Effects of household characteristics on low-income family businesses: evidence from India Open Access

We examine the impact of household characteristics on family business outcomes in low-income settings. Family business is a means for survival in low-income settings, aiming to meet consumption of basic needs (Sciascia and Mazzola, 2008). The primary objective of an enterprise initiated by these households is to meet the consumption requirements of the household rather than increase wealth or assets. Such necessity entrepreneurs typically use their existing skills to meet the needs of the household by starting an enterprise. In contrast, family businesses from upper middle- and upper-income levels engage in what is termed opportunity entrepreneurship (Giacomin et al., 2011; Reynolds et al., 2001). The primary objective of these opportunity entrepreneurs is to increase their wealth and quality of their life rather than just meet basic consumption needs.

The low-income end of the income spectrum in emerging economies has also been referred to as subsistence marketplaces. In such economies, the family context of the entrepreneur ^[1] plays an important role in managing small businesses (Prahalad, 2009), blurring the boundaries between the household and the enterprise. Past research has examined such issues as women’s empowerment through entrepreneurship (e.g. Banihani, 2020; Agarwal and Lenka, 2016) in the context of household responsibilities. Prior research in marketing in emerging economies reveals that the household-enterprise link is particularly salient for low-income (i.e. necessity) family businesses (Viswanathan et al., 2010), meeting the basic consumption needs of the household being the primary motivation. Indeed, the necessity of consumption drives entrepreneurship for the consumer–entrepreneur or subsistence consumer–merchant, with household characteristics and the family being front and center. The central role of the household in necessity entrepreneurship has led to the need for academic research on “the nature of the ‘intertwined’ relationship between family business household and business venture” (Carter, 2010, p. 50).

In this paper, we examine a research question of central importance to low-income family businesses. How do individual and household-level characteristics influence economic returns from the enterprise for low-income family businesses? We do so by comparing low-income family businesses driven by necessity, i.e. at lower-income levels, with entrepreneurship at relatively higher incomes, driven by opportunity, analyzing economic benefits and performance of enterprises. Thus, we examine the range of entrepreneurship in emerging markets to provide comparison and contrast for our research question focused on lower-income entrepreneurs. Our study provides insights into the role of the consumer–entrepreneur in growth through different types of enterprises and associated household characteristics, with implications for entrepreneurship theory and practice.

Past studies in the field of entrepreneurship have examined the impact of factors such as the involvement of household members (Sciascia and Mazzola, 2008), gender of the entrepreneur (Kalleberg and Leicht, 1991), education level (Liñán et al., 2011) and the type of enterprise (Mead and Liedholm, 1998) on entrepreneurial motivations and performance. The subsistence marketplaces stream of work has evolved in the discipline of marketing, unpacking the duality of consumer and (necessity) entrepreneur in terms of how consumption and entrepreneurship are two sides of the same coin and how the necessity entrepreneur navigates the domains of family, customer and supplier (Viswanathan et al., 2010). However, a gap in the literature in both disciplines relates to the impact of household characteristics on entrepreneurial outcomes. Such a focus is of particular relevance, given the motivation to engage in entrepreneurship to begin for necessity entrepreneurs of enabling basic consumption for their families (versus higher-income entrepreneurs; heretofore referred to as lower-income entrepreneurs versus higher-income entrepreneurs). We explore the influence of household-level characteristics, such as gender, level of education and involvement of household members in the enterprise and type of business, on the income of the entrepreneur. The study of household characteristics where the necessity for basic consumption drives entrepreneurship is at the intersection of marketing and entrepreneurship. It is also specifically relevant in juxtaposing entrepreneurship in lower-income with higher-income contexts, further understanding necessity versus opportunity entrepreneurship. Insights gained can be used by practitioners to enable effective support of entrepreneurs and their ecosystem and for engendering better returns for low-income entrepreneurs.

Our empirical analysis is based on unique household data from India to highlight the intertwining of household and entrepreneurship. Our findings provide insights into the differential impact of individual and household-level factors on economic returns from lower versus higher-income entrepreneurship. In turn, we derive important implications for marketing and public policy research and poverty alleviation in emerging economies, as our research speaks to the dual roles of consumer and entrepreneur and of consumption and entrepreneurship.

We provide an overview of necessity versus opportunity entrepreneurship, followed by a brief review of household factors that impact entrepreneurial outcomes. Entrepreneurship can be distinguished as being motivated by necessity versus opportunity. The motivations to start an enterprise depend on a variety of factors, ranging from the dire need to earn a basic income due to hardships or lack of other employment alternatives (necessity motivation) to the motivation to exploit lucrative business opportunities (opportunity motivation) (e.g. Reynolds et al., 2001). Among other factors, household income, a significant socioeconomic characteristic, is a strong predictor of entrepreneurial motivations of necessity or opportunity-seeking (Giacomin et al., 2011; Robichaud et al., 2010). In turn, socioeconomic characteristics of the entrepreneur and the entrepreneur’s household have been found to be predictors of these two types of entrepreneurial motivations (Giacomin et al., 2011). Necessity entrepreneurship is prevalent in emerging economies, where the market infrastructure is usually weak and socioeconomic disadvantages are high (Gurtoo and Williams, 2009). Research underscoring the distinction between opportunity and necessity entrepreneurs has shown how the former have higher profits, better management practices and superior cognitive and noncognitive skills compared to necessity entrepreneurs (Calderon et al., 2017). This highlights the significant performance advantages of those who start businesses out of opportunity, typically having relatively higher-income levels. Socioeconomic characteristics of the entrepreneur and the entrepreneur’s household have been found to be predictors of these two types of entrepreneurial motivations (Giacomin et al., 2011).

Necessity entrepreneurs typically run small income-generating activities, microenterprises run by individuals with help from their families. A number of household characteristics may influence the outcomes from such businesses, such as education, gender of the entrepreneur, the number of family members and the type of business as well as the interactions among these factors. The subsistence marketplaces stream in marketing has documented how such entrepreneurs negotiate the domains of customer, supplier and family, the latter serving as a buffer (Viswanathan et al., 2010). The very necessity or motivation for entrepreneurship is for the family to survive and subsist. Ideally, the family supports the entrepreneur who prioritizes the needs of customers and develops relationships with suppliers for the smooth running of the enterprise. Thus, household characteristics and family are intertwined with the activities of necessity entrepreneurs in ways that are distinct from the reality for opportunity entrepreneurs. Furthermore, these subsistence marketplaces have been characterized in terms of one-to-one interactions with responsive, fluid and customized exchanges, enduring relationships with interactive empathy and pervasive interdependence and oral communications (Viswanathan et al., 2012) . Family members can play a central role in the enterprise in such a 1–1 interaction setting. A summary of the literature in this domain is presented in Table 1.

Given the important role that households often have on businesses in emerging markets, we now develop hypotheses about household characteristics and enterprise performance (Figure 1). A number of demographic factors influence the survival of small enterprises. Businesses owned by individuals with higher education and management experience are more likely to survive, whereas older owners and male owners face higher risks of enterprise failure (Ptak-Chmielewska, 2014). Sociocultural factors such as family support, education and social networks play critical roles in entrepreneurial success. Moral and financial support from family significantly impacts business expansion, whereas government policies and access to financial resources are essential enablers for entrepreneurial success in developing economies (Barik et al., 2017). Personality traits and socioeconomic factors, such as economic independence and personal growth, drive women’s entrepreneurship. Challenges such as social barriers and limited access to resources need to be addressed to promote gender equity in entrepreneurship (Gadar and Yunus, 2009). A study of agripreneurs suggested that factors such as education, asset endowment and entrepreneurial skills play a crucial role, with external factors including access to credit, market conditions and infrastructure (Apostolopoulos et al., 2020). Education impacts agripreneurship by enhancing individuals’ ability to access information, make informed decisions and innovate. Another study shows how individual characteristics such as education, family size and membership in social groups, along with external factors such as access to finance and infrastructure, affect diversification decisions for farm business owners (Igwe et al., 2020). These variables impact diversification by providing the necessary skills, resources and social support needed to explore nonfarm activities. For example, education equips individuals with the knowledge and skills to identify and exploit new market opportunities, whereas access to finance enables the acquisition of capital required for business expansion.

To summarize, household characteristics can influence business outcomes for entrepreneurs in general and low-income entrepreneurs in particular. Key factors here include education, gender and family. Next, we develop hypotheses comparing low to higher-income entrepreneurs based on these household characteristics – specifically covering education, gender of the entrepreneur and number of family members involved in the enterprise and, additionally, the type of enterprise. We use the terminology lower versus higher-income entrepreneurs, reflecting necessity versus opportunity entrepreneurs, respectively. Furthermore, we study entrepreneurial outcomes in terms of income and enterprise failure, capturing the level of returns and whether the business is ongoing, respectively.

The human capital perspective emphasizes the role of education in leading to economic growth. Bates (2005) found a positive relationship between the educational background of the owner and the survival of small businesses. Robinson and Sexton (1994) examined US census data and showed a positive relationship between education level and becoming self-employed as well as in succeeding at it [see also similar findings from Rees and Shah (1986) in a study in the UK]. Although experience has a similar effect, education has a stronger effect in comparison (Robinson and Sexton, 1994). Through an examination of a household panel of self-employed individuals in Europe, authors find that formal education has a positive impact on self-employment survival and a negative effect on exiting self-employment (Millán et al., 2012). Whereas prior studies have looked at the education of the entrepreneur, we build on this literature and suggest that the level of education at the household level will also have a positive effect on income for both lower-income entrepreneurs and higher-income entrepreneurs. However, our predictions focus on the differential effects on lower-income entrepreneurs. Whereas prior studies have been at the individual level, our focus is also at the household level.

Examining the determinants of entrepreneurship survival in Spain, Belda and Cabrer-Borrás (2018) find that secondary education does not have a significant impact on the survival of opportunity entrepreneurs but has a positive impact on the survival of necessity entrepreneurs. Fossen and Büttner (2013) studied the returns to education for opportunity and necessity entrepreneurs in Germany. They found that returns to education of the entrepreneur were lower for necessity entrepreneurs relative to opportunity entrepreneurs. Thus, a higher level of entrepreneur education is likely to increase business efficacy and have a more positive impact on enterprise income for lower-income entrepreneurs. By a similar rationale, a higher level of entrepreneur education is likely to have a negative impact on enterprise failure. A similar pattern is predicted for household (adults) education level, given the embeddedness of the business with family for low-income entrepreneurs:

The mutual relationship between entrepreneurs and their families has been studied in entrepreneurship and family business literatures. Dyer and Handler (1994) provided a detailed account of how the family influences various stages of the entrepreneurial process and how household dynamics and entrepreneurial strategy intersect at various points in time. The role of resources in the household is critical for nascent entrepreneurs at the time of starting their entrepreneurial journey. The literature has examined relationships between business strategy and household socioeconomic goals (e.g. Aldrich and Cliff, 2003).

In the context of emerging economies, as noted, a majority of businesses are small businesses, where the individual and the family are involved and the role of the family is important (Prahalad, 2009), accentuated at for low income. There is a blurring of boundaries between developing business strategy and achieving the socioeconomic goals of the household (Aldrich and Cliff, 2003; Viswanathan et al., 2010). Among low-income households, families are not only an active support system for the entrepreneur but also function as risk-sharing systems for family members (Viswanathan et al., 2010). Entrepreneurial motivations, on which the necessity-opportunity classification is hinged, develop dynamically in synchrony with the situation within a household (Jayawarna et al., 2011). Under such conditions, financially disadvantaged individuals start an enterprise for survival to meet the basic consumption necessities of their household, with few, if any, alternatives to do so (Viswanathan et al., 2010). Given this context, it is noteworthy that most studies in entrepreneurship have examined the benefits to the individual entrepreneur and not the household (Carter, 2010) and most databases on entrepreneurs focus on the individual entrepreneur (e.g. Global Entrepreneurship Monitor).

In entrepreneurship, scholars have argued for a family embeddedness perspective to entrepreneurship as family and business are intertwined, and families influence financial, human and physical (e.g. space) resources for the enterprise (Aldrich and Cliff, 2003). The number of family members associated with the enterprise is a proxy for social capital as well as for the size of the team involved in the enterprise. In fact, family capital is a form of social capital that is derived from members of the family (Coleman, 1988). Beyond inexpensive labor, family capital enables trusted relationships and family members are involved in activities such as selling products and collecting payments (Bhagavatulah et al., 2010). As noted, the subsistence marketplaces stream from the marketing discipline speaks to the role of family as a buffer as the entrepreneur navigates the marketplace to serve customer needs (Viswanathan et al., 2010). Family members can play a central role in the enterprise in a 1–1 interactional marketplace described earlier (Viswanathan et al., 2012). Further, family members are likely to have a stronger impact on enterprise income for lower-income entrepreneurs when compared to higher-income entrepreneur households. We suggest that the number of family members involved in the enterprise will have a positive impact on the income from the enterprise for both types of entrepreneurs and the impact will be higher for lower-income entrepreneurs:

Women entrepreneurs face numerous challenges, from access to finances (Baiyegunhi and Fraser, 2014) to household responsibilities (Parvin et al., 2012). Research in this arena has spanned the personal traits of successful women entrepreneurs, including balance, resilience and determination (Banda, 2018). Indeed, women play a major role in a variety of business functions (Pudjiastuti, 2015). Family involvement occurs through such means as labor for microenterprises (Moss et al., 2015). Significant performance disparities exist between male- and female-owned businesses, with male-owned firms generally exhibiting higher turnover (Tandrayen-Ragoobur and Kasseeah, 2017). Women entrepreneurs face significant barriers, including a lack of credit, education and social capital, which impact their business performance compared to male counterparts. These issues are accentuated for low-income entrepreneurs. In this regard, women are more likely to be associated with necessity entrepreneurship than men due to the pressure to overcome their economic insecurity (Calderon et al., 2017).

As noted, prior literature has noted that women face significant social constraints relative to men (Baughn et al., 2006). A consistent finding is that enterprises run by men have higher-income when compared to women (Robichaud et al., 2010; Vossenberg, 2013). A variety of factors place women at a large disadvantage and create a gender gap in developing countries, ranging from access to resources, safety and gender-based violence (affecting hours of operations, location, etc.), work and family balance, lack of societal support (including norms and attitudes), lack of access to information or training and lack of access to financial resources (Vossenberg, 2013). Further, these disadvantages are likely to be higher for lower-income entrepreneurs in terms of the discrimination that women face (Strier, 2010):

We develop hypotheses for the interaction of gender with household-level education and the number of household members involved in the enterprise. As discussed above, in emerging markets, women lower-income entrepreneurs face higher barriers. The higher levels of constraints and discrimination faced by women at lower-income levels suggest that, for lower-income entrepreneurship, the positive effects of household education and number of household members involved in the enterprise-on-enterprise income will be higher for women relative to men:

We consider the type of enterprise, the focal phenomenon, although not a household characteristic. This is a variable that interfaces with household characteristics and entrepreneurship for low-income contexts. We specifically study manufacturing versus retail enterprises, as they bear on the capabilities of lower-income entrepreneurs. Manufacturing enterprises in developing contexts have been associated with difficulty in obtaining capital, the need for specialized skills, larger players and survival in the face of disrupted supply chains (UNIDO, 2013). The literature has emphasized the importance of access to value chains and market linkages in developing countries (Samiee, 1993). Poor infrastructure, transportation costs, inefficient channels and supply chain issues often limit sales to local markets, particularly for low-income entrepreneurs. The challenges to access markets and capital are often higher for low-income entrepreneurs. Retail enterprises may be easier to run in countries with uncertain regulations as they require fewer workers (McPherson and Liedholm, 1996). Necessity entrepreneurs are likely to choose businesses that allow a cost leadership strategy than a product differentiation strategy (Block et al., 2015). Necessity entrepreneurs who have been unemployed for a long duration are likely to choose “home” based enterprises rather than pursue opportunities in external industries (Nikiforou et al., 2019). Furthermore, retail enterprises are closer to the marketplace than manufacturing enterprises and typically in less need for capital. In this regard, low-income entrepreneurs, along with their family members, can use inherent social skills to run retail enterprises, irrespective of income and literacy levels, through 1–1 interactions with customers. All our hypotheses are presented in Figure 1:

To test our hypotheses, we use the nationally representative household panel data set, India Human Development Surveys (IHDS I and II), conducted in 2004 and 2011 by the University of Maryland and the National Council of Applied Economic Research (Desai and Vanneman, 2005, 2011). The first survey wave, IHDS I, conducted in 2004, covers 42,554 households across 33 states. Data was captured from 28 states and 5 union territories (2 union territories – Andaman and Nicobar and Lakshadweep – are islands and were not covered in the surveys). We refer to these states and union territories as “states,” and hence, the survey has data from 33 states (currently, India has 29 states and 7 union territories). The second survey wave, IHDS – II, conducted in 2011, covers 42,152 households (83% of the households from the first wave were reinterviewed). We combine the data from both survey waves to create panel data of 34,621 households. Further, we also used the deflators (based on month-adjusted consumer price index values) specified in IHDS II to convert all amounts (income and expenses) in 2011 to 2004 values.

The use of data from 2004 to 2011 remains relevant and insightful as the study’s focus is on understanding the foundational relationships between household characteristics and entrepreneurial outcomes, which are rooted in structural and socioeconomic dynamics that persist over time. These dynamics, such as the role of education, gender and family involvement, go beyond specific temporal contexts, providing insights into entrepreneurship in low-income settings. Furthermore, the nature of the data set in being panel-based allows for robust longitudinal analysis, capturing changes over time within households, making the findings valuable for both historical understanding and contemporary policy implications. Whereas newer data might reflect updated trends, the fundamental mechanisms explored in this research remain applicable to understanding entrepreneurial behavior in emerging economies today.

The waves of the survey captured a number of variables – whether the household members owned an enterprise (of any scale) and, if so, the details, including revenues, expenses, net income and the list of household members participating in the business. We identify our sample for the main analysis using the following criteria. First, we classify the households into either relatively lower-income (lower-income entrepreneur) or higher-income (higher-income entrepreneur) groups in 2004. Specifically, we use state-level quintiles of households (in terms of income in 2004) and classify the bottom 20 percentile as subsistence households and the top 20 percentile as higher-income households. Second, using the data from both survey waves, we retained a subset of households that matched the following criteria – households that reported owning (at least) an enterprise in both 2004 and 2011, termed as “entrepreneur households.” There were 3,913 entrepreneur households – 317 entrepreneur households in the bottom 20 (lower-income entrepreneurs) and 1,312 entrepreneur households in the top 20 (higher-income entrepreneurs). Thus, in our main analysis, we have a total of 1,629 households that contribute 3,258 observations (over two periods).

We took a conservative approach to operationalizing lower- versus higher-income entrepreneurship using the lowest and highest 20 percentiles. Our data provided income distributions and we used the extreme segments in 2004 to operationalize lower-income entrepreneurship with the lowest 20th and higher-income entrepreneurship with the highest 20th percentiles. As noted, the main analysis sample consists of 1,629 households (317 lower-income entrepreneurs and 1,312 higher-income entrepreneurs) that contribute 3,258 observations.

IHDS data captures the annual net income earned by the household from the enterprise. We use the log of annual net income from the enterprise as the metric of enterprise performance in our analysis.

IHDS I and II capture the number of years of education of all members (adults and children) of a household. We operationalize human capital in two ways – (i) education level of the primary decision-maker of the enterprise, i.e. entrepreneur education; and (ii) average education of all members of the household, i.e. household education.

IHDS I and II capture the number of family members involved in the enterprise. We use this metric to operationalize the household social capital.

The gender of the primary decision-maker is identified for 2004 and 2011. In 2011, the primary decision-maker is specified in the data set. However, in 2004, we used the number of hours spent by every family member involved in the enterprise to identify the primary decision-maker.

For each enterprise, IHDS I and II record the industry classification code – broadly classifying the type of enterprise into nine categories [(i) agriculture, hunting, forestry and fishing, (ii) mining and quarrying, (iii) manufacturing, (iv) electricity, gas and water, (v) construction, (vi) wholesale, retail trade, restaurants and hotels, (vii) transport, storage and communication, (viii) financing, insurance, real estate and business services and (ix) community, social and personal services]. We use the industry codes and classification provided by IHDS to identify enterprises in two categories – retail and manufacturing. When we assess the effect of the type of enterprise, the sample is restricted to those households that run enterprises from these two categories.

IHDS captures detailed household demographic information, enabling us to account for several factors in our analysis. Specifically, we account for the number of adults and children in the household, social group of the household [households are divided into seven social groups (exclusive) – Brahmins, forward castes, other backward classes, Dalits, Adivasis, Muslims and Christians, Sikhs and Jains], location of the household (urban vs rural) and state to which the household belongs. The definition and operationalization of all key variables are presented in Table 2 and their descriptive measures are presented in Table 3.

We use a panel fixed-effects regression framework (using data from 2004 to 2011) to estimate the effect of various factors on enterprise performance and household consumption. Specifically, we estimate the following equation:

where i = 1,.1,629 households, t = 1, 2 (panel time period: 1–2004 and 2–2011), Y_it is the outcome variable – log of income from enterprise for household i, Factor_it represents the factor of interest (education, number of family members, gender and type of enterprise), SCE_i takes the value of 1 if household i is a lower-income entrepreneur household (bottom 20 percentile) or 0 (higher-income) otherwise, HH_i refers to the additional control variables – household composition (no. of adults and children). We add several fixed-effects to our model – industry-year fixed effects (we drop the industry-year fixed effects from our analysis when we assess the effect of type of enterprise), social group-year fixed effects, urban-year fixed effects and state-year fixed effects (to account for time-varying industry-specific, social group specific, location specific and state specific effects on the outcome variables). The coefficients of interest are β₂ (coefficient of Factor), which captures the effect of Factor on the outcome variable, and β₄ (coefficient of Factor * SCE), which estimates the differential effects of the factor between higher-income and lower-income entrepreneur households. In cases where we are interested in additional moderating variables (along with gender, as per H4a and H5a), we include another factor, resulting in a three-way interaction and interpret the results accordingly. Specifically, we include (Gender * SCE * Education) and (Gender * SCE * No. of Family Members) for testing H4a and H5a, respectively.

Results of the analysis evaluating the effects of household characteristics are presented in Table 4. In Columns 1 and 2 (Table 4), we find that the coefficients of interest – β₂ (Education) are positive and significant, irrespective of the operationalization. Results suggest that education (both at the entrepreneur and household level) has a positive effect on enterprise income. Specific to our hypotheses, we find the coefficient of the interaction term (Education * SCE) to be significant when the average education of all household members is used, thus supporting H1c (whereas H1a is not supported). In summary, education has a positive effect on enterprise income. However, the effect of average household education is accentuated for lower-income entrepreneurs.

In Column 3 (Table 4), we find that the coefficient of interest – β₂ (No. of family members in business) is positive and significant (0.206, p < 0.001). Every additional member participating in the enterprise has an incremental effect of 20.6% income from the enterprise. Specific to the hypothesis, we find the coefficient of the interaction term (No. of family members * SCE) to be insignificant; thus, H2a is not supported. In summary, the number of family members involved in the enterprise has a positive effect on enterprise income across the board that is not accentuated for lower-income entrepreneurs.

In Column 4 (Table 4), when we examine the gender effects on enterprise income, we find that the coefficient of interest – β₂ (Gender) is positive and significant (0.516, p < 0.001), suggesting that enterprises run by men, on average, earn 51.6% more than those run by women. However, we find the coefficient of the interaction term (Gender * SCE) to be insignificant; thus H3a is not supported.

In Column 5 (Table 4), when we examine the effects of type of enterprise on enterprise income, we find that the coefficient of interest – β₂ (Type of Enterprise) is negative and marginally significant (−0.146, p < 0.01), suggesting that retail enterprises, on average, earn 14.6% lesser than manufacturing firms. However, we find the coefficient of the interaction term (Type of Enterprise * SCE) to be insignificant; thus, H6a is not supported.

The results of the three-way interaction (Gender * Factor * SCE) are presented in Table 5. Results of the moderating effect of number of years of education of the enterprise decision-maker (human capital) on enterprise income are presented in Table 5 – Column 1. We find the three-way interaction term (Gender * Education * SCE) to be insignificant; thus, H4a is not supported.

Finally, we examine the moderating effect of the number of family members participating in the enterprise (social capital) on enterprise income (Table 5 – Column 2). We find that the three-way interaction term (Gender * No. of family members * SCE) to be negative and significant (−0.415, p < 0.01). We find that the effect of the number of family members involved in the enterprise on enterprise income depends on the gender of the primary decision-maker. Specifically, among lower-income entrepreneurs, the effect of the number of family members on enterprise income is relatively less for enterprises run by men compared to those run by women, thus supporting H5a. Earlier, we reported on findings that a number of family members involved in the enterprise has a positive effect across the board on enterprise income that is not differentially higher for lower-income entrepreneurs. Here, we find that the relationship between the number of family members involved in an enterprise and enterprise income is stronger for women lower-income entrepreneurs when compared to men. Perhaps, this result points to the capacity of women, when compared to men, to leverage social capital within the family to counteract the gender disadvantage that is pervasive.

In summary, gender disadvantages for women persist across income levels for enterprise income. However, where gender begins to have a differential effect for lower-income entrepreneurs is through its interaction with family size. Number of family members involved in their enterprise serves to counter the disadvantages for women.

We use the IHDS household panel data set (IHDS – I and II) and select a subsample for analysis. Similar to earlier analyses, we classify the households into either relatively lower-income (lower-income entrepreneur) or higher-income (higher-income entrepreneur) groups in 2004, using state-level quintiles of households (in terms of income in 2004). We then divide this sample into two groups – (i) households that reported an enterprise in both 2004 and 2011 (“control” group – 1,629 households, of which 317 are lower-income entrepreneurs) and (ii) households that reported an enterprise in 2004, but not in 2011 (“treatment” group – 1,204 households, of which 385 are lower-income entrepreneurs). To summarize, both the control and treatment households own an enterprise in 2004. In the treatment households, the enterprise has failed (enterprise failure being the treatment) sometime between 2004 and 2011 (but due to data limitations, we do not have the actual date of closure of the enterprise). Thus, our analysis sample has 2,833 households. The descriptive measures for the sample are presented in Table 6.

As we have identified the control and treatment households, we use a cross-sectional probit regression using 2004 data to identify the predictors of treatment – enterprise failure by 2011. Specifically, we estimate the following probit model:

where i = 1,0.2,833 households, Y_i takes value 1 if business failed between 2004 and 2011 and 0 otherwise (indicating the household reported a business in 2004 and 2011), Factor_i represents the factor of interest (education, number of family members, gender and type of enterprise), SCE_i takes the value of 1 if household i is a lower-income entrepreneur household (bottom 20 percentile) or 0 (higher-income) otherwise, X_i refers to a vector of covariates including household composition (number of adults and number of children), industry, social group, location and state. We also account for annual income from enterprise in 2004, total annual income of the household in 2004 and monthly household consumption in 2004. To include zero and negative values in the analysis (and enable meaningful interpretation), instead of log transformation, we use an inverse hyperbolic sine transformation [x = log(x+(x²+1)^1/2)] for income from enterprise, household income and household consumption variables.

The coefficients of interest are β₁ (coefficient of Factor), which captures the effect of Factor on outcome variables, and β₃ (coefficient of Factor * SCE), which estimates the differential effects of the factor between higher-income and lower-income entrepreneur households. In cases where we are interested in additional moderating variables (along with gender, as per H4b and H5b), we include another factor, resulting in a three-way interaction and interpret the results accordingly. Specifically, we include (Gender * SCE * Education) and (Gender * SCE * No. of Family Members) for testing H4b and H5b, respectively.

Results of the probit model evaluating the effects of household characteristics are presented in Table 7 (Columns 1–5). In Columns 1 and 2, we find that the coefficients of interest – β₁ (Education) are negative and significant when education is operationalized as the education of the enterprise decision-maker and the average education of the household. Overall, results suggest that education (both at the entrepreneur and household level) reduces the likelihood of enterprise failure. However, we find the coefficient of the interaction terms (Education * SCE) to be insignificant, not supporting H1b and H1d.

In Column 3 (Table 7), we find that the coefficient of interest – β₁ (No. of family members in enterprise) is negative and significant (−0.238, p < 0.001). Every additional member participating in the enterprise reduces the likelihood of enterprise failure. Further, we find the coefficient of the interaction term (No. of family members * SCE) to be positive and marginally significant (0.138, p < 0.10), indicating that, among lower-income entrepreneurs, an increase in the number of members participating in the enterprise increases the likelihood of enterprise failure; thus, H2b is not supported. Juxtaposing the findings from enterprise income, the number of family members involved in the enterprise has a positive effect across the board on enterprise income that is not differentially higher for lower-income entrepreneurs. However, in terms of enterprise failure, the involvement of a larger number of family members increases the likelihood of enterprise failure. Thus, the number of family members has a unique negative effect on enterprise failures.

In Column 4 (Table 7), when we examine the gender effects on enterprise failure, we find that the coefficient of interest – β₁ (Gender) is negative and significant (−0.454, p < 0.001), suggesting that enterprises run by men are less likely to fail compared to those run by women. However, we find the coefficient of the interaction term (Gender * SCE) to be positive and marginally significant (0.297, p < 0.10), indicating that, among lower-income entrepreneurs, enterprises run by women are less likely to fail compared to those run by men, thus supporting H3b. Thus, women subsistence consumer entrepreneurs reverse the across-the-board gender advantage of men in terms of reduced enterprise failures. Juxtaposing the earlier findings on enterprise income, we reported that the gender advantage for men manifests across the board for the enterprise. Thus, it appears that women lower-income entrepreneurs play a unique role in reducing enterprise failures.

In Column 5 (Table 7), we examine the effects of type of enterprise-on-enterprise income and find that the coefficients of interest – β₂ (Type of Enterprise) and the coefficient of (Type of Enterprise * SCE) are insignificant; thus, H6b is not supported.

The results of the three-way interaction (Gender * Factor * SCE) are presented in Table 8. Results of the moderating effect of the number of years of education of the enterprise decision-maker (human capital) on enterprise failure are presented in Table 8 – Column 1. We find the three-way interaction term (Gender * Education * SCE) to be insignificant, thus not supporting H4b. This result is consistent with the lack of a gender effect for enterprise income.

Finally, we examine the moderating effect of a number of family members participating in the enterprise (social capital) on enterprise failure (Table 8 – Column 2). We find that the three-way interaction term (Gender * No. of family members * SCE) is negative and marginally significant (−0.359, p < 0.10). We find that, among lower-income entrepreneurs, enterprises run by women are less likely to fail as the number of family members increase, compared to those run by men, providing support for H5b. This finding suggests that women have a higher capacity to leverage social capital within their families to reduce enterprise failure. Juxtaposing the earlier findings on enterprise income, we reported that the relationship between the number of family members involved in an enterprise and enterprise income is stronger for women lower-income entrepreneurs when compared to men. Thus, the pattern of finding further reinforces the capacity of women lower-income entrepreneurs to leverage social capital within the family.

Our findings are summarized in Table 9. Education (both individual and household level) has a positive effect on enterprise income (for both lower-income entrepreneurs and higher-income entrepreneurs). Further, the effect of the average education of the household on enterprise income is accentuated for lower-income entrepreneurs. Results indicate that education at the individual and household level, as expected, reduces the likelihood of enterprise failure. Interestingly, we find education at both the individual and household levels reduces the likelihood of enterprise failure across the board, but the effect is not stronger for lower-income entrepreneurs.

The number of family members involved in the enterprise has a positive effect on enterprise income across the board. Whereas the number of family members reduces the likelihood of enterprise failure across the board, the effect is lesser for lower-income entrepreneurs compared to higher-income entrepreneurs, counter to our prediction. Thus, the findings are different between income and failure, with the latter leading to an accentuated effect among lower-income entrepreneurs, opposite to what was hypothesized. This finding is unexpected in light of the role of family members in a highly personal 1–1 interactional marketplace for lower-income entrepreneurs.

In terms of gender effects, enterprises run by men had higher enterprise income across the board for all types of entrepreneurs. Enterprises run by men, on average, are less likely to fail compared to women. However, among lower-income entrepreneurs, enterprises run by women are less likely to fail compared to those run by men. This interaction is not found for income. Thus, the findings are different between income and failure, with the latter leading to an accentuated effect among lower-income entrepreneurs.

Interestingly, we do not find any differential gender effects of education for enterprise income or enterprise failure. The gender advantage for women as a function of the number of members in the family involved in the enterprise was higher for lower-income entrepreneurs compared to higher-income entrepreneurs for enterprise income. Thus, women-run enterprises in the subsistence marketplaces benefit more when a higher number of family members is involved in the enterprise. Finally, we find that, among lower-income entrepreneurs, enterprises run by women are less likely to fail as the number of family members increases, compared to those run by men. Thus, the findings are consistent for enterprise income and enterprise failure.

A fundamental quandary in development relates to how entrepreneurship can be enabled in low-income settings. Our research takes a household perspective, given the centrality of the household as the unit of analysis for consumer–entrepreneurs with low income, who engage in entrepreneurship out of necessity to meet basic household consumption needs. The implications of our work for entrepreneurship theory and practice, encompass questions relating to whom to focus on and how. We present the implications of our findings in Table 10.

At a broad level, we note the cross-fertilization between the disciplines of entrepreneurship and marketing. The literature we draw from on women and family entrepreneurship motivated by necessity versus opportunity provides the backdrop of income-based comparisons. The subsistence marketplaces literature provides granular detail about necessity entrepreneurship, i.e. how necessity-driven consumption drives entrepreneurship. The subsistence consumer–entrepreneur or consumer–merchant navigates a different domain, with the domain of family being front and center and serving as a buffer. It also paints a picture of a 1–1 interactional marketplace, delving below umbrella notions of social capital to provide a detailed understanding at the microlevel (Viswanathan et al., 2012). Thus, the research unpacks the nature of exchanges, relationships and the larger context, which the entrepreneur must navigate, often with assistance from their families. Together, these disciplines have the potential to inform understanding of entrepreneurship and consumption, and of consumer–entrepreneurs.

In terms of gender effects, enterprises run by men earned more across-the-board for all types of entrepreneurs, reiterating the need to address the disadvantages that women have to overcome in these settings, irrespective of income. Enterprises run by men had higher household consumption across the board for all types of entrepreneurs. Different factors are likely in operation for lower-income entrepreneurs versus higher-income entrepreneurs, and future research should aim to provide a more nuanced understanding.

The positive impact of education for all entrepreneurs in terms of enterprise income points to the importance of engendering this form of human capital and doing so even within the narrower range of education for lower-income entrepreneurs. Whereas education positively impacts enterprise income, its effect on reducing enterprise failure does not differ significantly between lower- and higher-income entrepreneurs. Hence, the focus should be on education for all members (irrespective of their age group or involvement in the business). Whereas prior findings have highlighted the importance of the individual entrepreneur’s education on the performance of their enterprise, our findings clearly indicate that household education level also impacts performance. As suggested by the literature on the family-embeddedness of entrepreneurship (Aldrich and Cliff, 2003) and the role of family in subsistence marketplaces (Viswanathan et al., 2010), the importance of the households as a unit of analysis for understanding lower-income entrepreneurship in emerging economies cannot be underestimated.

There is no differential effect of gender and education in terms of enterprise income in subsistence households, whereas an advantage for women emerges for enterprise failure. This suggests that among lower-income entrepreneurs, enterprises run by men are able to earn more on average, irrespective of educational differences between men and women. Our findings highlight that, whereas there are benefits from entrepreneurship, they are not evenly distributed. Moreover, educational level among women by itself does not address the gender advantage for men for income but does so for failure. Enterprises headed by women lower-income entrepreneurs continue to face high barriers and their education does not increase their enterprise performance when compared to men (Ahl, 2006).

The number of family members involved in the enterprise has a positive effect on enterprise income for both types of entrepreneurs. Whereas number of family members reduces the likelihood of enterprise failure across the board, the effect is lesser for lower-income entrepreneurs compared to higher-income entrepreneurs.

Among lower-income entrepreneurs, the effect of the number of family members on enterprise income is relatively less for enterprises run by men compared to those run by women. Among lower-income entrepreneurs, enterprises run by women are less likely to fail as the number of family members increase, compared to those run by men. Thus, a consistent reduction in gender disadvantage emerges for women as a function of number of family members in the enterprise for enterprise income and enterprise failure. A tentative conclusion is that women are better able than men to leverage family member participation toward the success of their enterprise. Despite the centrality of social capital and the 1–1 interactional nature of low-income settings, we do not find a differential effect when compared to higher-income contexts for enterprise income but find it for enterprise failures.

At a broad level of the consumer–entrepreneur, practical implications flow from how the understanding of each facet can inform the other – how understanding of the household, family members and their consumption (needs) can inform ways of supporting entrepreneurship and vice versa. In terms of education, given that there are global efforts at entrepreneurial education for women (Viswanathan et al., 2008; Wilson et al., 2007), our findings suggest that social constraints for women in emerging economies need to be simultaneously addressed for women entrepreneurs to benefit from education. Despite the constraints that women face, they show some advantage in preventing enterprise failure.

In terms of number of family members, implications extend to providing a support ecosystem and community networking that would serve as a proxy for families of different sizes in subsistence marketplaces, with a particular emphasis on women entrepreneurs. Examples of possible vehicles include self-help groups in India or Vikobas in Tanzania. Such vehicles can be used to explicitly incorporate marketing and entrepreneurial support at a community level. Whereas such community-level enterprises emerge organically, a research-based approach to the design of a marketing and entrepreneurship ecosystem would systematically incorporate this very important facet.

In terms of practical implications, our research provides insights for governmental, private and social sectors involved in creating marketing or entrepreneurial ecosystems (Table 10). Our research identifies household characteristics that engender entrepreneurial performance and can aid development in this arena of lower-income entrepreneurs. However, public policy and commercial and social sector initiatives should introduce initiatives to change the underlying circumstances that create disadvantages for women when compared to men to begin with.

Lower-income entrepreneurship represents a major proportion of the type of enterprises in emerging economies and an important force for economic growth and development. Taken in total, our findings highlight the reciprocal relationship between households and enterprises of lower-income entrepreneurs in emerging economies, building on and contributing to prior literature in marketing as well as entrepreneurship. Our research provides nuanced household-level insight into the impact of household and individual characteristics on enterprise income and enterprise failure. Together, our conceptualization and findings address a number of research questions and provide a number of avenues for future research at the intersection of marketing and entrepreneurship and important implications for practice.

At a broad level, future research should explore the interplay between consumption and entrepreneurship, with household characteristics being central in such endeavors. In terms of gender, future research should examine a host of issues, including the cultural and social barriers that women in the lower-income group face and how this may play a role in shaping business outcomes. Indeed, gender is a central variable and its interactions with other variables need to be studied. In terms of education, our findings call for renewed enquiry into the constraints faced by lower-income entrepreneurs in emerging economies. Whereas lower-income entrepreneurs have greater returns from entrepreneurship at the household level than higher-income entrepreneurs, lower-income entrepreneurs also face social and economic constraints that limit household income from entrepreneurship. In terms of the number of family members, future research should examine the nature of the educational, marketing and entrepreneurial support that should be part of such ecosystems and their relationships with outcomes for the enterprise and household. Further, future research may examine how the macroeconomic and regulatory frameworks may interact with the microlevel (household) characteristics in determining business outcomes. Finally, although we did not find very high variations in other measures, household size (number of family members) did show some variation between 2004 and 2011. Future research should examine the impact of changes in household composition on business outcomes.

We note some limitations and related directions for future research. Our reliance on self-report data suggests the need for future research to use more objective measures of entrepreneurial performance. Furthermore, beyond entrepreneurial performance, innovation and sustainability also need to be studied. Another direction for future research is in change over time, for instance, how changes in household characteristics affect business outcomes, capturing dynamic aspects of the relationship between household characteristics and business outcomes. In terms of processes, how entrepreneurs exhibit resilience and adaptation is worthy of future research. Future research should also conduct a comparative analysis with family businesses in other emerging economies at different income levels to generalize findings and identify key moderators.