Jan 22, 2025 Gender Enterprise Note No. 45 Understanding Women’s Lower Participation than Men as Workers, Top Managers, and Owners in Private Firms in the EU-27 Countries* Mohammad Amin T his Brief examines issues related to women’s participation as workers, top managers, and owners of private rms in 27 European Union countries (EU-27), using the rich database of the World Bank Enterprise Surveys. e analysis focuses on EU regions varying between about 800,000 and 3 million inhabitants (NUTS2-level groupings). Overall, women’s participation as workers, top managers, and rm owners is statistically signi cantly less than that of men. Surprisingly, richer NUTS2 regions experience a larger gender gap favoring men in employment, top manager positions, and rm ownership. Another worrying feature is that relative to men, women workers tend to be concentrated in rms that are less productive and pay low wages. us, closing gender gaps in income requires not just more jobs but also better quality of jobs for women. Having a woman as the top manager of the rm is associated with a higher share of women workers in the rm, but this e ect is much stronger when the rm initially has a relatively high share of women workers. A gender gap also exists in labor productivity, which is lower for women-run rms than men-run rms, and for rms with higher women’s ownership. ese gaps in labor productivity are much larger at lower quantiles of labor productivity, implying the presence of “sticky oors” but not necessarily “glass ceilings” in the EU-27 countries. e Brief identi es some of the factors that are correlated with the average gender labor productivity gap and estimates their contribution to the gap . ere is no systematic di erence in the level of constraints, including access to nance, faced by men-run versus women-run rms and/or by rms at di erent levels of women’s ownership. Understanding women’s lower participation as relationship between the gender of the top manager and the workers, top managers, and firm owners in EU labor productivity of the rm, the controls ensure that this relationship is not because rms with women top managers firms di er from rms with men top managers in terms of size, age, is Brief examines women’s participation in the private exports, foreign ownership, industry, and being part of a sector as workers, top managers, and rm owners.1 Women’s larger establishment. e baseline controls consist of industry participation as workers is measured by the share of workers dummies (at the ISIC 2-digit level); the log of age of the rm; in a rm who are women. Women’s participation as top the log of the number of workers at the rm; a dummy manager is based on the top manager of the rm being a indicating whether the rm exports directly or not; a dummy woman. Women’s participation as rm owners is measured indicating whether the rm has foreign owners or not; and a by the share of the rm that is owned by women, as well as a dummy indicating whether the rm is part of a larger dummy variable indicating women’s share above a speci ed establishment. level. Across the EU-27 regions, women’s participation as All rm-level regression results discussed in the Brief workers, top managers, and rm owners is statistically account for baseline controls that ensure that the relationship signi cantly lower than men’s ( gure 1). For a typical rm in under consideration is not spuriously driven by di erences in the region, a mere 35.3 percent of workers are women. e the baseline rm characteristics. For instance, for the share of rms that have a woman top manager is 18.1 *A liations: World Bank, Development Economics, Enterprise Analysis. For correspondence: mamin@worldbank.org. Acknowledgments: is Brief is a part of a series focusing on issues of regional disparities and growth opportunities in the EU-27 area. e series is a product of the World Bank’s Enterprise Analysis team (DECEA) and has bene tted from generous support from the EU DG REGIO directorate. e team would also like to thank Norman V. Loayza and Jorge Rodriguez Meza for comments and guiding the publication process. Nancy Morrison provided excellent editorial assistance. Objective and disclaimer: e ndings in this series of Briefs do not necessarily represent the views of the World Bank Group, its Executive Directors, or the governments they represent. All Briefs in the series can be accessed via: https://www.worldbank.org/en/research/brief/global-indicators-briefs-series. Enterprise Note No. 45 Figure 1 Women’s participation in private firms in the EU-27 is lower than that of men 60 55 51 50 45 44 41 40 40 39 40 39 38 38 40 35 36 37 37 37 35 32 32 33 33 34 29 30 30 26 21 22 Percent 21 21 22 20 20 18 20 20 22 19 17 18 18 16 18 17 15 16 15 15 12 10 0 Retail Small Medium Manufacturing Transition Most developed Young (<=10 years) Other services Old (>10 years) All firms Large Least developed All Firm size Industry Income level Age of firm Women workers Women top manager Women owners present Women’s ownership Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. percent. e share of rms having one or more women in the least developed NUTS2 regions than in the most owners is 39.9 percent, with the share of the rm owned by developed regions. Women’s participation as top managers is women averaging 22 percent. us, overall, there is a statistically signi cantly higher in the least developed signi cant gender-based gap in workers, top managers, and NUTS2 regions than in the most developed ones. Women’s rm owners throughout EU-27 countries. is result holds share in rm ownership is statistically signi cantly higher in separately for rms of di erent sizes and age, and for rms in the least developed and transition NUTS2 regions than in NUTS2 regions at di erent income levels (see gure 1). One the most developed regions. exception is the retail sector. In this sector, the share of ere are several potential reasons for the di erence in women workers is statistically signi cantly higher than that women’s participation across income groups. In the sample of men, and rms with a woman owner are as likely as rms of rms with data available for the controls listed above, the with no women owners. share of women workers in the least developed NUTS2 regions is higher by 5.6 percentage points than in the Women’s participation by income levels transition and most developed regions (the “gap”). Using a decomposition known as the Kitagawa-Oaxaca-Blinder e literature is ambiguous about the relationship (KOB) (Kitagawa 1955; Oaxaca 1973; Blinder 1973), it is between economic development and women’s labor market possible to identify several factors that contribute to this gap participation. Some studies suggest a nonlinear relationship, and are statistically signi cant. Table 1 provides the details. with women’s participation as workers increasing during the initial phases of development and falling thereafter (see Some factors widen the gap. Goldin 1995; Gaddis and Klasen 2014). Structural • Country-speci c factors. Di erences in the level of transformations associated with economic development, country-speci c factors widen the gap by about 85 percent institutions that protect women’s interest, and work-leisure of the total gap (a di erence of 4.7 percentage points of trade-o s are some of the driving factors. Across NUTS2 the total of 5.6 percentage points). regions in the EU-27 countries, the share of women workers • Greater prevalence of women top managers and higher share in a rm, the share of rms having a woman top manager and of rms with women owners in the least developed NUTS2 the share of rms with a woman owner decline statistically regions than in the other regions. ese di erences signi cantly with income per in habitant of the NUTS2 together widen the gap by 0.9 percentage points, or about regions. ese negative relationships hold at the rm level 16 percent of the total gap, because women’s employment and the NUTS2 level. Figure 2 shows these relationships at tends to be higher in rms with women top managers and the NUTS2 level. in rms with higher women’s ownership. ese results are observed across the broad income groups • Higher prevalence of retail rms in the least developed (shown in gure 1), but not within groups on average. For regions. Retail rms employ more women relative to men instance, at the rm level and at the NUTS2 level, the share than the other sectors, and there are proportionately more of women workers in a rm is statistically signi cantly higher retail rms in the least developed regions than in the other 2 Enterprise Note No. 45 Women’s participation as workers, top managers, and firm owners declines with income per Figure 2 inhabitant across NUTS2 regions a. Share of women workers in a firm b. Share of firms with a woman top manager c. Share of the firm owned by women Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. regions. is widens the gap by 0.63 percentage points, or Taken together, these results provide a useful starting by more than 11 percent of the total gap. point for understanding the causal factors for the gender gap in employment observed between least developed NUTS2 Other factors narrow the gap. regions and the rest. • Di erential e ects of retail rms across least and more developed regions. An increase in the share of retail rms Labor productivity effects increases the share of women workers in a rm much more in the transition regions and most developed regions than Consistent with the broader literature, the labor in the least developed regions. is di erence narrows the productivity of women-run rms is statistically signi cantly gap by 0.83 percentage points, or about 15 percent of the lower than that of men-run rms. It is lower by 0.29 log total gap. points, or about 25.2 percent; this gap is reduced to 16.5 • E ects of research and development. R&D activity increases percent once the baseline controls are included. e the share of women workers in the transition and most di erence in these two estimates of the labor productivity gap developed countries but lowers it in the least developed is almost entirely due to the industry dummies. A similarly countries. is di erence narrows the gap by about 1.35 large and statistically signi cant labor productivity gap is percentage points, or 24 percent of the total gap. found at the NUTS2 level ( gure 3, panel a) and the country • Prevalence of foreign owners. Firms in the transition and level ( gure 3, panel c). at is, countries and NUTS2 most developed regions are more likely to have foreign regions with proportionately more female-managed rms owners, and such rms tend to hire proportionately more have lower labor productivity, on average. At the rm level, women workers. is e ect narrows the gap by 0.17 NUTS2 level, and the country level, the labor productivity percentage points, or about 3 percent of the total gap. gap continues to be signi cant and large even after 3 Enterprise Note No. 45 Decomposition results for women's employment gap between least developed NUTS2 Table 1 regions and the rest (1) (2) (3) Endowment Structural e ects e ects Average share of women workers in a rm in 33.482*** Transition and most developed regions (%) (0.751) Average share of women workers in a rm in least 39.053*** developed regions (%) (1.433) Women's employment gap (women workers’ share in a 5.571*** rm in the least developed regions minus the same in the (1.615) transition and most developed regions) Country dummies (normalized) 4.896*** -0.404 (1.421) (0.685) Retail sector dummy 0.628** -0.831* (0.279) (0.494) Number of workers (logs) 0.002 3.590 (0.023) (3.083) Age of the rm (logs) -0.151 3.346 (0.424) (4.948) Foreign ownership Y:1 N:0 -0.165* 0.173 (0.088) (0.184) Exporter Y:1 N:0 -0.097 -0.149 (0.130) (0.530) Multi-establishment rm Y:1 N:0 -0.568*** 1.343*** (0.184) (0.399) Proportion of rm owned by women 0.300* -1.298* (0.159) (0.699) Woman top manager Y:1 N:0 0.611** 0.657 (0.258) (0.506) Firms spent on R&D last year Y:1 N:0 0.378 -1.315** (0.260) (0.525) Crime losses Y:1 N:0 -0.351* 0.236 (0.185) (0.479) Constant -5.260 (4.566) Total (aggregate) 5.483*** 0.088 (1.602) (0.862) Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Note: Huber-White robust standard errors clustered on country. All coefficients shows are percentage points. A positive coefficient implies a bigger difference or gap in the share of women workers in a typical firm in favor of the least developed NUTS2 regions versus the rest. NUTS = Nomenclature of Territorial Units for Statistics. *** p<0.01, ** p<0.05, * p<0.1 accounting for di erences in income per inhabitant between point relative to men. Studies have also found larger gender the NUTS2 regions ( gure 3, panel b) and GDP per capita gaps at the lower end of the income distribution (Booth, between countries ( gure 3, panel d). Francesconi, and Frank 2003; Duraisamy and Duraisamy 2016). is is the “sticky oors” e ect. It implies that women “Glass ceiling” and “sticky oor” e ects face more challenges initially than later in the process of e broader literature on gender gaps in labor markets development relative to men. suggests that the di culties that women face relative to men ere is little research on the causes of “glass ceilings” and may vary by income (in the case of workers) and labor “sticky oors” e ects. Bertrand (2018) summarizes possible productivity level (in the case of rm managers). Several explanations for the “glass ceilings” e ect as follows: greater studies report a sharp increase in gender gaps (favoring men) discrimination against women at higher levels of at the upper tail of the income distribution (Bertrand 2018; performance; relatively less technical education among Blau and Kahn 2017). is is the “glass ceilings” e ect. It women that hinder their success, particularly at the top level; implies that women’s prospects are limited after a certain greater risk aversion among women that disadvantages them 4 Enterprise Note No. 45 Figure 3 Women-run firms have lower labor productivity than men-run firms in the EU-27 a. Gender labor productivity gap at the b. Gender labor productivity gap at the NUTS2 NUTS2 level level after controlling for income per inhabitant Labor productivity (logs), residuals Labor productivity (logs) Percent of women-run firms Percent of women-run firms, residuals c. Gender labor productivity gap at d. Gender labor productivity gap at the country the country level level after controlling for GDP per capita Labor productivity (logs), residuals Labor productivity (logs) Percent of women-run firms Percent of women-run firms, residuals Source: Original calculations for this Brief based on World Bank Enterprise Surveys (WBES) data. at the highest level of performance; and greater demand on reducing gender discrimination and the di culties that time at higher levels of performance. Booth (2009) argues women face in running their businesses. However, across that “sticky oors” may exist because women at the bottom NUTS2 regions, the gender-based labor productivity gap is of the wage/earnings distribution may have less bargaining higher in relatively richer regions. For instance, at the power or be more vulnerable to a rm’s market power than NUTS2 level, the labor productivity of a rm with all men comparable men. is could be due to unobservable family owners is higher than that of a rm with all women owners obligations or social customs that prioritize men’s careers. by 0.49 log points in NUTS2 regions with below-median Figure 4 examines these e ects. e gure shows the labor income, but this result is statistically insigni cant. e productivity gap (measured as log di erences, favoring men) corresponding gure for NUTS2 regions with above-median at the 5th, 10th, …, 95th quantiles of labor productivity income is much higher, equaling 2.27 log points, and is across the EU-27. ere is a clear downward trend in the statistically signi cant at the 1 percent level. productivity gap, moving from lower to higher labor productivity quantiles. e gap is about twice as large at the Factors associated with lower productivity of lowest quantiles than at the highest quantiles. It is statistically women-run compared to men-run rms signi cant between the 5th and 80th quantiles and ere are several potential reasons or channels by which insigni cant above the 80th quantile. e gure indicates labor productivity is lower for women-run than men-run that women top managers of rms in the EU-27 countries rms. e KOB decomposition methodology helps identify appear to face “sticky oors” but no “glass ceilings.” some possible channels. e results are provided in table 2. Economic development is often seen as a vehicle for e focus here is on select variables known to a ect the labor 5 Enterprise Note No. 45 The sticky floors effect: The labor productivity gap (favoring men) between men-run and women-run Figure 4 firms is larger at lower levels of labor productivity 0.50 Log of labor productivity of men-run firms minus log of 0.45 labor productiivty of women-run firms 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 5th 10th 15th 20th 25th 30th 35th 40th 45th 50th 55th 60th 65th 70th 75th 80th 85th 90th 95th Quantiles of log of labor productivity Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Note: The gap shown is statistically significant at the 1 percent or 5 percent level at the 5th to 80th quantile values. It is insignificant at the 10 percent level or less above the 80th quantile. The estimates shown are obtained after accounting for all the baseline controls. productivity gap between women-run and men-run rms. As is Fourth, men-run rms are much more likely to export and evident from the table, the labor productivity of men-run rms have quality certi cation than women-run rms. ese is higher by 0.28 log points, or about 32 percent. is di erences signi cantly widen the gap by about 0.026 log productivity gap is signi cant at the 1 percent level. Several points and 0.016 log points, or by 9 percent and 6 percent of individual factors contribute statistically signi cantly to this gap. the total gap, respectively. First, about 27 percent (0.075 log points of the total 0.28 Fifth, labor productivity among men-run rms is gap) of the gap is attributable just to the country in which a signi cantly adversely a ected by nancial constraints. men-run or women-run rm is located (country-level However, there is no such impact on women-run rms. As a dummies). at is, a large part of the productivity gap is result, the productivity gap is narrowed by 0.023 log points, because women-run rms are disproportionately or about 8 percent of the total gap. concentrated in those countries where country-speci c Last, the share of the rm that is owned by women is factors lower labor productivity. much higher among women-run rms than men-run rms. Second, in contrast to country-speci c factors, is widens the gap as rms with a higher women’s industry-speci c factors (manufacturing, retail, and other ownership tend to have lower labor productivity. e gap is services) narrow the productivity gap by 0.05 log points, or widened by about 0.37 log points, or 131 percent of the total 18 percent of the total gap. is happens because labor gap. Further, a higher share of women in rm’s ownership productivity is highest in the retail sector, followed by other leads to a large decline in labor productivity among services, and lowest in the manufacturing sector. is women-run rms, but there is no impact on men-run rms. di erence narrows the gap because relative to men, is di erence widens the gap by 0.054 log points, or about women-run rms are most concentrated in the retail sector. 19 percent of the total gap. ird, heavier regulation, measured by the percentage of senior management’s time spent in dealing with government Men-run firms outperform women-run firms in regulations, is associated with a signi cant and much larger other performance measures decline in labor productivity of women-run rms than men-run rms. is di erence narrows the productivity gap Some other performance indicators also reveal a by 0.10 log points, or about 36 percent of the total gap. us, signi cant gender gap favoring men-run rms. ese regulatory burden seems to be an important driver of the indicators include R&D activity, the employment growth productivity gap. rate over the last three years, exporting status, rm size 6 Enterprise Note No. 45 Table 2 Decomposition results for labor productivity gap between women-run and men-run firms (1) (2) (3) Endowment Structural e ects e ects Average labor productivity of men-run rms 11.543*** (0.053) Average labor productivity of women-run rms 11.262*** (0.078) Labor productivity gap 0.281*** (0.065) Industry dummies (normalized) -0.050*** 0.031 (0.015) (0.024) Country dummies (normalized) 0.075* 0.031 (0.039) (0.026) Number of workers (logs) -0.008 0.138 (0.011) (0.140) Age of the rm (logs) 0.006 -0.038 (0.005) (0.175) Foreign ownership Y:1 N:0 0.007 -0.010 (0.005) (0.013) Multi-establishment rm Y:1 N:0 0.005 0.001 (0.004) (0.021) Exporter Y:1 N:0 0.026** 0.014 (0.011) (0.029) Quality certi cation Y:1 N:0 0.016* -0.050 (0.009) (0.034) % of a rm’s senior management’s time spent in 0.012 0.100** dealing with business regulations (time tax) (0.009) (0.043) Share of women in rm's ownership 0.282*** 0.054*** (0.069) (0.016) Fully credit constrained Y:1 N:0 -0.003 -0.023* (0.005) (0.013) Constant -0.335 (0.217) Total or aggregate 0.368*** -0.087 (0.084) (0.075) Number of observations 13,260 Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Note: Huber-White robust standard errors clustered on country. All coefficients shows are percentage points. A positive coefficient implies a bigger difference or gap (favoring men) in the log of labor productivity of men-run and women-run firms. Fully credit constrained firms are identified based on whether the firm applied for a bank loan or not, reasons for not applying, and the outcome of the loan application if it applied. *** p<0.01, ** p<0.05, * p<0.1 (measured by the number of workers (logs), annual sales of total hours of power outages, reported bribe amount (logs). For example, the probability of spending on R&D (percentage of annual sales) that rms usually pay to public activity by a woman-run rm is lower than that of a man-run o cials to “get things done”, percentage of senior rm by 4 percentage points (against the sample mean of 15.3 management’s time spent in dealing with business percent). is di erence is signi cant at the 5 percent level. regulations, and the number of visits or required meetings However, there is no signi cant di erence between with tax o cials. women-run and men-run rms in the overall quality of management practices, the likelihood of formal planning of How women- and men-run firms perceive production, and the likelihood of using manual versus obstacles to running a business automated production methods. Problems and obstacles experienced by women-run rms are not too di erent from e WBES also provides rms’ self-reports on whether those of men-run rms. For instance, there is no signi cant elements of the business environment are a major or bigger di erence between men-run and women-run rms in terms obstacle for their current operations. A mixed picture 7 Enterprise Note No. 45 emerges for these obstacles. For some obstacles, such as tax points and 2.2 percentage points, respectively. Both these rates, tax administration, access to nance, crime, access to declines in women workers are signi cant at the 5 percent land, customs and trade regulations, political instability, and level or less. In gure 5, panels a and b illustrates the point at electricity, there is no signi cant di erence between the the NUTS2 level while panels c and d do so at the country likelihood of a woman-run and a man-run rm reporting the level. e relationships shown in gure 5 continue to hold obstacle as a major or bigger constraint. However, for some even after accounting for di erences in income per other obstacles—such as competition from informal rms, inhabitant and/or the share of tertiary educated adults in the obtaining licenses and permits, transport, functioning of NUTS2 regions or countries as applicable. However, there courts, inadequately educated workers, labor regulations, and are sharp di erences by industry. e greater presence of corruption—women-run rms are signi cantly more likely women workers in lower-productivity and lower-wage rms to report them as a major or bigger obstacle than men-run is much stronger in the manufacturing sector than in retail rms. and other services sectors. In fact, it is statistically Di culties that women workers face often cause them to insigni cant in the retail sector. be concentrated in low-paying and low- productivity jobs relative to men. is is true at the level of rms, NUTS2 Does exporting encourage women’s employment regions, and countries. at is, at the rm-level, a one in private firms in EU-27? standard deviation increase in log of labor productivity and in the log of the average wage bill is associated with a decrease Several studies have indicated that exporting causes rms in the share of women workers by about 1.8 percentage to hire more women workers. One reason for this is that the Figure 5 There are proportionately more women workers in low-productivity and low-wage NUTS2 regions a. Labor productivity and women workers b. Labor cost per worker and women workers relationship at the NUTS2 level relationship at the NUTS2 level Percent of women workers Percent of women workers Labor productivity (logs) Labor cost per worker (logs) c. Labor productivity and women workers d. Labor cost per worker and women workers relationship at the country level relationship at the country level Percent of women workers Percent of women workers Labor productivity (logs) Labor cost per worker (logs) Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. 8 Enterprise Note No. 45 level of competition is typically higher in international be greatest at the intermediate level of women workers in the markets than domestically. More competition raises the cost EU-27 countries. While this is a plausible explanation, it is of discrimination, leading to higher women’s employment possible that rms with intermediate levels of women (see Amin and Islam 2023). Depending on the country’s workers have certain characteristics that make women comparative advantage, exporting can also increase demand workers especially suitable for exporting purposes. for products where women work. Overall, across all rms in the EU-27 countries, there is a How does having a woman as the top manager positive but statistically insigni cant relationship between affect women’s employment prospects at lower the share of women workers in a rm and rm’s exports as a levels? percentage of its sales. However, there are two important heterogeneities. First, the relationship is much larger and Some studies nd that women top managers are better at statistically signi cant in the manufacturing sector, but weak understanding the productivity of women workers than men and insigni cant in the retail and other services sectors. top managers, which reduces statistical discrimination Second, the relationship is weak and statistically insigni cant against women workers—and thus women in top for rms with a very low and high initial share (quantile) of management can help narrow the gender gap in employment women workers, but large and statistically signi cant for and wages. . However, the impact may not always be linear. rms with an intermediate share of women workers. One For instance, Flabbi, Moro, and Schivardi (2019) nd that reason for the latter result could be that women prefer to for the Italian manufacturing sector, having a woman top work in rms where there is a substantial presence of women manager narrows the gender wage gap at the top of the wage workers to begin with (see Wasserman 2023). us, the distribution and widens it at the bottom, with essentially no impact of exports should be bigger on rms with a high e ect on average. A somewhat similar result obtains for the initial share of women workers. A counter mechanism could gender gap in the workforce across rms in the EU-27 be the limited avenue for an increase in women workers countries. at is, on average, women’s employment is among rms that already have a high share of women higher in rms that have a woman top manager. While this workers. is implies a weaker relationship between exports e ect holds at most points of the women’s employment and women workers for rms with an initially high share of distribution, it is much larger at higher quantiles (excluding women workers. e net of these con icting e ects seems to the 95th quantile) than at lower quantiles ( gure 6). For The share or women workers is higher among women-run than men-run firms, especially at higher Figure 6 quantiles of women workers’ share 18 16 % of workers who are women in women-run firms minus the same in men-run firms 14 12 10 8 6 4 2 0 h th th th th th th th th th th th th th th th th th 5t 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Quantiles of the share of women workers in a firm Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Note: The difference in women’s employment shown is statistically significant at the 1 percent level at all the quantiles. The estimates shown are obtained after accounting for all the baseline controls. 9 Enterprise Note No. 45 Figure 7 Higher women’s ownership of a firm is associated with lower labor productivity 11th 15th 20th 25th 30th 35th 40th 45th 50th 55th 60th 65th 70th 75th 80th 85th 90th 95th Impact on log of labor productivity -0.06 -0.05 -0.04 -0.04 -0.10 -0.18*** -0.19*** -0.21*** -0.22*** -0.23*** -0.23*** -0.23*** -0.25*** -0.27*** -0.29*** -0.28*** -0.29*** -0.33*** Quantiles of log of labor productivity Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Notes: The figure shows the impact on labor productivity gap of an increase in the proportion of women’s ownership of the firm from 0 to 1. Convergence was not achieved for the quantile regression at the 5th and 10th quantile values of labor productivity (logs). The estimates shown are obtained after accounting for all the baseline controls. *** p<0.01, ** p<0.05 instance, moving from a man-run to a woman-run rm signi cant impact of higher women’s ownership on the increases the share of women workers in a rm by 11.5 likelihood that a rm has an overdraft facility or an percentage points at the 20th percentile value of women’s outstanding loan; uses banks to nance investment; uses employment and by 16.6 percentage points at the 80th banks to nance working capital; or the proportion of percentile value. working capital and/or investment nanced by external sources and/or banks. e same holds for the likelihood that Women’s ownership and productivity a rm is nancially constrained, based on whether the rm relationship applied for a loan last year, the reason for not applying if it did not apply, and the outcome of the loan application. Higher women’s ownership of a rm is associated with ere are a couple of di erences when analyzing these lower labor productivity. is holds for the share of the rm issues at the NUTS2 level. at is, NUTS2 regions with a owned by women: whether majority (more than 50 percent) higher share of women owners on average in a rm have a ownership by women; 10 percent ownership; or more versus statistically signi cantly smaller share of rms that are not less. For instance, labor productivity of rms with majority credit constrained ( gure 8). In contrast, the likelihood of women’s ownership is lower by 0.19 log points, or about using bank nance for investment and the proportion of 17.3 percent, compared to rms with majority men’s investment nanced by banks is statistically signi cantly ownership. However, the negative relationship between labor higher in NUTS2 regions that have a higher share of productivity and each of the three indicators of women’s women’s ownership in the average rm ( gure 9). Other ownership mentioned above is much stronger at low levels or access to nance measures studied show no signi cant quantiles of labor productivity. In fact, the relationship is correlation with women’s ownership at the NUTS2 level. weak and statistically insigni cant at the 75th and higher quantiles of labor productivity (see gure 7). is result Concluding remarks resembles the result found regarding the labor productivity gap between women-run and men-run rms. ere is a signi cant gender gap in favor of men in employment, top managerial positions, and rm ownership Access to finance across rms and NUTS2 regions in the EU-27 countries. Somewhat surprisingly, the gender gap is larger in the ere is only limited evidence that women’s ownership transition and most developed NUTS2 regions than in the across rms and NUTS2 regions in the EU-27 countries least developed NUTS2 regions. us, economic a ects access to nance. At the rm level, there is no development alone does not guarantee gender equality in 10 Enterprise Note No. 45 labor productivity across rms in the EU-27 countries is Higher women’s ownership leads to much larger at lower quantiles of labor productivity, Figure 8 fewer financially unconstrained firms across NUTS2 regions implying the presence of “sticky oors.” Another worrying feature in the EU-27 countries is that women’s employment tends to be concentrated in rms that have lower labor productivity and that paid lower wages. ese ndings Proportion of firms that are not credit suggest that achieving greater gender parity in incomes requires not only providing more jobs to women, but also better-quality jobs. constrained Notes 1 e data used for the EU-27 countries is described in detail in the rst brief in this series. Share of firm owned by women (average value at the NUTS2 level) References Source: Original calculations for this Brief based on World Bank Enterprise Survey (WBES) data. Amin, M., and A. M. Islam. 2023. “Export Intensity and Its E ect on NUTS = Nomenclature of Territorial Units for Statistics. Women's Employment.” Kyklos 76 (4): 676–704. Bertrand, M. 2018. “ e Glass Ceiling.” Working Paper No. 2018-38, Becker Friedman Institute for Research in Economics, University of Chicago. Blau, F. D., and L. M. Kahn. 2017. “ e Gender Wage Gap: Extent, Use of bank finance for investment is Trends, and Explanations.” Journal of Economic Literature 55(3): Figure 9 higher in NUTS2 regions with higher 789–865. share of women in firm’s ownership Blinder, A. S. 1973. “Wage Discrimination: Reduced Form and Structural Estimates.” Journal of Human Resources 8 (4): 436–55. % of firms using bank finance for investment Booth, A. 2009. “Gender and Competition.” Labour Economics 16, no. 6: 599-606 Booth, A. L., M. Francescopni, and J. Frank. 2003. “A Sticky Floors Model of Promotion, Pay, and Gender.” European Economic Review 47 (2): 295–322. Duraisamy, M. and P. Duraisamy. 2016. “Gender Wage Gap across the Wage Distribution in Di erent Segments of the Indian Labour Market, 1983–2012: Exploring the Glass Ceiling or Sticky Floor Phenomenon.” Applied Economics 48 (43): 4098–4111. Flabbi, L., M. M., A. Moro, and F. Schivardi. 2019. “Do Female Executives Make a Di erence? e Impact of Female Leadership on Gender Gaps and Firm Performance.” Economic Journal 129 (622): 2390–2423. Gaddis, I., and S. Klasen. 2014. “Economic Development, Structural Change, and Women’s Labor Force Participation.” Journal of Share of firm owned by women Population Economics 27: 639–681. (average value at the NUTS2 level) Goldin, C. 1995. “ e U-Shaped Female Labor Force Function in Economic Development and Economic History.” In Investment in Source: Original calculations for this Brief based on World Bank Women’s Human Capital, edited by T. P. Schultz, 61–90. Chicago: Enterprise Survey (WBES) data. University of Chicago Press. NUTS = Nomenclature of Territorial Units for Statistics. Kitagawa, E. M. 1955. “Components of a Di erence between Two Rates.” Journal of the American Statistical Association 50 (272): 1168–94. employment, rm management, and entrepreneurship. A Oaxaca, R. 1973. “Male–Female Wage Di erentials in Urban Labor gender gap also exists in the performance of rms, labor Markets.” International Economic Review 14 (3): 693–709. Wasserman, M. 2023. “Hours Constraints, Occupational Choice, productivity lower in women-run rms and rms with and Gender: Evidence from Medical Residents.” Review of higher shares of women’s ownership. is gender gap in Economic Studies 90 (3, May): 1535–68. e Enterprise Note Series presents short research reports to encourage the exchange of ideas on business environment issues. e notes present evidence on the relationship between government policies and the ability of businesses to create wealth. e notes carry the names of the authors and should be cited accordingly. e ndings, interpretations, and conclusions expressed in this note are entirely those of the authors. ey do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its a liated organizations, or those of the Executive Directors of the World Bank or the governments they represent.