87096 APRIL 2014 • Number 141 Urbanization, Gender, and Business Creation in the Informal Sector in India Ejaz Ghani, Ravi Kanbur, and Stephen D. O’Connell This Economic Premise examines the intersection of four important development themes: urbanization, agglomeration, gender, and informality. Although urbanization has continued at a rapid pace, formalization appears to have stalled. Women comprise an increasing share of the informal sector in many countries, but are increasingly underrepresented in the formal sector relative to their presence in the informal sector. Firm-level evidence suggests informal enterprise creation, particularly by women, has important connections to urbanization. Female-specific market access, especially to inputs, is a key factor for women-owned enterprise creation in the informal sector. Given the persistence of the informal sector, and given the importance of women-owned enterprise creation for jobs and gender equity, more policy measures focused on enhancing access to inputs for female-owned enterprises are important to maximize women’s contributions to India’s economic growth. The pace and content of urbanization in developing countries Unpacking the specific nature of agglomeration exter- are the subject of much discussion and analysis. The famous nalities, and going beyond a simple specification of impact on “tipping point,” at which more than half the world’s popula- unit costs, is not straightforward. One way to carefully ex- tion became urban, and which was driven largely by develop- plore this topic is to analyze the industrial structure of a loca- ing countries, has further spurred the debate on the benefits tion and measure market access effects based on the connect- and costs of urbanization (Glaeser and Joshi-Ghani 2013; edness of an enterprise to the suppliers of its inputs and the Beall, Guha-Khasnobis, and Kanbur 2010). Among the key purchasers of its outputs (Ellison and Glaeser 1997; Glaeser benefits of urbanization are the advantages of agglomeration. and Kerr 2009). The hypothesis is that urbanization, which is The simplest way of conceptualizing these benefits has been associated with greater spatial density of economic activity, through the impact of location externalities. It is hypothe- also brings greater proximity of suppliers and customers, im- sized that being located in a dense network of production and proved market access, and hence the benefits of agglomera- market access links increases productivity and lowers the unit tion on costs and productivity to each individual enterprise. costs of each individual enterprise in the network (Fujita, Such “Marshallian” conceptualization underlies much of the Krugman, and Venables 1999). The combination of agglom- recent work on agglomeration benefits in industrialized econ- eration and market access externalities leads to a positive feed- omies (Glaeser and Kerr 2009; Jofre-Monseny et al. 2011; back loop, until they are eventually countered by the costs of Dauth 2011). congestion, which can also result from agglomeration (Over- Two further topics often arise in the dialogue on urban- man and Venables 2010). ization and its costs and benefits. First, although urbaniza- 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise tion has continued at a rapid pace in developing countries, ployment (rather than employment in women-owned estab- formalization appears to have stalled, or at least does not lishments), the trend is relatively stable: women comprise seem to be proceeding as rapidly as might be expected given 27–29 percent of manufacturing employees in any of the country growth rates. Further, in India, informal enterprises years (table 3, top right). This highlights that, while there may seem to be moving into urban areas en masse, while formal not be huge employment shifts, there is an important owner- enterprises seem to be moving out of these areas. Thus, in- ship dynamic emerging. creasingly, the effects of urbanization are as likely to be Women-owned enterprises are mainly household-based found in the outcomes for informal enterprises as for formal establishments, and their number has remained relatively enterprises. stable (table 1, bottom right panel). These establishments Second, the relative overrepresentation of women in in- continue to account for only a fraction of total output, and formal sector employment has been an important departure this share of output has remained small during 1994–2005, point in the literature. However, there is a corresponding un- even given the changes in employment (table 2, top right). derrepresentation of women in the ownership of enterprises Among female proprietorships, there has been some shift in in the informal sector. The exact roles of market access and the distribution of output away from household-based estab- urbanization in this underrepresentation, and how these fac- lishments toward non-household-based establishments (table tors impact the creation of new women-owned enterprises in 2, bottom right). These features may become relevant when the informal sector, are important policy questions yet to be interpreting the impacts of market access on female enter- addressed. This note presents the findings of a recent paper prise creation. on the effects of urbanization on informality and gender Table 4 looks at the employment growth for male- and (Ghani, Kanbur, and O’Connell 2013). female-owned enterprises separately across employment- based size classes. The largest employment growth among fe- Emerging Trends male-owned enterprises was in the 1-employee size category Tables 1, 2, and 3 present trends in employment in enterpris- (both in terms of growth rate and number employed); the es in manufacturing, disaggregated by the gender of the own- second largest growth in terms of sheer numbers was the 2–4 er and workers in India. The striking finding is that there is a category. For men, largest growth in both share and numbers large increase in the share of employment in women-owned was in the 11+ category. Thus female-owned enterprises have establishments in the manufacturing sector over a relatively distinctly different growth patterns than male-owned enter- short period of time: from 9 percent of total manufacturing prises in this respect. This differentiation in gender trend by employment (organized and unorganized) in 1994 to 19 per- size of enterprise can also be a channel of differential impact cent in 2005 (table 1, top right panel). If one looks only at of Marshallian agglomeration economies on the growth of fe- women’s employment as a share of total manufacturing em- male-owned enterprises. Table 1. Manufacturing Employment Counts by Sector and Owner Gender Persons engaged in Indian manufacturing, 1994–2005 Persons engaged in Indian manufacturing, 1994–2005 (in thousands) (percentage of total) Sector 1994 2000 2005 Sector 1994 2000 2005 Total 34,424 40,702 40,336 Total 100% 100% 100% Organized 6,775 6,723 7,470 Organized 20 17 19 Unorganized 27,649 33,979 32,866 Unorganized 80 83 81 Female proprietorships 3,180 5,554 7,555 Female proprietorships 9 14 19 Male proprietorships 22,813 26,576 23,265 Male proprietorships 66 65 58 All others 1,656 1,849 2,046 All others 5 5 5 Persons engaged in Indian manufacturing, 1994–2005 Persons engaged in Indian manufacturing, 1994–2005 (female-owned proprietorships only, by location, in thousands) (percentage of female-owned proprietorships, by location) Sector 1994 2000 2005 Sector 1994 2000 2005 Female proprietorships 3,180 5,554 7,555 Female proprietorships 100% 100% 100% Household 2,882 4,934 6,800 Household 91 89 90 Nonhousehold 296 619 751 Nonhousehold 9 11 10 Other/unknown 2 1 4 Other/unknown 0 0 0 Source: Authors' calculations using Annual Survey of Industries/National Sample Survey 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 2. Manufacturing Output Value by Sector and Owner Gender Total output in Indian manufacturing, 1994–2005 Total output in Indian manufacturing, 1994–2005 (in MM 2005 US$ at PPP) (percentage of total) Sector 1994 2000 2005 Sector 1994 2000 2005 Total 459,689 650,566 870,224 Total 100% 100% 100% Organized 384,375 501,638 705,215 Organized 84 77 81 Unorganized 75,314 148,927 165,009 Unorganized 16 23 19 Female proprietorships 3,154 7,142 10,362 Female proprietorships 1 1 1 Male proprietorships 51,548 116,450 119,072 Male proprietorships 11 18 14 All others 20,613 25,336 35,575 All others 4 4 4 Total output in Indian manufacturing, 1994–2005 Total output in Indian manufacturing, 1994–2005 (female-owned proprietorships only, by location) (percentage of female-owned proprietorships, by location) Sector 1994 2000 2005 Sector 1994 2000 2005 Female proprietorships 3,154 7,142 10,362 Female proprietorships 100% 100% 100% Household 2,071 3,945 4,624 Household 66 55 45 Nonhousehold 1,080 3,194 5,730 Nonhousehold 34 45 55 Other/unknown 2 3 8 Other/unknown 0 0 0 Source: Authors' calculations using Annual Survey of Industries/National Sample Survey. Table 3. Manufacturing Employment Counts by Sector and Employee Gender Persons engaged in Indian manufacturing, 1994–2005 Persons engaged in Indian manufacturing, 1994–2005 (in thousands) (percentage of total) Sector 1994 2000 2005 Sector 1994 2000 2005 Total 34,420 40,701 40,333 Total 100% 100% 100% Organized 6,775 6,723 7,470 Organized 20 17 19 Female workers 652 654 728 Female workers 2 2 2 Male workers 3,702 3,414 3,361 Male workers 11 8 8 Supervisory/ 2,421 2,656 3,382 Supervisory/ 7 7 8 contracted/other contracted/other Unorganized 27,645 33,978 32,863 Unorganized 80 83 81 Female workers 9,191 10,649 11,594 Female workers 27 26 29 Male workers 18,454 23,329 21,269 Male workers 54 57 53 Persons engaged in Indian manufacturing, 1994–2005 Persons engaged in Indian manufacturing, 1994–2005 (female-owned proprietorships only, by location, in thousands) (percentage of female-owned proprietorships, by location) Sector 1994 2000 2005 Sector 1994 2000 2005 Female workers 9,842 11,303 12,321 Female workers 100% 100% 100% Organized 652 654 728 Organized 7 6 6 Unorganized 9,191 10,649 11,594 Unorganized 93 94 94 Source: Authors' calculations using Annual Survey of Industries/National Sample Survey. Table 5 looks at the breakdown of employment by plant doubled in rural areas). Urbanization seems to be promoting ownership across rural and urban areas. In urban areas, em- the growth of female-owned enterprises. ployment in female-owned establishments increased from 6 Urbanization, Market Access, percent in 1994 to 14 percent in 2005, and in rural areas, and Enterprise Creation from 11 to 22 percent, respectively. Thus there is a slightly stronger pattern of growth in female-owned business employ- Table 6 presents the results of a cross-sectional estimation of ment in urban areas than in rural areas—the total employment the natural log of total employment in new enterprises by in- in female-owned businesses nearly tripled in urban areas (and cumbent employment levels and size-neutral Marshallian 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 4. Within Various Employment Size Classes, What Has Been the Growth Rate Over the Time Period? Persons engaged, female-owned establishments, Establishment size share of persons engaged, by establishment size female-owned establishments Size 1994 2000 2005 Change Change % Size 1994 2000 2005 Change % Total 3,180 5,554 7,555 4,375 138 Total 100% 100% 100% 1 1,112 2,925 4,160 3,048 274 1 35 53 55 20 2–4 1,866 2,312 2,975 1,109 59 2–4 59 42 39 -19 5–7 125 162 221 95 76 5–7 4 3 3 -1 8–10 28 80 85 57 205 8–10 1 1 1 0 11+ 49 75 115 66 133 11+ 2 1 2 0 Persons engaged, male-owned establishments, Establishment size share of persons engaged, by establishment size male-owned establishments Size 1994 2000 2005 Change Change % Size 1994 2000 2005 Change % Total 22,809 26,575 23,262 453 2 Total 100% 100% 100% 1 2,940 4,183 3,519 579 20 1 13 16 15 2 2–4 14,045 16,010 12,845 -1,200 -9 2–4 62 60 55 -6 5–7 3,089 3,188 3,198 110 4 5–7 14 12 14 0 8–10 1,195 1,330 1,393 199 17 8–10 5 5 6 1 11+ 1,540 1,865 2,306 766 50 11+ 7 7 10 3 Source: Authors' calculations using Annual Survey of Industries/National Sample Survey. Table 5. Where Is the 9– >19% Pattern Coming From? Persons engaged in Indian manufacturing, 1994–2005 Persons engaged in Indian manufacturing, 1994–2005 (in thousands) (percentage of total) Sector 1994 2000 2005 Sector 1994 2000 2005 Total 34,413 40,701 40,333 Total - - - Total urban 13,301 16,529 16,414 Total urban 100% 100% 100% Organized 4,595 4,124 4,274 Organized 35 25 26 Unorganized 8,706 12,405 12,141 Unorganized 65 75 74 Female proprietorships 764 1,814 2,245 Female proprietorships 6 11 14 Male proprietorships 6,892 9,564 8,820 Male proprietorships 52 58 54 All others 1,051 1,026 1,075 All others 8 6 7 Total rural 21,111 24,172 23,919 Total rural 100 100 100 Organized 2,173 2,599 3,197 Organized 10 11 13 Unorganized 18,938 21,573 20,722 Unorganized 90 89 87 Female proprietorships 2,417 3,739 5,310 Female proprietorships 11 15 22 Male proprietorships 15,917 17,011 14,441 Male proprietorships 75 70 60 All others 605 823 971 All others 3 3 4 Persons engaged in Indian manufacturing, 1994–2005 (female- Persons engaged in Indian manufacturing, 1994–2005 owned proprietorships only, by location, in thousands) (percentage of female-owned proprietorships, by location) Sector 1994 2000 2005 Sector 1994 2000 2005 Total urban female proprietorships 764 1,814 2,245 Total urban female proprietorships 100% 100% 100% Household 617 1,479 1,810 Household 81 81 81 Nonhousehold 145 336 435 Nonhousehold 19 19 19 Other/unknown 2 0 0 Other/unknown 0 0 0 Total rural female proprietorships 2,417 3,739 5,310 Total rural female proprietorships 100 100 100 Household 2,265 3,455 4,990 Household 94 92 94 Nonhousehold 151 283 316 Nonhousehold 6 8 6 Other/unknown 0 1 3 Other/unknown 0 0 0 Source: Authors' calculations using Annual Survey of Industries/National Sample Survey. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 6. Estimation of ln (entry employment), India Unorganized Manufacturing, 2005 employment in new female- and male-owned Employment in Employment in plants, respectively. The trend, with respect Dependent variable Employment in new female- new male-owned to the importance of the interaction term, (in natural log): new plants owned plants plants holds for both, and there does not appear to (1) (2) (3) be a substantial difference in the urbaniza- Total incumbent -0.066+++ -0.015 -0.170+++ tion gradient across female versus male entry employment (0.023) (0.013) (0.034) employment. Total female incumbent 0.202+++ Looking at the interaction terms across employment (0.013) columns 1–6, there is a robust pattern of sig- Total male incumbent 0.107+++ nificance of the interaction term between ur- employment (0.028) banization and the Marshallian input metric Market access index for 0.239+++ 0.100+ 0.172++ for material input flow, but a far less robust pat- input market (0.076) (0.060) (0.073) tern for the labor inputs–based metric or for Market access index for 0.559+++ 0.176+++ 0.491+++ the output-based metric. Thus, the impact of a output market (0.061) (0.044) (0.060) match between input requirements and input Market access index for 0.188+++ 0.062 0.157+++ availability on enterprise creation is greater in labor inputs (0.060) (0.043) (0.059) more urbanized areas. However, there does not Observations 6985 6985 6985 appear to be a gender-differentiated impact of R-squared 0.377 0.453 0.342 this Marshallian metric on enterprise cre- Adjusted R-squared 0.337 0.417 0.300 ation—the effects are similar across male- and District fixed effects Y Y Y female-owned young enterprises. But this is Industry fixed effects Y Y Y with a non-gender-differentiated Marshallian Notes: Robust standard errors in parentheses. Regressions weighted by [log 2001 population * log metric. The next section deepens the analysis employment]. All specifications include a constant term that is not shown. by using gender-differentiated market access measures. metrics capturing the strength of local input, output and la- Gender-Differentiated Market Access bor markets across district-industry clusters in 2005. The Marshallian metrics are significant as determinants of em- Column 1 of table 8 presents estimates of total employment ployment in new enterprises controlling for total employ- entry with market access measures constructed on separate ment in incumbent enterprises. Table 7 presents results inter- local industrial distributions of male- and female-owned en- acting the Marshallian metrics by measures of density and terprises. When predicting total entry: (i) the distribution of urbanization. Compared to the results from the specification both male and female businesses strongly predict entry via used in table 6, we conclude that the main effect of the Mar- output market conductivity, with a positive but smaller in shallian metric is dominated by the interaction term. For ex- magnitude and statistically insignificant effect of male- and ample, in column 1, there are no statistically significant ef- female-owned input markets; and (ii) labor market conductiv- fects of the Marshallian metrics on their own: the effect is ity based on male-owned businesses is a stronger determinant entirely dependent on the interaction term with the popula- of total entry than that based on female-owned industry. tion density gradient. The interaction term is positive, sug- How well does gender-differentiated market access pre- gesting the Marshallian effects are strongest in the densest dict gender-differentiated enterprise creation? In predicting districts. female entry (column 2 of table 8): (i) local female-owned Column 2 performs the same analysis interacting with input markets strongly predict female entrepreneurship, the urbanization rate. Since these two measures are some- while male-owned input markets have no effect; (ii) the la- what correlated across districts (correlation = 0.48 in the un- bor input market based on female-owned industry predicts derlying measures used to bucket districts), it is not surprising female entrepreneurship, with again no significant effect of that the pattern of interaction importance holds for the most labor market compatibility based on local male-owned in- part for the input-based metric. However, for the urbaniza- dustry; and (iii) there are very different results for the out- tion rate measure, the main effect of the output index re- put market metric—both local male- and female-owned out- mains a statistical significant determinant and is not affected put markets have a weak but significant effect on female by urbanization. entry—suggesting a potential “upstream” relationship of fe- When comparing female- and male-owned enterpris- male producers supplying intermediate goods to both local es, is there an urbanization gradient in the effect of local female- and male-owned businesses. Across metrics predict- market access? Columns 3–4 and 5–6 of table 7 estimate ing female entry, the largest effect comes from female-based 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 7. Estimation of Entrant Employment across Districts: Density and Urbanization Hypotheses, India Unorganized Manufacturing, 2005 Dependent Employment in Employment in Employment in Employment in variable (in Employment in Employment in new female- new female- new male-owned new male-owned natural log): new plants new plants owned plants owned plants plants plants Sample Whole district Whole district Whole district Whole district Whole district Whole district Interaction Density Urbanization rate Density Urbanization rate Density Urbanization rate (1–3) (1–3) (1–3) (1–3) (1–3) (1–3) (1) (2) (3) (4) (5) (6) Total incumbent -0.064+++ -0.065+++ -0.013 -0.014 -0.167+++ -0.169+++ employment (0.023) (0.023) (0.013) (0.013) (0.034) (0.034) Total female 0.201+++ 0.199+++ incumbent (0.013) (0.013) employment Total male 0.105+++ 0.106+++ incumbent (0.028) (0.028) employment Market access index 0.086 0.037 -0.008 -0.126+ 0.019 0.003 for input market (0.094) (0.100) (0.073) (0.073) (0.091) (0.095) Market access index 0.591+++ 0.570+++ 0.160++ 0.258+++ 0.541+++ 0.501+++ for output market (0.120) (0.113) (0.077) (0.076) (0.116) (0.106) Market access index 0.062 0.124 0.012 -0.018 0.032 0.134+ for labor inputs (0.079) (0.079) (0.057) (0.047) (0.076) (0.078) Market access index 0.145+++ 0.152+++ 0.101+++ 0.169+++ 0.143+++ 0.128++ for input market (0.052) (0.052) (0.039) (0.037) (0.051) (0.051) *interaction term Market access index -0.037 -0.018 0.001 -0.067 -0.049 -0.014 for output market (0.068) (0.065) (0.046) (0.045) (0.066) (0.061) *interaction term Market access index 0.115++ 0.089+ 0.050 0.104+++ 0.113++ 0.041 for labor inputs (0.053) (0.053) (0.037) (0.038) (0.050) (0.054) *interaction term Observations 6985 6985 6985 6985 6985 6985 R-squared 0.388 0.387 0.457 0.461 0.352 0.350 Adjusted R-squared 0.348 0.348 0.422 0.426 0.310 0.308 District fixed effects Y Y Y Y Y Y Industry fixed effects Y Y Y Y Y Y Notes: Robust standard errors in parentheses. Regressions weighted by [log 2001 population * log employment]. All specifications include a constant term that is not shown. local input market strength, as opposed to labor or output seen most clearly in comparing rows 4 and 5 of columns 2 and markets. 3 in table 8. The gender-specific market access metric is sig- In predicting male entry, local male-owned industrial dis- nificant for gender-specific enterprise creation. The result is tributions matter far more than local female-owned industri- also present for the labor flows input metric (rows 8 and 9 of al distributions (column 3 of table 8). This is certainly the columns 2 and 3). However, for the output-based metric, case for input and labor markets; again, for output markets, there is a different pattern: plant entry for both genders is both local female- and male-owned industrial distributions equally predicted by their “own-gender” and “other-gender” matter, and these relationships are far larger in magnitude output market strength. and statistical significance for male entry than for female en- Conclusion and Policy Implications try. Across metrics predicting male entry, the effect is stron- ger for the output markets consisting of male-owned busi- The above analysis suggests that the effect of local input mar- nesses rather than material or labor input markets. kets (in terms of intermediate goods and labor) on enterprise The strongest gender-differentiated pattern for the Mar- creation in the informal sector is greater in more urbanized shallian material flow–based input metric of market access is areas, although this gradient does not exist for the strength of 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 8. Estimation of Entrant Employment across Districts: Gender-Specific Market Access, prise creation for achieving the goal of India Unorganized Manufacturing, 2005 gender equity, policy measures to en- Employment in Employment in hance access to inputs for female-owned Dependent variable Employment in new female- new male- enterprises are key to success. (in natural log): new plants owned plants owned plants The analysis presented here can be Sample Whole district Whole district Whole district extended and refined in a number of di- (1) (2) (3) rections. Analytically, one of the most Total incumbent employment -0.081+++ 0.030+ -0.171+++ important next steps is to better under- (0.025) (0.016) (0.036) stand the difference that has emerged Total female incumbent 0.168+++ between the input-based and the output- employment (0.018) based perspective of Marshallian agglom- Total male incumbent employment 0.052+ eration effects. The two types of metrics, (0.031) whether gender differentiated or not, Market access index for input 0.135 0.165++ -0.027 have different effects on enterprise cre- market (based on local female- (0.085) (0.072) (0.073) ation when interacting with levels of ur- owned industrial distribution) banization. Higher levels of urbanization Market access index for input 0.107 -0.071 0.227+++ intensify the effect of the input market market (based on local male- (0.077) (0.058) (0.074) owned industrial distribution) metric, but not that of the output market metric. A second direction of research is Market access index for output 0.396+++ 0.071 0.299+++ market (based on local female- (0.045) (0.044) (0.046) to better understand the nature and owned industrial distribution) properties of the gender-differentiated Market access index for output 0.423+++ 0.073+ 0.481+++ market access metrics, which are shown market (based on local male- (0.062) (0.044) (0.061) to have gender-differentiated impacts on owned industrial distribution) enterprise creation. Such further work in Market access index for labor 0.037 0.079+ -0.032 this area will have potential implications inputs (based on local female- (0.056) (0.045) (0.053) for policy affecting the livelihoods of owned industrial distribution) many. Market access index for labor 0.151++ -0.002 0.183+++ inputs (based on local male- (0.059) (0.040) (0.058) About the Authors owned industrial distribution) Ejaz Ghani is Lead Economist, Economic Observations 6985 6985 6985 Policy and Debt (PREM Network) at the R-squared 0.386 0.453 0.356 World Bank. Ravi Kanbur is T. H. Lee Pro- Adjusted R-squared 0.346 0.418 0.314 fessor of World Affairs, International Pro- District fixed effects Y Y Y fessor of Applied Economics and Manage- Industry fixed effects Y Y Y ment, and Professor of Economics, Cornell Notes: Robust standard errors in parentheses. Regressions weighted by [log 2001 population*log employment]. All University. Stephen D. O’Connell is Ph.D. specifications include a constant term that is not shown. candidate in economics at City University of New York—Graduate Center. local output markets. We also find that disaggregation of the Marshallian market strength metrics by gender adds addition- References al insight into the determinants of business creation by men Beall, Jo, Basudeb Guha-Khasnobis, and Ravi Kanbur. 2010. versus women, as gender-specific market strength metrics “Beyond the Tipping Point: A Multidisciplinary Perspective on tend to be stronger in predicting entrepreneurship by their Urbanization and Development.” In Urbanization and Develop- ment: Multidisciplinary Perspectives, ed. J. Beall, B. Guha-Khasno- own gender, at least for input markets. However, this differen- bis and R. Kanbur, 3–16. Oxford University Press. tial effect is not apparent for output markets. Dauth, Wolfgang. 2011. “The Mysteries of the Trade: Interindustry What are the policy implications? Urbanization interacts Spillovers in Cities.” Working Paper, Institute for Employment with market access to inputs so as to encourage informal en- Research, Nuremberg, Germany. terprise creation, and it does so equally for female and male Ellison, Glenn, and Edward Glaeser. 1997. “Geographic Concentra- tion in U.S. Manufacturing Industries: A Dartboard Approach.” enterprises. Female-specific market access, especially access to Journal of Political Economy 105: 889–927. inputs, is important for female-owned enterprise creation in Fujita, Masahisa, Paul Krugman, and Anthony Venables. 1999. The the informal sector. Given the persistence of the informal sec- Spatial Economy: Cities, Regions and International Trade. Cam- tor, and given the policy importance of female-owned enter- bridge, MA: MIT Press. 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Ghani, Ejaz, and Ravi Kanbur. 2013. “Urbanization and (In) Glaeser, Edward, and William Kerr. 2009. “Local Industrial Condi- Formalization.” In Rethinking Cities: A Roadmap Towards Better tions and Entrepreneurship: How Much of the Spatial Distribu- Urbanization and Development, ed. Ed Glaeser and Abha Joshi- tion Can We Explain?” Journal of Economics and Management Ghani. Washington, DC: World Bank. Strategy 18 (3): 623–63. Ghani, Ejaz, Ravi Kanbur, and Stephen D. O’Connell. 2013. “Ur- Jofre-Monseny, Jordi, Raquel Marín-López and Elisabet Viladecans- banization and Agglomeration Benefits : Gender-Differentiated Marsal, 2011. “The Mechanisms of Agglomeration: Evidence Impacts on Enterprise Creation in India’s Informal Sector.” from the Effect of Inter-Industry Relations on the Location of Policy Research Working Paper Series 6553, World Bank, New Firms.” Journal of Urban Economics 70 (2–3): 61–74. Washington, DC. Overman, Henry G., and Anthony J. Venables. 2010. “Evolving Ghani, Ejaz, William Kerr, and Stephen D. O’Connell. 2013. “Local City Systems.” In Urbanization and Development: Multidis- Industrial Structures and Female Entrepreneurship in India.” ciplinary Perspectives, ed. J. Beall, B. Guha-Khasnobis and R. Journal of Economic Geography 13(6): 929–64. Kanbur, 3–16. Oxford University Press. Glaeser, Edward, and Abha Joshi-Ghani, eds. 2014. Rethinking Cit- ies: A Roadmap Towards Better Urbanization and Development, Washington, DC: World Bank. 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. 8 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise