75524 FEBRUARY 2013 • Number 107 Promoting Women’s Economic Participation in India Ejaz Ghani, William Kerr, and Stephen D. O’Connell Despite rapid economic growth, gender disparities in women’s economic participation have remained deep and persistent in India. What explains these gender disparities? Is it poor infrastructure, limited education, or the composition of the labor force and industries? Or is it deficiencies in social and business networks and a low share of incumbent female entrepreneurs? This note analyzes the spatial determinants of female entrepreneurship in India in the manufacturing and services sectors. It finds that good infrastructure and education predict higher female entry shares. Gender networks also influence women’s economic participation, as strong agglomeration economies exist in both manufacturing and services. A higher female ownership among incumbent businesses within a district-industry predicts a greater share of subsequent female entrepreneurs. Moreover, higher female ownership of local businesses in related industries (similar labor needs, input-output markets) predicts greater relative female entry rates. A central driver of economic growth over the past century has axis), but its score for women’s economic participation and been the increased role of women. This empowerment comes opportunity is worse than 95 percent of all countries in the in many forms: increased female labor force participation, re- sample (vertical axis). duced discrimination and wage differentials that encourage What explains these huge disparities in women’s eco- greater effort, and improved advancement practices that pro- nomic participation in India? Is it poor infrastructure, limited mote talented women into leadership and managerial roles. education, or the gender composition of the labor force and As the 2012 World Development Report highlights, empow- industries? Or is it deficiencies in social and business net- ering half of the potential workforce has significant economic works and a low share of incumbent female entrepreneurs? benefits beyond promoting just gender equality (World Bank Which Industries Attract Female 2012). Entrepreneurs? In India, increases in reservations for women in panchay- ats—rural local self-government—have gone a long way in in- In a recent paper, Ghani, Kerr, and O’Connell (2012) explore creasing political participation for women. However, when it the factors that encourage female entrepreneurship in India. comes to economic participation, gender disparities remain A representative sample of the Indian economy was captured deeply entrenched. The 2012 World Economic Forum’s Gen- using microdata on the unorganized service and manufactur- der Gap Index ranked India 123rd out of 135 countries on ing sectors during 2001–2 and 2005–6, respectively. The economic participation and opportunity. survey data were used to identify the presence of new en- Figure 1 visually represents data from the Global Gender trants as well as the gender of the owner of proprietary estab- Gap Report (Hausmann, Tyson, and Zahidi 2011). India lishments. This information was analyzed to find relative scores average on the gender gap index overall (horizontal rates of female entrepreneurship and business ownership at 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1. Women’s Economic Participation: Opportunity and Overall Gender Gap Index, 2011 1.00 (0= perfect inequality, 1=perfect equality, mean: 0.63) 0.90 economic participation and opportunity subindex 0.80 0.70 0.60 0.50 0.40 India, 0.69, 0.40 0.30 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 overall gender gap index excluding economic participation, 2011 (0= perfect inequality, 1=perfect equality, mean: 0.69) Source: Hausmann, Tyson, and Zahidi 2011. the district-industry-year level, visually represented in figure tion of state gender ratios between manufacturing and ser- 2 by district. vices is about 0.5 on a count basis, and above 0.9 on an Overall, the average female business ownership share in- employment-weighted basis. creased from 26 percent in 2000 to 37 percent in 2005. On The states with the highest female service sector owner- an employment-weighted basis, the rate increased from 17 ship rates are Kerala, Tamil Nadu, and Andhra Pradesh, with percent to 25 percent. The female ownership rates across ma- average female ownership shares exceeding 12 percent. The jor cities have a distribution that is mostly similar to the distri- lowest female ownership rates are in Rajasthan, Bihar, Orissa, bution across states. and Uttar Pradesh, each with 6 percent or less. The average The districts containing India’s major cities have higher female business ownership share, with and without employ- than average rates of female entrepreneurship. Karnataka, ment weights, was between 8 percent and 9 percent for 2001 Kerala, and Tamil Nadu have relatively high female business and 2006, respectively. Among service industries, female ownership rates in unorganized manufacturing, with an aver- ownership shares exceed 30 percent in industries related to age female establishment ownership rate exceeding 45 per- sanitation and education. Industries related to research and cent. In contrast, Delhi, Bihar, Haryana, and Gujarat have low development, water transport, and land transport have the female ownership and entrepreneurship shares. lowest female ownership rates, at 1 percent or less. Within the manufacturing sector, female ownership What Drives the Gender Balance of New shares are highest and typically exceed 50 percent in indus- Enterprises? tries related to chemicals and chemical products, tobacco products, and paper and paper products. At the opposite end, The data on female business ownership were then translated female ownership shares are 2 percent or less in industries re- into metrics that combine the incumbent industrial struc- lated to computers, motor vehicles, fabricated metal prod- tures of cities, with the extent to which industries interact ucts, and machinery and equipment. through clustering or agglomeration mechanisms (Marshall In the service sector, female ownership rates in major cit- 1920). Essentially, these metrics condense complex local in- ies tend to be higher than overall state averages. The correla- dustrial structures into simple indicators, looking at the suit- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 2. Female Entrepreneurship in Unorganized Manufacturing and Service Sectors a. Manufacturing female-ownership share of young (< 3 years old) proprietary establishments (2005–6) 0–10% 20.1–30% 10.1–20% 30.1–40% 40.1–50% no data > 50% ^ major cities Delhi ^ ^ Kanpur Ahmadabad ^ Kolkata ^ ^ Surat Mumbai ^ ^Pune ^ Hyderabad Bangalore Chennai ^ ^ figure continued on next page ability of a given area for an industry in terms of local labor infrastructure, travel time to biggest cities, and the stringency force compatibility or input-output relationships. The met- of labor laws. These estimations control for industry-year rics are developed separately using female- and male-owned fixed effects. More rigorous models and instrumental variable incumbent businesses to identify how gender-specific ag- strategies largely confirm these findings. glomeration benefits affect new entrants. Initial explanatory measures focus on basic demographic Table 1 highlights the factors influencing the share of fe- traits of a district. Population control captures the size of the male-owned new establishments at the district-industry level. local consumer market, which can be especially important for Variables include broader district-level traits (population, ed- service businesses, and the overall level of surrounding eco- ucation), indicators of women’s welfare in the area (female nomic activity (for example, general availability of workers). literacy rate, total fertility rate), indicators of local physical Higher entry levels partially correlate with greater popula- 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 2. Female Entrepreneurship in Unorganized Manufacturing and Service Sectors (continued) b. Services female-ownership share of young (< 3 years old) proprietary establishments (2005–6) 0–10% 20.1–30% 10.1–20% 30.1–40% 40.1–50% no data > 50% ^ major cities ^Delhi ^Kanpur ^Ahmadabad Kolkata ^ ^Surat Mumbai ^ ^Pune ^Hyderabad Chennai Chennai Bangalore ^ ^ Source: Authors’ calculations using national sample survey data. tion, but there is no theoretical reason to suspect population basic infrastructure services like electricity are essential for influences the gender balance after controlling for other dis- all businesses, new entrants and the informal sector can be trict attributes. particularly dependent upon local infrastructure (estab- Empirical results suggest that a district/industry with lished firms are better able to provision their own electricity more incumbent female employment has a greater female en- if necessary). Inadequate infrastructure also affects women try share. Among district-level traits, a higher female-to-male more than men, because women are often responsible for a ratio, an age profile emphasizing working-age population, and larger share of, and often more time consuming, household better quality infrastructure appear important. activities. Infrastructure Interestingly, empirical findings suggest that access to The relationship between infrastructure and female entry major cities does not influence the gender balance of entrepre- share is perhaps the most relevant for policy makers. While neurship, but infrastructure access within a district does. In 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 1. Unconditional Estimations of Female Entrant Share: Agglomeration effects Manufacturing and Service Sectors Development economists frequently mention the role of Manufacturing Services business networks among women in developing countries. Estimations (2005–6) (2001–2) However, few studies systematically look at female business Log female-owned incumbent busi- ownership across regions and industries in multiple sectors +++ ++ nesses in district and explore the importance of the gender profile of the in- Log female-owned incumbent busi- cumbent industrial structures. +++ +++ nesses in district-industry The agglomeration metrics suggest that female connec- District traits: tions in labor markets and local buyer/seller (input-output) Female literacy rate 0 ++ markets contribute to a higher entry share. A 1 standard de- Sex ratio +++ 0 viation increase in either of these incumbent conditions cor- relates with a 2–3 percent increase in the share of new en- Population density --- -- trants that are female. This compares to a base female entry Education level 0 0 ratio of 21 percent. Age profile + + The first rationale is that proximity to customers and Infrastructure level +++ ++ suppliers reduces transportation costs and thereby increases Labor regulations stringency + ++ productivity (for example, Fujita, Krugman, and Venables Local industrial conditions of 1999). Within the manufacturing sector, the extent to which incumbent firms: districts contain potential customers and suppliers for new Index of labor market strength, entrepreneurs can be measured. Beyond material inputs, la- +++ n/a female-owned businesses bor is perhaps the most important input into any new firm, Index of input-output strength, and entrepreneurship is quite likely to be driven by the avail- +++ n/a female-owned businesses ability of a suitable labor force (for example, Combes and Du- Observations 4,336 4,458 ranton 2006). However, while a district’s education and basic Adjusted R-squared 0.328 0.220 demographics can determine the suitability of the local labor force, these aggregate traits can miss the specialized nature of Source: Authors’ compilation. Note: Coefficient direction and significance level shown; “+++� implies positive many occupations. coefficient significant at the 1 percent level; two and one +/- signs imply 5 percent and Most of the basic district-level links observed for manu- 10 percent significance, respectively. facturing hold true for services as well. Somewhat surprising- ly, a higher female entry ratio is not associated with a greater particular, transport infrastructure and paved roads within female ratio in the district, but female literacy rates and gen- villages play an important role. eral education levels are more predictive. This link may be Travel in India can be restrictive and unpredictable, and due to services being more skill intensive than manufacturing women face greater constraints in geographic mobility im- in India (Ghani 2010). Stronger female-owned incumbent posed by safety concerns and social norms. In addition, bet- businesses again predict a greater female entrepreneurship in ter electricity and water access may reduce the burden of service industries. women in providing essential household inputs for their These results support the conclusion that female entre- families, and allow for more time to be directed toward en- preneurship in India follows from incumbent female-owned trepreneurial activities. businesses in a district/industry that encourage subsequent Labor regulations entry. The strength of local input-output conditions are im- The positive association for stringent labor regulations is also portant, and their effects appear to be driven primarily by the relevant for female entrepreneurship. Several studies link la- presence of other local female-owned businesses. bor regulations in Indian states to economic performance Correcting Gender Imbalances through (Basu and Maertens 2009; Besley and Burgess 2004; Aghion Policy et al. 2008). These regulations may affect the gender balance of entrepreneurs by shifting activity into industries or sec- Economic growth and development depend upon successful- tors that female entrepreneurs tend to be more involved in, ly utilization of the entire workforce, both male and female. or influencing occupational decisions within the family. Despite its recent economic advances, India’s gender balance This note does not investigate this further, given that the fo- in economic participation and entrepreneurship remains cus is on the networks evident in local industrial structures— among the lowest in the world. although the partial correlation is worthy of additional re- To encourage more equitable economic participation search. and growth, better access to education and infrastructure is 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise needed. Due to the nature of household responsibilities, inad- fessor at Harvard Business School, Harvard University. Stephen equate infrastructure particularly affects women. The lack of D. O’Connell is Chancellors Fellow at City University of New specific transport infrastructure and paved roads within vil- York Graduate Center. lages is a bottleneck, given the constraints in geographic mo- References bility imposed by safety and social norms. Investment in local transport infrastructure may thus directly alleviate a major Aghion, Philippe, Robin Burgess, Stephen Redding, and Fabrizio constraint to female entrepreneurs in accessing markets. Zilibotti. 2008. “The Unequal Effects of Liberalization: Evi- dence from Dismantling the License Raj in India.� American There is also strong evidence of agglomeration econo- Economic Review 98: 1397–1412. mies in both manufacturing and services. Higher female own- Basu, K., and A. Maertens. 2009. “The Growth of Industry and ership among incumbent businesses within a district/indus- Services in South Asia and Its Impact on Employment.� In Ac- try leads to a greater share of subsequent female entrepreneurs. celerating Growth and Job Creation in South Asia, ed. E. Ghani et Moreover, higher female ownership of local businesses in re- al. Oxford University Press. lated industries (for example, similar labor needs, input-out- Besley, Timothy, and Robin Burgess. 2004. “Can Labor Regulation put markets) predict greater relative female entry rates, even Hinder Economic Performance? Evidence from India.� Quar- after taking into account the particular district/industry con- terly Journal of Economics 91–134. ditions. Promoting gender networks can directly stimulate Combes, Pierre-Philippe, and Gilles Duranton. 2006. “Labour Pool- female entrepreneurship. ing, Labour Poaching, and Spatial Clustering.� Regional Science and Urban Economics 36: 1–28. However, more research is needed to understand how gender networks influence aggregate efficiency. An impor- Duflo, Esther. 2005. “Gender Equality in Development.� MIT BREAD Policy Working Paper. tant message is that these links and spillovers across firms Fujita, Masahisa, Paul Krugman, and Anthony Venables. 1999. The can depend a lot on common traits of business owners. Like- Spatial Economy: Cities, Regions and International Trade. Cam- wise, interactions between the informal and formal sectors bridge, MA: MIT Press. may not be as strong as interactions within each sector. Fur- Ghani, Ejaz, ed. 2010. The Service Revolution in South Asia. New ther research needs to identify how these economic forces York: Oxford University Press. vary by the composition of local industry. This will be espe- Ghani, Ejaz. Forthcoming. “Local Industrial Structures and Female cially helpful for evaluating the performance of industry Entrepreneurship in India.� Journal of Economic Geography. concentrations in developing economies and guiding appro- Ghani, Ejaz, William Kerr, and Stephen O’Connell. 2012. “What priate policy actions. Explains Big Gender Disparities in India? Local Industrial This Economic Premise emphasizes the connection that Structures and Female Entrepreneurship.� Policy Research female entrepreneurs have to favorable incumbent industrial Working Paper Series 6228, World Bank, Washington, DC. structures, and the high degree to which existing female busi- Hausmann, R., L. Tyson, and S. Zahidi. 2011. Global Gender Gap Report. World Economic Forum. ness ownership enables future female entry. While achieving economic equality sometimes requires tough choices (for ex- Klapper, Leora, and Simon Parker. 2011. “Gender and Business Environment for New Firm Creation.� World Bank Research ample, progressive taxation that may discourage effort), the Observer. opposite is true in the case of gender. Unlocking female em- Marshall, Alfred. 1920. Principles of Economics. London, UK: Mac- powerment and entrepreneurship is a direct path to shared Millan and Co. prosperity and a more dynamic and sustainable growth. Munshi, Kaivan, and Mark Rosenzweig. 2005. “Economic Develop- About the Authors ment and the Decline of Rural and Urban Community-Based Networks.� The Economics of Transition 13:3: 427–43. Ejaz Ghani is a Lead Economist in the PREM Economic Policy, World Bank. 2012. Gender Equality and Development: World Debt, and Trade Department. William Kerr is an Assistant Pro- Development Report. Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise