What Decides Women Entrepreneurship in India?

Abstract

The available research literature considers women entrepreneurship as a source of economic growth and women empowerment. Entrepreneurship is credited for increases in chances of participation of women in economic growth and their overall empowerment. The present study serves two objectives. First, it highlights the recent trend and progress of women entrepreneurship in India. Second, it identifies the determinants of women entrepreneurship in the country. The study reveals the spatial concentration of women enterprises. It discusses the common problems of women entrepreneurship in the country. The results of the regression analysis reveal that female labour force participation rate, affordable credit and women’s participation in decision-making are significant factors that enhance entrepreneurship for women. Female literacy rate, despite being widely accepted as an important determinant of women entrepreneurship, is found to be statistically insignificant for women entrepreneurship. Any considerable relation could not be established between physical infrastructure and women entrepreneurship.

Keywords

Women entrepreneurship determinants and factors women empowerment trends and pattern in India

Introduction

Women in India have remained traditionally confined to household chores. Their major activities involve child rearing and looking after the family. They remain overburdened with the family responsibilities and resort to entrepreneurship only to support the family income. Social taboos and biases keep them devoid of opportunities and they undertake, if any, subsistence entrepreneurship. The disengagement in productive activities and prevalent illiteracy among women can be some of the prominent reasons that limit the expansion of women entrepreneurship in India.

A review of previous studies suggests that women entrepreneurship plays a crucial role in inclusion and empowerment of women in any society. Accordingly, multilateral organisations and policymakers have recognised the urgent need to promote women entrepreneurship and bridge the gender divide in entrepreneurship. But, entrepreneurship does not come naturally to Indian women. The patriarchal nature of Indian society disallows them to undertake entrepreneurship opportunities. This biased form of society has given rise to gender gaps in ownership of enterprises in India. For instance, out of the total 58.5 million establishments in India, only 13.76 per cent (Government of India, 2014) are run by women entrepreneurs. Furthermore, India is placed at a poor 17th position out of a total of 77 nations in the Female Entrepreneurship Index (FEI) ratings for the year 2015 (Terjesen & Lloyd, 2015). Importantly, this poor performance cannot be totally ascribed to the patriarchal nature of the Indian society. There may be some other persistent politico-economic factors responsible for this undesirable trend. Consequently, an extended investigation is required to get to the depth of the reasons behind the gender gaps in entrepreneurship in India.

Therefore, it is highly pertinent to expose what enables or hinders women entrepreneurship in India. The present study aims to identify the factors that inhibit or promote the growth of women entrepreneurship in India. Such information can guide the efficient distribution of resources, formulation of favourable regulations and the extension of required skills to the target groups.

The remaining part of the article is organised as follows: Second section provides a brief review of the available literature on women entrepreneurship in the global and Indian context. Third section discusses the trends and problems of women entrepreneurship in India. Fourth section describes the research design of the study. It also presents the description of various variables included in the regression analyses. Fifth section presents the results and implications of the regression analyses. The final section concludes the findings of the study.

Literature Review

It is widely accepted that women entrepreneurship is a source of women empowerment as well as the economic growth through the participation of women in economic activities. Policymakers cannot afford to ignore this force (Goyal & Yadav, 2014). It raises personal income for women and helps in their inclusion (Khokhar & Singh, 2016a). While female-run enterprises are steadily growing all over the world, contributing to household incomes and growth of national economies, women still face challenges in starting and successfully maintaining enterprises due to lack of funding, fewer skills and social constraints (Goyal & Yadav, 2014; World Bank, n.d.). Review of the global literature on women entrepreneurship suggests that a variety of factors lead to its growth. Also, persistent barriers to women entrepreneurship too exist. Minniti (2009) and Henry, Foss, and Ahl (2016) present a detailed and up-to-date synthesis of the available literature on women entrepreneurship and stress the need to conduct in-depth, innovative qualitative investigation.

Studies on women entrepreneurship, almost non-existent until the 1980s, witnessed a significant growth in the 1990s coinciding with the rise in the number of women entrepreneurs (Brush, 1992). In the early years, the research was restricted to unveiling similarities and differences between male and female entrepreneurs. Mirchandani (1999) identifies and explains the characteristics of entrepreneurs belonging to the two sexes. Studies reported educational, motivational, occupational and business method differences between the two sexes. Burke, Fitzroy, and Nolan (2002) underline the striking differences between the determinants of male and female entrepreneurial performance. While female entrepreneurship arises out of necessity in developing countries like India, in developed countries, opportunity and promotion are the driving forces (Jennings & Brush, 2013; World Bank, 2012). Goyal and Yadav (2014) posit that women in developing economies are more likely to be confronted by systemic barriers of access to resources.

Langowitz and Minniti (2007) investigate the entrepreneurial propensity of women. The study found that women across countries in the chosen sample tend to consider themselves and the entrepreneurial atmosphere in a less favourable light as compared to men. This perception of women is said to be the universal significant factor that influences women’s motivation to join entrepreneurship. Minniti and Naude (2010) present a detailed discussion of the patterns and determinants of female entrepreneurship across the countries. The study discusses the motivations, issues and constraints faced by female entrepreneurs. Further, previous studies conclude that the prevalence rates of women entrepreneurship are generally higher in developing countries than the developed ones (Minniti et al., 2005). Despite deepened interest in exploring women entrepreneurship in recent decades, empirical studies investigating this phenomenon are lacking. Remarkably, most of the available empirical studies are in developed economy context. More empirical research referring to developing economies is required (Gundry, Miriam, & Posig, 2002).

In the Indian context, women entrepreneurship is predominantly concentrated in the informal or unorganised sector (Khokhar & Singh, 2016b). A larger share of the studies on women entrepreneurship in India remains content with theoretical discussions on its trends and patterns. Few studies have tried to reveal the determining factors of women entrepreneurship in India. Some of them have found that the overall participation of women in economic activities is low in India. Sorsa (2015) analyses the determinants of low female economic participation in India and recommended policies for raising it. Besides, there have been some sincere attempts to expose what exactly affects the growth of women entrepreneurship in India. Ghani, Kerr, and O’Connell (2011) find links between better education and higher entry rates in entrepreneurship. But, the study remained inconclusive about the possible influence of education on the gender balance of entrepreneurs. Among the studies, particularly focusing on identifying the determinants of women entrepreneurship, Ghani, Kerr, and O’Connell (2012) analyse the spatial determinants of female entrepreneurship in India in the manufacturing and services sectors. It also looks into the various factors that affect the economic participation of women in India and found that good infrastructure and education ease the entry of women in entrepreneurship. This finding is reinforcement to the view posited by Minniti and Arenius (2003) that nations with a better literacy rate and a higher rate of participation of women in productive activities have a natural advantage for the development of women entrepreneurship. Adding to that, a significant share of female ownership in any industry/location predicts the larger share of subsequent women entrepreneurs (Ghani, Kerr, & O’Connell, 2013). Also, the decision-making ability of women is reflected through their participation in leadership roles which encourages women to enter entrepreneurship (Ghani, Kerr, & O’Connell, 2014). Further, Ghani et al. (2014) analyse the impact of access to affordable finance on women entrepreneurship. Succinctly, the works of Ghani et al. (2012, 2013, 2014) as well as Daymard (2015) focus on identification of determinants of women entrepreneurship, particularly in India. The expansion of women entrepreneurship has picked up some pace in the last few years. Daymard (2015) finds that the rise in female entrepreneurship can be due to a lack of other employment opportunities. Minniti and Naude (2010) too endorse similar view in the context of their cross-national study. The present study takes the research further and conducts regression analyses to identify the various factors that play a crucial role in deciding the extent of women entrepreneurship in India.

Women entrepreneurship in India

Trends and Developments

The recently released All India Report of Sixth Economic Census is a rich source of data and trends pertaining to women entrepreneurship in India. It recognises the important role that women entrepreneurship can play when it says that—‘Women’s equal access and control over economic and financial resources is critical for the achievement of gender equality and empowerment of women as well as equitable and sustainable economic growth and development’. The report highlights the concentration of establishments owned by women in few states. Figure 1 shows the share of the top five states/territories in women establishments. The figure reveals that over 60 per cent of the own account establishments (establishments that have no hired workers) are located in just 5 states out of a total of 28 states and 7 union territories (UT) (over 50%, even if the establishments of Telangana state formed in 2014 are excluded from Andhra Pradesh). The same 5 states also account for nearly 50 per cent of the establishments which have at least 1 hired worker. This spatial concentration of establishments indicates the possible impact of the underlying political-economic environment on the growth of women entrepreneurship.

Interestingly, out of a total of 8.05 million establishments owned by women, 83.19 per cent are single-person (owner only) establishments with no hired worker. The owner and the family members perform all kind of duties in such establishments. The rest of the 16.81 per cent establishments hire at least 1 worker. A total of 5,243,044 establishments which constitute 65.12 per cent of the total establishments are located in rural areas. The share of rural areas in a total establishment is almost in proportion to the percentage of national population residing there.

Figure 1.

Source: Government of India.

The distribution of total establishments on the basis of nature of activity reveals that about one-third (34.7%) of the establishments owned by women are agricultural establishments. Manufacturing (45.36%) and trading (28.57%) constitute the major share in the non-agricultural establishments, which are themselves 65.3 per cent of the total establishments owned by women.

The Problems

Decision-Making Opportunity for Women

Women’s entry into entrepreneurship faces barriers of unequal access to resources or opportunities, and women are subject to different social stereotypes and expectations (Sullivan & Meek, 2012). The situation of women in India is no different. Though the Indian society has modernised to some extent, especially in urban pockets, the culture continues to be largely patriarchal in nature. Such cultural dynamics and mindset are major constraints to the development of entrepreneurship among women in India. Women are confined to household work and rarely get the opportunity to assert themselves. They lack financial autonomy and are dependent on elder male family members for investment or purchase decisions. To emancipate them from this bondage, the reservation for women in gram panchayat (village local self-government) and municipal bodies plays an important role. The pending women reservation bill (in Parliament of India) proposes to reserve 33 per cent seats for women in the lower and upper house of the Parliament and the State Legislative Assemblies. The idea of women empowerment through quotas is the need of the hour. A women representative is generally more accessible to women and is expected to take better initiatives for the cause of women. She is expected to introduce more legislation favourable to women. Consequently, the present study takes the number of female members in state legislative assemblies as an independent variable.

Figure 2.

Source: Government of India.

Women’s Access to Education and Training

It is a known fact that education plays critical role in empowerment as well as in the development of entrepreneurship. But, there are gender gaps in skills and knowledge to start and successfully operate an enterprise. Van der Sluis, Van Praag, and Vijverberg (2005) analyse the relationship between schooling and entrepreneurship selection in developing economies and has found a positive relation. Afridi (2010) and Afridi, Mukhopadhyay, and Sahoo (2012) underlines that education helps women empowerment through awareness about rights and capability enhancement through training and exposure. The government in India established the National Entrepreneurship Development Institutes (NEDI) as a major initiative to fill this gap. These institutes run training programmes to train and educate the youth to inculcate the entrepreneurial culture among the first generation entrepreneurs. To test for the impact of education on women entrepreneurship, the state-wise female literacy rate has been picked up as another independent variable for the study.

Opportunity for Women to Work

Participation of women in labour force indicates their willingness to get engaged in economic activity (Afridi et al., 2012). They learn relevant skills through engagement in various activities and develop self-confidence. Engagement in any kind of work raises the entrepreneurial potential of women. It boosts their self-esteem and women become more assertive in terms of entrepreneurial ambitions. Allegedly, out of suspicion of creating a dent in male dominion, women at times are dissuaded to shun their entrepreneurial plans. Kobeissi (2010) finds that labour force participation along with education abandons the traditional gender stereotypes. Further, working women set an example for other women who remain idle at home. Ghani et al, (2013) asserts that a high share of women entrepreneurs in any sector and location positively affects the subsequent share of women in entrepreneurship. Therefore, female Labour Force Participation Rate (LFPR) has also been picked as an independent variable.

Figure 3.

Source: Government of India.

Women’s Access to Financial Resources

The access to easy credit and good infrastructure may be other factors vital for the spread of women entrepreneurship in any country. Many women shy away from entrepreneurship because they believe that getting credit is very difficult and costly. Figure 3 reveals the percentage share of various sources of finance for women entrepreneurs in India. Almost four-fifths of the total women-owned establishments are self-financed. It highlights the over dependence of women entrepreneurs on personal savings as a source of finance to start and run their establishments. The total share of organised sector finance for women-owned establishments is merely a little over five per cent. The absence of significant public lending system may be a major impediment to the growth of large women-owned establishments. Accordingly, the credit to small enterprises by scheduled commercial banks (SCBs) and the credit by banks to self-help groups (SHGs) have been taken as an independent variable in the regression analyses to test the impact of access to credit on women entrepreneurship in India.

Infrastructural Support

The availability of proper physical infrastructure is as important for the expansion of women entrepreneurship as any other factor. Poor road quality confines women to their living place and disallows any training or exposure available in big cities. It hinders the access of women to the latest knowledge and skills to start and successfully run any business. It also makes tough for their goods and services to reach the big markets and trade becomes virtually impossible. Under insufficient infrastructure, women entrepreneurial ventures are highly likely to fail. This will further suppress entrepreneurial motivation among women. Therefore, the surfaced road length in the state has been taken as a variable to represent the availability of physical infrastructure. All the factors discussed here have been considered for testing their impact on the growth of women entrepreneurship in India.

Methodology

The study undertakes a theoretical and empirical investigation to unveil the problems and causes of women entrepreneurship in the country. The major objectives of the study are listed below:

Objectives of the Study

To discuss the recent trend and progress of women entrepreneurship in India.

To identify the determinants of women entrepreneurship in India.

The study employed secondary data from the Sixth Economic Census to illustrate the pattern and trends of women entrepreneurship across the country. This part of the research is discussed in third section. The extended discussion has in fact assisted the choice of explanatory variables for the regression analysis. The study conducts regression analysis to identify the determinants of women entrepreneurship in India.

Sample

The present study conducts regression analysis on 21 states and 1 UT of India out of a total of 29 states and 7 UTs. The sample is selected through carefully examining the extent of female entrepreneurial activity in the states/UTs. All major states/UTs are selected leaving only those states/UTs which observe the sparse entrepreneurial activity. Table 1 presents the list of the selected states.

Database

The study employs secondary data from various published sources. For the regression analysis, pooled data on the number of entrepreneurship is taken from 62nd and 67th National Sample Survey Organization (NSSO) surveys pertaining to the year 2005–2006 and 2010–2011 and All India Report of Sixth Economic Census pertaining to the year 2013–2014. The number of enterprises owned by women in each state/UT is taken as the dependent variable in the econometric model. Thus the dependent variable is the number of unincorporated women manufacturing enterprises in the selected state. The rationale behind the selection of this variable comes from the fact that more than 90 per cent of the enterprises in India are from the unincorporated sector. Separate data for women entrepreneurs is available for such enterprises through NSSO surveys and most recently the Sixth Economic Census report too provides data on women entrepreneurship. The data sources and other specifications of the various independent variables tested in the regression are further explained in Table 2.

Table 1.

The List of Indian States/UTs Selected for the Study

Andhra Pradesh¹ 1 Andhra Pradesh in the context of the present study means the state before its bifurcation into Telangana State and Andhra Pradesh under Andhra Pradesh Re-organisation Act, 2014.	Assam	Bihar	Chhattisgarh	Delhi
Goa	Gujarat	Haryana	Himachal Pradesh	Jammu & Kashmir
Jharkhand	Karnataka	Kerala	Madhya Pradesh	Maharashtra
Odisha	Punjab	Rajasthan	Tamil Nadu	Uttarakhand
Uttar Pradesh	West Bengal

Source: The author.

Table 2.

The Description and Database of the Variables in Regression Analysis

Variables	Description	Database
Dependent Variable
Log(Enterprises)	No. of manufacturing enterprises owned by women	62nd Round NSSO survey on Unorganised Manufacturing Enterprises of India. 67th Round NSSO survey on Unincorporated Non-agricultural (excluding construction) Enterprises in India. All India report of Sixth Economic Census.
Independent Variables
Log (Literacy)	Female literacy rate	61st and 66th Round NSSO survey on Status of Education and Vocational Training in India. 71st Round NSSO survey on Social Consumption in India (Education).
Log (SBusiness)	Credit by scheduled commercial banks to small enterprises	Database of the Indian economy, Reserve Bank of India.
Log (Decision)	Number of women Member of the Legislative Assembly (MLA) in State Legislative Assemblies	Website of Election Commission of India.² 2 See www.eci.nic.in
Log (Road)	Surfaced road length in state	Road transport, TRW, Ministry of Road Transport and Highways 2001–2008. Basic Road Statistics of India 2008–2011 and 2013.
Log (BSHG)	Bank credit to self-help groups	Microfinance India: A state of the sector report 2006, 2010 and 2013.
Log (LFPR)	Labour force participation rate of women	62nd, 66th and 68th Round NSSO survey on Employment and Unemployment situation in India.

Source: The author.

The Econometric Model

The discussion in third section suggests that there are multiple factors that may determine the progress of women entrepreneurship in India. The discussion assisted with a detailed literature review guided the selection of variables to be tested for their impact on women entrepreneurship. In order to identify the determinants of women entrepreneurship in India, the present study performs fixed effect regression analysis through the econometric model explained as under.

I n Y_{i t} = α + β I n X_{i t} + τ_{t} + μ_{i t}

α = intercept term u_it = stochastic error term i = 1 to n (n = 22 states/UT of India)

t = 1, 2 3 (2005–2006, 2010–2011 and 2013–2014)

ln Y_it represents the natural log of the state-wise number of female manufacturing enterprises. The βs here are the slope coefficients for the different explanatory variables. ln X_it represents the natural log of a vector of state-level variables. A total of six regressors are tested for their impact on female entrepreneurship. τ_t represents the period of fixed effects. The model takes into account the period fixed effects in the sense that all the variations in the variables are due to time-variant factors and the changes are cross-section invariant. The model provides the elasticity of a number of women enterprises with respect to the selected independent variables. It gives the percentage change in a number of female establishments for a percentage change in one of the independent variables, keeping the rest of the independent unchanged. Reasonable adjustments have been made in the data to overcome the possible impact of the difference in gaps between the reference years. Pooled data regression analysis is conducted on six alternative models to achieve the investigative results.

Discussion

The present study is a kind of exploratory research as very few empirical studies have previously been conducted at sub-national level. If insight is drawn from the available theoretical literature on women entrepreneurship then several competing econometric models can be formulated to identify the determinants of women entrepreneurship in India. Choosing among these competing models or hypotheses is then a hard job. The appropriate way to decide the best among the competing econometric models is to test as many of them as possible and choose the one that best fits the data. Miller (1978, p. 176) asserts—‘no encounter with data is a step towards genuine confirmation unless the hypothesis does a better job of coping with the data than some natural rival. What strengthens a hypothesis, here, is a victory that is, at the same time, a defeat for a plausible rival’. Therefore, the study tests multiple hypotheses and allows the results of data analysis to figure out the determinants of women entrepreneurship in India.

Results and Implications

The study tests six alternative hypotheses. The alternative hypotheses are represented by six different econometric models. A total of six independent variables make the alternative econometric models. The results of the regression analysis as presented in Table 3 indicate that all models are reasonably good fit except the model presented in Column 2. More than 70 per cent variations in the dependent variable (LogEnterprises) are explained by variations in the independent variables in models presented in Columns 3 to 7. The model in Column 2 that consists only two regressors (Log(Literacy) and Log(LFPR)) does not adequately fits the data as only about 11 per cent variations in regress and are explained by variations in the regressors. The F-statistic value indicates that regressors in all econometric models except the one presented in Column 2 are jointly significant at 1 per cent level of significance. Out of the two regressors in Column 2 model, the literacy variable is not significant at any popular level of significance, while the other variable LFPR is significant only at 10 per cent level of significance. These indicators make econometric model presented in Column 2 worth rejection. It satisfies none of the basic statistical requirements of a good model and hence can be considered an inadequate model for identifying the determinants of women entrepreneurship in India.

Literacy and labour force participation has been taken as the base regressors in the regression analysis. They were expected to have the greatest impact on women entrepreneurship in India. However, the results are away from the intuition. Female literacy rate, though, positively related with women entrepreneurship, is statistically insignificant for all regressions except the one in Column 3 where it is significant at 5 per cent level of significance. Theoretically, a high literacy rate is conducive to skill development and innovation. The enterprises under investigation in the study are mostly small and informal sector enterprises. The owners of such units are generally under-educated, which is inimical to their progress. Despite that, the empirical evidence suggests that the variable fails to have any significant impact on the development of women entrepreneurship in India.

Further, the regression analysis results reveal that out of the total six regressors, LFPR, SBusiness and Decision are significant at popular levels of significance in every econometric model where they are included. The regressor Decision is significant at five per cent level of significance. Comparatively, the variable called SCB loans to small businesses (SBusiness) has the strongest impact on women entrepreneurship. It accentuates the fact that access to formal sources of credit is the most important determinant of women entrepreneurship. The coefficient for SBusiness in Column 7 tells that a 100 per cent increase in the SCB lending to small businesses will raise the entrepreneurial activity among women by over 102 per cent. The beta coefficient is statistically significant at 1 per cent level of significance. This highlights the important role that this source of lending can play in the emergence and development of women entrepreneurship in India. The enterprises covered in the study are small unincorporated units. Access to affordable credit has remained a challenge for such women owners. There is a need to ensure their reach to formal sources of finance. The credit by commercial banks is more dependable as it is available at affordable rates and for longer time periods. There is a need to ease the procedural barriers to such loans. In absence of formal sources of credit, there is a high likelihood for the small entrepreneur to fall in the trap of private moneylenders. These moneylenders fleece the poor borrowers and the entrepreneur is stuck in never-ending debt trap. So, access to formal credit is not just prudent but also the call of social justice. Figure 3 revealed that the share of lending by banks and government in the overall sources of finance for women establishments is less than five per cent which is considerably low. There is a need to raise the share in order to expand women entrepreneurship in India.

The study finds that the participation of women in the labour force plays a significant and positive role in the development of women entrepreneurship in India. The coefficient for LFPR is significant at 1 per cent level of significance in most econometric models tested for the present study. The coefficient for LFPR in Column 7 indicates that a 100 per cent rise in labour force participation will bring at least a 60 per cent rise in women entrepreneurial activity in the country. Importantly, the participation in labour force reflects the desire of women to get engaged in productive and income earning activities. Such inspired and motivated women are also likely to possess entrepreneurial acumen. This energy and passion are expected to translate them into good entrepreneurs. The rise in LFPR indicates an upswing in women’s willingness to work. A part of this rise in the women labour force will be diverted towards entrepreneurship and will result in an upsurge in women entrepreneurship.

The coefficient for the regressor BSHG is positive and significant at 10 per cent level of significance. The result augments the above finding that easy access to credit is advantageous for the expansion of women entrepreneurship. It also unveils the fact that efficient functioning of small institutions targeted at training and facilitating women entrepreneurship is crucial for it. These SHGs facilitate entrepreneurial activity among women by engaging women in productive work. They render skill-specific training and conduct awareness programme for women. The SHGs also facilitate the women entrepreneurs in finding a market for their produce. The empirical evidence tells that a 100 per cent rise in funding for SHGs will boost women entrepreneurship by about 19 per cent even if all other factors are kept constant.

The regressor Road was found to be insignificant at all popular levels of significance. The very small value of beta coefficient for this variable indicates that the elasticity of this variable with respect to a number of women enterprises is very low. Also, the beta coefficient values for this regressor have different signs in the two econometric models where this variable has been included. The inclusion of this variable slightly brings down the F-statistic value as compared to the models where this variable is not included. Theory tells that well-developed basic infrastructure is conducive for growth of entrepreneurship. But, the empirical evidence does not support this view. One reason could be that the establishments under study are miniscule units; hence, the road transport facilities may not be the chief requirement for the growth of such units. The redundant variable test results (results not presented here) confirm that the variable can be dropped from the model. The final results after dropping this variable are presented in Column 7 of the table.

Another important outcome of the regression is that women’s participation in decision-making is found to share positive and significant relation with women entrepreneurship. The coefficient for variable Decision tells that a 100 per cent surge in a number of women Member of the Legislative Assembly (MLAs) in state legislative assemblies will bring nearly 42 per cent rise in number of women enterprises. It confirms that women need to be provided a greater opportunity to play decision-making and management roles. A rise in a number of women MLAs in state assemblies is expected to inspire more women-friendly regulations. Such occurrence will encourage women to launch their enterprises. There is a need for social as well as administrative reforms to ensure greater space for women to participate in economic activities. The policymakers and society have to accept that women are not just 50 per cent population. They represent 50 per cent of the potential human capital and innovative brain power in India. The success of many women entrepreneurs in recent years confirms that women are an indispensable source of growth and development.

The above discussion has brought out the major determinants of women entrepreneurship in India. Although the result analysis has exposed the impact of major politico-economic factors, it has thrown surprises too. Female literacy rate, which was assumed to strongly influence women entrepreneurial activity in any society, was proved to be inconsequential and insignificant variable. Furthermore, the variable representing metalled road length in the state was tested insignificant. This highlights the fact that the study has yielded some unorthodox results. The study tested for a total of six independent variables. Although every care has been taken to ensure that the model and the empirical investigation do not violate any basic assumptions of the classical regression models, there are possibilities of failure of one or more assumptions. Collinearity among the regressors is a real possibility in social science research like in the present study. The next subsection discusses the problem of collinearity in the context of the present study and undertakes deliberation on the reliability of the presented regression results.

Diagnostic Tests

In the present study, the regress and measures the number of women enterprises. There are six regressors, namely female literacy rate (Literacy), credit by commercial bank to small enterprises (SBusiness), number of women MLAs in state legislative assemblies (Decision), surfaced road length in state (Road), bank credit to self-help groups (BSHG) and labour force participation rate of women (LFPR). There is a possibility in social science research that the predictors are correlated in some way. Putting it more formally, there is a likelihood of the existence of the problem of collinearity. The problem in case of multiple regression is referred to as multicollinearity. One of the common methods used to detect multicollinearity is to look at the correlation among the regressors. The diagnostics for multicollinearity include calculation of pairwise correlation coefficients as presented by a correlation matrix. For multivariable relationships, multiple correlation coefficient of each regressor with respect to all other regressors is calculated to measure intercorrelation among more than two regressors. A large value of correlation indicates the possible existence of multicollinearity. Another popular tool to confirm the existence of multicollinearity is the Variance Inflation Factor (VIF) values for regressors. Although the beta coefficients are still BLUE in the presence of multicollinearity, their variances are inflated. In the presence of high multicollinearity, the beta coefficients cannot be estimated precisely or accurately as they have large variances and covariances. The VIF quantifies the extent to which the variances for the coefficients are inflated. It is basically the speed with which the covariances and variances increase. A bigger multiple correlations among regressors lead to higher values of VIF.³

VIF can assume values from 1 to ∞. The VIF is defined as:

VIF = \frac{1}{1 - R_{t}^{2}}

The multiple correlation coefficient values of 0 and 1 yield VIF values of 1 and ∞ respectively. The beta coefficients of regressors are indeterminate and their standard errors are infinite in case of perfect multicollinearity (VIF = ∞), while they attain their original or uninflated values in case of no multicollinearity (VIF = 1).

The VIF is the reciprocal of tolerance, which is expressed as (1 – R_i²). As a consequence of high or near multicollinearity, the confidence intervals tend to be much wider, resulting in acceptance of the ‘zero null hypothesis’ more readily.

Table 3.

Results of the Regression Analyses

IndependentVariables	Dependent Variable: Log (Number of Women Manufacturing Enterprise)
1Log (Literacy)	21.786(1.400)	31.70**(2.337)	40.675(0.877)	50.674(0.860)	60.584(0.771)	70.555(0.743)
Log (LFPR)	0.921***(1.703)	0.925*(2.999)	0.953*(3.337)	0.948*(2.733)	0.684***(1.895)	0.614***(1.868)
Log (SBusiness)		1.429*(9.166)	1.134*(6.305)	1.130*(4.217)	1.117*(4.320)	1.027*(5.613)
Log (Decision)		—	0.535*(2.744)	0.533**(2.417)	0.459**(2.121)	0.418**(2.109)
Log (Road)		—		0.006(0.022)	–0.144(–0.496)
Log (BSHG)		—			0.201***(1.945)	0.187***(1.900)
Intercept term	5.983**(2.333)	0.247(.156)	–0.995(–0.64)	–0.999(–0.635)	–0.491(–0.319)	–0.616(–0.410)
R ²	0.106	0.716	0.763	0.763	0.786	0.784
F–statistic	1.59	24.684*	24.56*	19.930*	18.908*	22.476*
No. of observations	66	66	66	66	66	66
No. of states/territories	22	22	22	22	22	22

Source: The author.

Notes: 1. The results are based on period fixed effect panel data regression method.

2. *, ** and *** denote the significance at 1%, 5% and 10% levels, respectively.

3. Figures in the parentheses are the robust t-statistics.

Table 4.

The Correlation Matrix of the Independent Variables

	Literacy	Sbusiness	Decision	Road	BSHG	LFPR
Literacy	1
SBusiness	0.0802	1
Decision	–0.3745	0.5394	1
Road	–0.1164	0.7914	0.6582	1
BSHG	–0.1200	0.4851	0.4844	0.6753	1
LFPR	–0.0502	–0.0625	–0.0450	0.2460	0.3864	1

Source: The author.

The study conducts regression diagnostics in order to judge the goodness of the fit of the econometric models. It calculates pairwise correlation coefficients presented in the form of the correlation matrix in Table 4. The pairwise correlation coefficients, as presented in the correlation matrix, dismiss the existence of perfect correlation among the regressors. Most of the correlations among the regressors are weak. They lie between no and low correlation range (correlation coefficient ranging from –0.25 to 0.25). Some correlation values lie in the low to medium range (correlation coefficient ranging from 0.25 to 0.5 or –0.5 to –0.25). There are few moderate correlations (correlation coefficient ranging between 0.6 and 0.8) and all these involve the variable Road. A high pairwise correlation coefficient could be the first indicator of the existence of multicollinearity problem.

Although the level of correlation between the regressor Road and other regressors is manageable as per popular norms,⁴

As a rule of thumb, pairwise correlation up to 0.8 is considered acceptable, and multicollinearity is not a serious problem.

the study conducts a further diagnosis to validate the regression results and to intensify the reason for exclusion of the regressor Road, if required compulsorily.

The study calculates Tolerance and VIF to test for the presence of multicollinearity. There are recommendations in the literature about acceptable levels of VIF. Major literature recommends a maximum acceptable VIF value as 10 (Kennedy, 2003; Marquaridt, 1970). However, recommendations can also be found for a maximum acceptable VIF value of 5 (Rogerson, 2001). The calculated values of VIF and Tolerance are presented in Table 5. The table shows that the highest VIF value turns out to be 5.2387 for regressor Road. The VIF values for all other regressors are well within the strictly an acceptable range of 1 and 5. The mean VIF value in this study is 2.7250, which is comfortably within acceptable range. Therefore, there is no problematic multicollinearity issue in the present study. The greater than 5 VIF value for Road is a matter of concern as per recommendations of Rogerson (2001). The variable can be dropped adducing the high VIF value. The VIF value of 5.2387 for the regressor Road means that the variance (standard error) for the coefficient of that predictor variable is 5.2387 () times as large as it would be when the predictor variable is uncorrelated with the rest of the predictor variables.

Notably, a high VIF value is not sufficient reason for discarding any regressor. High VIF is neither necessary nor sufficient to get high variances and high standard errors (Gujarati & Sangeetha, 2011). But, the redundant variable test too confirmed that the variable Road can be dropped. Doubt is also created by the correlation matrix analysis about the desirability of this regressor in the econometric model for identification of determinants of women entrepreneurship in India. Hence, there are sufficient reasons for discarding the regressor Road. Thus, the diagnostic tests justify the exclusion of regressor Road. The tests validate the regression results as presented in Table 3 and confirm that the econometric model presented in Column 7 in Table 3 is most appropriate out of all the models analysed in the present study. The study, therefore, suggests that credit by commercial banks to small enterprises (SBusiness), number of women MLAs in the state legislative assemblies (Decision), bank credit to self-help groups (BSHG) and labour force participation rate of women (LFPR) are the factors that decide the level of women entrepreneurship in India. It also unveils that there are no significant reasons to believe that female literacy plays an important role in the development of women entrepreneurship in India.

Table 5.

The VIF and Tolerance for the Independent Variables

	R ²	Tolerance	VIF
Literacy	0.2592	0.7408	1.3499
Decision	0.5843	0.4157	2.4056
BSHG	0.5237	0.4763	2.0997
Road	0.8091	0.1909	5.2386
LFPR	0.3797	0.6203	1.6120
SBusiness	0.7256	0.2744	3.6441

Source: The author.

Limitations and Directions for Future Research

The major contribution of the present study is to unveil the trend and pattern as well as determinants of women entrepreneurship in India. The present section mainly addresses the limitations of the study and the implications for future research. The execution of empirical research on a social science issue at sub-national level imposes methodological limits on the investigators. The present study would have benefitted from larger longitudinal data. The limited entrepreneurial activity in some of the Indian states/UTs restricted the size of the sample. The limitations of the present study and directions for future research are:

Future studies on women entrepreneurship can follow alternative research designs. The present study and Ghani et al. (2011, 2012, 2013, 2014) have made a significant contribution towards empirical and theoretical research on women entrepreneurship at sub-national level. The future research can extend the work of these studies by taking larger samples as well as longer period of time.

A primary data based study may be conducted to verify the findings of other studies, like the present one, based on secondary data. Secondary data has its own limitations. Although it is easy to obtain secondary data, it brings its own complications. The quality and reliability of secondary data are hard to judge. This data is not collected specifically for the sake of the investigator’s research. Such reasons make a primary study very much relevant.

Future projects may also conduct a comparison study on different groups such as urban women, rural women, first generation entrepreneurial families and traditionally entrepreneurial families. Such comparison study could unveil how women’s perception of entrepreneurship changes according to their entrepreneurial exposure.

The list of variables selected for the study is not exhaustive. A host of other factors may promote or inhibit the growth of women entrepreneurship in India. There is ample scope for an investigation into psychosocial determinants of women entrepreneurship in India.

Conclusion

One of the objectives of this study has been to discuss and identify the recent trends and patterns of women entrepreneurship in India. The detailed discussion assisted with illustrations has been able to reveal that women entrepreneurship in India is in its nascent stage. There is a large concentration of women entrepreneurship in the informal sector. Also, there is the spatial concentration of women establishments. Over 50 per cent establishments are located in just 5 states. Although, women entrepreneurship possesses huge potential for growth in India, the excessively large share of informal sector units raise doubts over the sustainability of such establishments.

The results of the regression analysis revealed that easy access to credit and women labour force participation rate have a positive and strong impact on women entrepreneurship in India. A rise in participation of women in the labour force and improvement in access to credit are expected to ease the barriers for women to assume entrepreneurial roles. Further, the opportunity for women to participate in decision-making, as is represented by the share of women in state legislative assemblies, is another important factor that positively affects the development of women entrepreneurship. Although education or literacy is generally assumed to play a positive role in the development of women entrepreneurship, the present study finds that the impact of literacy is statistically insignificant in five out of six econometric models. Infrastructure, which may not have a direct impact on growth, is generally assumed to foster it. Interestingly, the study did not find any significant relation between the existence of better roads (a proxy for infrastructure) and women entrepreneurship. In fact, the variable representing physical infrastructure is diagnosed as an offending variable and it is dropped accordingly from the econometric model. The bank credit to SHGs has a positive and statistically significant coefficient in the final econometric model. It reveals that the bank credit to SHGs plays an important role in the expansion of entrepreneurial opportunities for women. This finding highlights the fact that SHGs are a catalyst to women empowerment through the development of women entrepreneurship in the country. They provide skill training at a local level and act as an institutional source of small credit.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

References

Afridi

(2010). Women’s empowerment and the goal of parity between the sexes in schooling in India. Population Studies , 64(2), 131–145.

Afridi

Mukhopadhyay

Sahoo

(2012). Female labour force participation and child education in India: The effect of the national rural employment guarantee scheme (IZA Discussion Paper No. 6593). Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2085180

Brush

C. G.

(1992). Research on women business owners: Past trends, a new perspective and future directions. Entrepreneurship Theory and Practice , 16, 5–30.

Burke

A. E.

Fitzroy

F. R.

Nolan

M. A.

(2002). Self-employment wealth and job creation: The roles of gender, non-pecuniary motivation and entrepreneurial ability. Small Business Economics , 19(3), 255–270.

Daymard

(2015). Determinants of female entrepreneurship in India (OECD Economics Department Working Paper No. 1191). Paris: OECD Publishing.

Ghani

S. E.

Kerr

W. R.

O’Connell

(2011a, November). Local Industrial Structures and Female Entrepreneurship (NBER Working Paper No. 17596). Cambridge, MA: NBER.

Ghani

Kerr

O’Connell

(2011b). Promoting entrepreneurship, growth, and job creation. Reshaping Tomorrow , 168–201.

Ghani

S. E.

Kerr

W. R.

O’Connell

(2012, October 1). What explains big gender disparities in India? (Policy Research Working Paper No. 6228). Washington, DC: World Bank.

Ghani

Kerr

O’Connell

(2013). Promoting women’s economic participation in India. Economic Premise , 107, 1–6.

10.

Ghani

Kerr

W. R.

O’Connell

S. D.

(2014). Political reservations and women’s entrepreneurship in India. Journal of Development Economics , 108, 138–153.

11.

Government of India. (2014). All India Report of Sixth Economic Census 2013–2014 . New Delhi: Ministry of Statistics and Programme Implementation, Government of India. Retrieved from http://www.indiaenvironmentportal.org.in/files/file/All%20India%20Report%20of%20Sixth%20Economic%20Census.pdf

12.

Goyal

Yadav

(2014). To be or not to be a woman entrepreneur in a developing country. Psychosociological Issues in Human Resource Management , 2(2), 68–78.

13.

Gujarati

D. N.

Sangeetha (2011). Basic econometrics . New Delhi: Tata McGraw-Hill.

14.

Gundry

L. K.

Miriam

B. Y.

Posig

(2002). Contemporary perspectives on women’s entrepreneurship: A review and strategic recommendations. Journal of Enterprising Culture , 1(10), 67–86.

15.

Henry

Foss

Ahl

(2016). Gender and entrepreneurship research: A review of methodological approaches. International Small Business Journal , 34(3), 217–241.

16.

Jennings

J. E.

Brush

C. G.

(2013). Research on women entrepreneurs: Challenges to (and from) the broader entrepreneurship literature? Academy of Management Annals , 7(1), 663–715.

17.

Kennedy

(2003). A guide to econometrics . Cambridge, MA: MIT press.

18.

Khokhar

A. S.

Singh

(2016a). Status of women entrepreneurship in India. Asian Journal of Multidisciplinary Studies , 4(5), 200–203.

19.

Khokhar

A. S.

Singh

D. B.

(2016b). Role of women entrepreneurship in inclusive growth. South Asia , 11(9.6), 18–3.

20.

Kobeissi

(2010). Gender factors and female entrepreneurship: International evidence and policy implications. Journal of International Entrepreneurship , 8(1), 1–35.

21.

Langowitz

Minniti

(2007). The entrepreneurial propensity of women. Entrepreneurship Theory and Practice , 31(3), 341–364.

22.

Marquaridt

D. W.

(1970). Generalised inverses, ridge regression, biased linear estimation, and nonlinear estimation. Technometrics , 12(3), 591–612.

23.

Miller

R. W.

(1987). Fact and method: Explanation, confirmation and reality in the natural and the social sciences . Princeton, NJ: Princeton University Press.

24.

Minniti

(2009). Gender issues in entrepreneurship . Delft: Now Publishers Inc.

25.

Minniti

Arenius

(2003, April). Women in entrepreneurship: The entrepreneurial advantage of nations—First annual global entrepreneurship symposium . Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.471.7819&rep=rep1&type=pdf

26.

Minniti

Naudé

(2010). What do we know about the patterns and determinants of female entrepreneurship across countries? The European Journal of Development Research , 22(3), 277–293.

27.

Minniti

Arenius

Langowitz

(2005). The 2005 Global Entrepreneurship Monitor Special Topic Report: Women in entrepreneurship . Babson Park, MA: Center for Women’s Leadership at Babson College.

28.

Mirchandani

(1999). Feminist insight on gendered work: New directions in research on women and entrepreneurship. Gender, Work & Organization , 6(4), 224–235.

29.

Rogerson

(2001). Statistical methods for geography . Thousand Oaks, CA: SAGE Publications.

30.

Sorsa

(2015). Raising the economic participation of women in India (OECD Economics Department Working Paper No. 118). Paris: OECD Publishing.

31.

Sullivan

D. M.

Meek

W. R.

(2012). Gender and entrepreneurship: A review and process model. Journal of Managerial Psychology , 27(5), 428–458.

32.

Terjesen

Lloyd

(2015). Female Entrepreneurship Index report 2015 . Washington, DC: Global Entrepreneurship Development Institute. Retrieved from http://thegedi.org/download/11536/

33.

Van der Sluis

Van Praag

Vijverberg

(2005). Entrepreneurship selection and performance: A meta-analysis of the impact of education in developing economies. The World Bank Economic Review , 19(2), 225–261.

34.

World Bank. (2012). World Development Report 2012: Gender equality and development . Washington, DC: World Bank.

35.

World Bank. (n.d.). Female entrepreneurship resource point—Module 2: How to make change . Retrieved from https://www.worldbank.org/en/topic/gender/publication/female-entrepreneurship-resourcepoint-module-2-how-to-make-change