mixed movement on the women's empowerment indicators. Uptake of the post-training advising sessions was low: fewer than four in ten trained women attended even one. The evaluators attribute the muted effects to the participants' social and economic context rather than to the programme design, and they note that wherever mentoring did land, outcomes improved further. COVID-19 arrived in the middle of delivery, and the study was confined to Dhaka. We publish these limits because a funder who is not told them cannot trust the numbers that sit above them.
Verified & documented
Read the public record.
WSDFM was designed, monitored, independently evaluated and published. The 92-page impact evaluation was written by an eight-person team at the BRAC Institute of Governance and Development, released in May 2022, and presented at a national workshop on inclusive ecosystems for women freelancers. It is now cited in the wider literature on women's labour force participation in Bangladesh, including on VoxDev. The links below take you to the primary sources.
Report · May 2022
Impact Evaluation Report of the WSDFM Program
Read the full evaluation →
Publication · BRAC University
BIGD publication record, with the full author team and abstract
View on bigd.bracu.ac.bd →
Case study · Learnio
WSDFM Project: Let a Thousand Flowers Bloom
Read the case study →
Dhaka Tribune · Apr 5, 2022
Why is there a lack of women in freelancing?
Read on dhakatribune.com →
The Business Standard · Aug 2022
Freelancing training can make women self-reliant, study finds
Read on tbsnews.net →
New Age · Editorial
Women's skills for freelancing
Read on newagebd.net →
Note: Learnio operated as CodersTrust Bangladesh throughout the WSDFM project, and the BIGD
evaluation refers to the organisation as CTBD. The company, mission and core team are continuous. The evaluation
anonymises the funder as a private philanthropy based in the Netherlands; Learnio's own project record names it as
Porticus.