Women's share of the labor market ranges from roughly one in five workers to near parity depending on the country. Women approach near-parity in Europe and North America, making up roughly 47% to 48% of the total workforce. They also constitute nearly half of all workers in Sub-Saharan Africa. Meanwhile, in South and West Asia, as well as North Africa, women constitute only one in five participants in the formal workforce. Culture, infrastructure, policy, and social norms all shape who gets to work. The rise of gig work in the past decade has given women more opportunities to enter the workforce, particularly in countries where societal norms largely limit their participation.
Freelancing offers flexibility and autonomy. Women can set their own schedules, choose which clients to work with, and often work from home without face-to-face interactions. But as we showed in our earlier analysis of the gender wage gap in US gig work, a gender gap in pay persists even when workers set their own prices. On the freelancing platform Upwork, where workers set their own prices, men charge more per hour than women with equivalent experience and reviews. The mechanisms differ from other forms of gig work like driving for Uber, but the gap persists.
This week, Revelio Labs uses data on active Upwork accounts from different countries to measure the gender pay gap in freelancing across borders. Gender is inferred from account holder names using our gender inference model. Because every freelancer in the sample competes on one platform under one set of rules, the comparison holds the market structure constant and isolates what differs between countries. The sample is limited to the 100 countries with at least 5,000 active accounts.
What share of gig economy workers are women?
Globally, women own over 40% of total active accounts on Upwork, which closely resembles the share of women in the global workforce. But that global average masks enormous variation across countries, as shown in the plot below. In some countries, women dominate Upwork accounts, holding over 60% of active accounts. High- and upper-middle-income countries dominate the top half of the plot, with more than 50% of the active accounts created by women. Meanwhile, low- and lower-middle-income countries are common at the bottom of this distribution. In countries like Nepal, Uzbekistan, and Yemen, women account for less than a quarter of active accounts on Upwork.
This is surprising, as one might think freelancing platforms could help women enter the workforce in countries where formal labor markets fail them. The social barriers that typically keep women out of the office largely disappear when women can accept jobs they are comfortable with and do from home without face-to-face interactions. Many factors can explain this variation in the share of women on Upwork. Internet access and digital literacy gaps play a role in women's participation on freelancing websites, as does the availability of payment infrastructure that lets women receive earnings independently. In some countries, cultural expectations around women managing money or working with foreign clients can make it harder for them to even access the digital infrastructure required to participate in the gig economy.
Does freelancing pull more women into the workforce?
We look at the relationship between the female share of Upwork accounts and the share of women in the workforce for the countries in the sample. We observe a positive correlation between the share of women in the workforce and their presence on Upwork, with a correlation of 0.4. We also observe that in many countries, women are overrepresented on the gig platform relative to their share in the workforce. Along the diagonal line, the share of women in the workforce exactly equals the share of Upwork accounts held by women. A larger number of countries in our sample cluster above the diagonal line, indicating that relative to their share in the workforce, women are overrepresented in running accounts on the gig platform. In Lebanon, for example, women make up about 28% of the workforce but about 40% of Upwork accounts. Lebanon is a country we have flagged before: our 2020 analysis of how global the gig economy has become found it had the highest freelancer share of any country in the world, at 2.4% of its workforce, nearly double its level a decade earlier. Interestingly, there is no clear pattern when we classify countries by the level of income. Women are overrepresented on the gig platform in both high- and upper-middle-income countries and low- and lower-middle-income countries.
Is there a gender pay gap in gig work?
Getting onto the platform is only half the story. In a previous newsletter, we showed that women on Upwork set lower hourly rates than men with equivalent levels of experience and ratings, documenting a gender wage gap that persists even when workers set their own prices. Here, we ask whether it holds up across countries.
Across most countries in our sample with at least 5,000 active accounts, after controlling for role, experience, and ratings, women quote lower hourly rates than men. The exceptions are France, where the median quoted hourly rate for men and women is similar, and the United Arab Emirates and Ireland, where women quote a higher median hourly rate than men, after controlling for experience, role, and reviews. The size of the gap varies significantly. Bangladesh stands out with a gap of about 30%: Women in Bangladesh quote hourly rates that are 30% lower than those listed by men. In the United States, women list hourly rates 26% lower than men. Interestingly, in the US, the gender pay gap in gig work is higher than the overall gender pay gap of about 15%.
Countries with higher female labor force participation tend to show a smaller gender pay gap on Upwork. This runs counter to what research on traditional labor markets would predict. Academic research on OECD countries finds that factors which suppress women's lifetime labor force participation, including higher fertility rates, larger husband-wife age gaps at marriage, and higher top marginal tax rates, are associated with larger, not smaller, gender wage gaps. On Upwork, the opposite pattern holds. One explanation is that freelance platforms introduce a selection mechanism that can reverse this relationship in the gig economy specifically. In countries where women have a harder time accessing formal employment, the women who make it onto freelance platforms may be more positively selected: more motivated, more educated, and more entrepreneurial than their counterparts within these countries. This selection effect could compress the gender pay gap in freelancing even when female labor force participation is low. This intuition finds some support from other studies across European countries and within regions in Turkey, which similarly find that higher female labor force participation does not always translate into a smaller gender wage gap.
If this selection story is right, we might expect to see evidence of it in how much women work on the platform: More capable or more committed freelancers could reasonably work more hours than their male counterparts. But hours worked are not simply a measure of effort or selection. They also respond to economic forces that are likely to pull the effect in opposite directions. On the one hand, the substitution effect predicts that as female freelancers' pay approaches men's, they will work more hours to capture the benefit of that higher return to their time. On the other hand, the income effect predicts the reverse: as female freelancers earn pay closer to men's, they may not need to work as much to reach a given income target, freeing up time for competing priorities like childcare, home production, and so on. Because these two forces push hours in opposite directions depending on the country's wage gap, there's no clean prediction for how hours and the wage gap should relate. The effect depends on which force dominates.
Empirically, the relationship between the gender gap in hours worked and the gender pay gap across countries is essentially flat (r = 0.07), indicating that the income and substitution effects cancel each other. Bangladesh, Nigeria, and France all sit at a similar point on the hours axis, where women work roughly 40 to 45% more hours than men, yet their wage gaps range from under 1% in France to nearly 18% in Nigeria to over 30% in Bangladesh. At the other end, countries like Egypt and Pakistan, where women work closer to parity in hours (or even fewer hours than men), still show wage gaps in the 15–20% range. This is consistent with the substitution and income effects roughly offsetting one another across our country sample: Women may end up working similarly elevated (or similarly modest) hours in both high-gap and low-gap countries, but for entirely different underlying reasons depending on which force dominates locally. The cross-country correlation alone can't distinguish between these explanations, though. A rigorous test involves testing within-country variation over time in the announced hourly rates and hours worked. Hours worked, in other words, is not a clean proxy for positive selection, since it can equally reflect the opposite dynamic: women working more simply to compensate for lower pay, or working less because higher pay has already let them scale back.
Does education explain the gender pay gap in freelancing?
Countries where more women have post-secondary education tend to show smaller gaps in the median hourly rates quoted on Upwork. Two mechanisms could plausibly connect education to how women price their work. The direct mechanism is straightforward: A college education builds technical, analytical, and communication skills that raise a freelancer's actual productivity and market value, which should translate into a higher rate regardless of gender. The indirect mechanism runs through confidence and pricing behavior rather than skill itself. Women who finish their bachelor's degrees are more likely to have operated in professional or academic environments where negotiation, self-advocacy, and quoting one's own value are normal, expected behaviors. On a platform where the hourly rate is often the first and only signal a client sees, that comfort with self-pricing can translate directly into dollars, independent of any underlying skill gain.
The pattern holds across very different country contexts. France and the United Arab Emirates, where female post-secondary attainment is high, show some of the smallest gaps in our sample. Meanwhile, Bangladesh and Pakistan, where the share of women with college degrees remains low, show some of the largest gaps, both exceeding 20%.
This doesn't mean education is the only factor, or that closing the gap is simply a matter of getting more women into universities. The United States is a reminder of that: Despite relatively high female educational attainment, the gig hourly pay gap sits above 25%, on par with much lower-education countries like Bangladesh.
What the data says about women in the gig economy
The data on women's status in freelancing tells two stories at once. Freelancing is opening doors for women in countries where formal labor markets offer them little. The flexibility and lack of physical barriers make gig work more accessible than most formal work options. But access is not equality. The gender wage gap persists in most countries in our sample, even after controlling for role, experience, and ratings. Education appears to be the strongest country-level predictor of how large that gap is, suggesting that what women bring to the negotiating table matters as much as the platform itself.
