Internet utilization and Indonesian female entrepreneurs during the COVID-19 pandemic

Abstract

The spread of COVID-19 was a severe setback to the labour market in the globe, including Indonesia. The policies implemented to prevent the spread of the virus have compelled the society to shift towards digital and internet-based activities. The rapid growth of digital technology and internet utilization presents a new potential for women to participate more actively in the labour market, particularly in entrepreneurial activities. This study has two objectives. First, this study examines the effect of using the internet on the participation of female entrepreneurs during the pandemic. Second, this study examines the relationship between internet utilization and income among female entrepreneurs. The data used in this study are individual data sourced from the 2020 National Labour Force Survey (Sakernas). A probit regression model is used to answer the first objective, while the correction for Heckman selection bias has been made when estimating the income equation in the second objective. The results of the first objective show that women who use the internet have a greater chance of becoming entrepreneurs than women who do not use the internet. The impact is higher for women living in urban areas. The finding in this study also emphasizes several important aspects that can encourage the growth of female entrepreneurs, such as training, experience, age, and marital status. Second, this study shows that female entrepreneurs who use the internet tend to have higher incomes. Females that have potential to have higher incomes by using internet are working in trade, food and beverages, and industry sectors.

Keywords:

• COVID-19 pandemic
• female entrepreneurs
• internet
• income
• probit

1. Introduction

Indonesia has been one of the most promising economies with an average annual economic growth of approximately 5–6% (Badan Pusat Statistik BPS, 2019). Employment has grown rapidly in the last two decades, and the unemployment rate has declined almost continuously over sixteen years. The Labour Force Participation Rate (LFPR) in Indonesia was also relatively stable at approximately 66%-67% in the past decade, indicating that the growth of the labour force is mostly affected by the natural growth in population. Although the working age populations of males and females were approximately the same, the LFPR for females was much lower than that for males. While male labour force participation in Indonesia is more than 80%, being one of the highest in the region, female labour force participation in Indonesia remains low at approximately a 52% participation rate in 2019. This number is lower than neighbouring countries, such as Thailand, Singapore and Vietnam, with approximately 59%, 62%, and 73%, respectively, discouraging a potentially greater contribution of working women to the economy.

One important reason for the low level of female labour force participation is women’s child-bearing and child-rearing years, particularly during the prime ages, indicating that many Indonesian female workers have to deal with balancing their careers and their domestic work. A study by Alatas et al. (2020) indicates that getting married and having children (especially young children) has significant effects in lowering women’s labour force participation in Indonesia. Asian Productivity Organization (APO) (2018) also mentioned that the existence of social norms that claim that females are better suited for being housewives is another factor inhibiting female labour force participation in Indonesia. The other reasons include low levels of education, limited job search activities, household welfare effects, and minimal access to migration (Alatas et al., 2020).

The massive growth of digital technology in Indonesia has lately brought new potential to Indonesian women participating more in the labour market. Digital technology has many benefits for women since such media can support them in starting or expanding their microenterprises while working from home (Srivastava & Manzar, 2013). The internet also provides more chances for women working as entrepreneurs, so they can produce additional income for their households (Suwana & Lily, 2017). Women might use digital media for pursuing economic activity, searching for information, and improving insights for better family welfare. They can also find information and inspiration for future economic opportunities in both online and offline business (Kominfo, 2014).

However, Melissa et al. (2015) showed that females in Indonesia generally use the internet at work less than males, partly because males are relatively better represented in formal wage employment, which requires more activities that use the internet and digital technology. While the number of working women in Indonesia has increased, they are still overrepresented in the self-employed category. This is consistent with the narrative of job flexibility: balancing income-generating work with domestic work may be easier in self-employment and entrepreneurship than it is in wage employment activities. According to a study conducted by the Global Entrepreneurship Monitor (GEM) (2020), women in Indonesia, India and Kazakhstan have very strong cultural perceptions in supporting entrepreneurship.

The spread of COVID-19 was a severe setback to the labour market in countries across the globe, including Indonesia. Female workers could be particularly affected by the pandemic because they are more vulnerable than males. Females accounted for close to 50–60% of all employees in the service sector, which has been deeply affected by the pandemic, such as trade, accommodation and restaurants (Manning, 2020). In addition, the job losses by men (i.e., husband or head of household) also negatively impacted their dependent housewife. Some policies to control the spread of COVID-19 in 2020, such as social distancing, lockdowns and restrictions on mobility, had significant effects on economic activity, which have also been reflected in jobs and labour incomes.

Using the individual data from the National Labour Force Survey (Sakernas) year (2020) (2020), this study examines the effects of internet utilization on women’s entrepreneurship activities in the context of the COVID-19 pandemic in Indonesia. Our problem statement, then is how does internet utilization affect the entrepreneurial participation activities for women during the COVID-19 pandemic in Indonesia. As pointed out by Srivastava and Manzar (2013), digital technology has been used to support many females in working to start or to expand their microenterprises. Wahid and Asiati (2021) also suggested that digital technology, particularly social media marketing, may help women in small and medium enterprises in Indonesia survive the COVID-19 pandemic and revive the economy.

The study also examines whether internet utilization affects earnings among female entrepreneurs in Indonesia. In other words, how does internet utilization impact the income of women entrepreneurs during the COVID-19 pandemic in Indonesia? We might hypothesize that women entrepreneurs working with internet utilization to receive higher earnings due to some limited offline activities during the pandemic. Mivehchi (2019) found that information technology plays a significant and vital role both in increasing the marketing and sales of products of entrepreneurs in Iran. The previous study of Indonesia also showed that the use of the internet provides an opportunity to earn higher profits than that without using the internet (Gustina et al., 2020).

There are two main research gaps that are the focus of this study. Firstly, discussions on the relationship between females and internet utilization in developing countries have received relatively little attention. Most of the studies in Indonesia also only focus on the internet for marketing activities and rarely discuss the elements of women entrepreneurs. Secondly, one can also observe a research gap in the changing pattern of female entrepreneurship in the context of the coronavirus outbreak in culturally different countries such as Indonesia. Thus, this study is important to be able to provide knowledge related to increasing the welfare of women entrepreneurs in developing country cases, especially in times of crisis when the mobility is limited, such as in the context of the COVID-19 pandemic.

The organization of the paper is as follows. The next section discusses the literature review, including previous studies and hypothesis. The paper then explains the research methodology used in the study. After the research methodology, the paper reports the results and discussions. Finally, the paper concludes and provides some policy implications, limitations and future studies.

2. Literature review

In the last decade, studies have explored the potential of female entrepreneurs in relation to economic growth (see Ali, 2018; Lock & Lawton Smith, 2016; Maden, 2015; Polas et al., 2022). Women’s entrepreneurship has a positive impact on economic development through job creation, innovation, and creative entrepreneurship (Ali, 2018; Sajjad et al., 2020; Welsh et al., 2016). In Indonesia, women’s entrepreneurial activities also create more jobs and have a positive impact on economic growth and development (Irjayanti & Azis, 2013). Therefore, the more women who participate in entrepreneurial economic activities, the greater the contributions not only to the household but also to the nation’s economy.

In relation to this matter, several previous studies have linked a positive correlation between the growth of digital technology, particularly the internet, and the increase in women’s entrepreneurship in a region. Studies conducted by Brush et al. (2019), Dettling (2017), Sussan and Acs (2017), and Ughetto et al. (2020) have found that the availability of internet-based connectivity is one of the factors contributing to the increasing number of women engaged in entrepreneurship or working from home activities in developed countries. A survey conducted by the OECD in the United States, Australia, Canada, the Philippines, the United Kingdom, and Thailand indicated that the percentage of female entrepreneurs using social media is higher compared to male entrepreneurs (Organisation for Economic Co-operation and Development OECD, 2017). This indicates that women entrepreneurs who utilize the internet are experiencing significant growth.

Compared to developed countries, women in developing countries have a relatively strong cultural perceptions in supporting entrepreneurship (GEM, 2020). Rijkers and Costa (2012) also added that firms led by women tend to be smaller and less productive, as observed in Bangladesh, Ethiopia, and Sri Lanka. Apart from their smaller size, women-led firms are often home-based and employ only one worker, resulting in lower average sales compared to male-led firms. Thus, this study is important to be able to provide knowledge related to increasing the welfare of women entrepreneurs, especially in times of crisis when mobility is limited, such as in the context of the COVID-19 pandemic. In addition, the role of the internet in maintaining the stability and sustainability of women’s businesses has not been widely discussed, especially in countries with diverse cultures such as Indonesia.

Theoretically, there are two mechanisms underlying why the internet utilization would impact entrepreneurship. These mechanisms fall into two categories: (i) social network mechanism and (ii) information and knowledge acquisition mechanism. The first, internet utilization helps to expand individual’s social network and thus promotes entrepreneurship. The interpersonal communication is an important way for people to obtain social and economic resources and maintain and develop social networks (Barnett et al., 2018). Face-to-face contact is the traditional means of interpersonal communication, while communication through the use of socially interactive technology through internet, such as social media, is becoming increasingly popular (Cesaroni et al., 2017). A number of studies demonstrate positive effects of social network on entrepreneurship. The other reason is that social network helps potential entrepreneurs obtain external financing (see Alesina et al., 2013; Basiglio et al., 2019; Ughetto et al., 2020). Potential entrepreneurs are often wealth constrained, and therefore obtaining external financing is central for entrepreneurship in an imperfect credit market.

Second, the initial phase in becoming an entrepreneur is the capacity to identify entrepreneurial opportunities, which are circumstances where the revenue generated surpasses the costs invested in product markets or factor markets (Barnett et al., 2019). People who use the internet products or digitalization can obtain more information, and thus they may have higher probabilities of becoming entrepreneurs than those who do not. Knowledge learned from internet products and applications can increase the cognitive properties necessary to entrepreneurship. Financial literacy, defined to be the basic understanding of economic and financial knowledge and the capability to use that knowledge and other instruments to manage financial resources effectively, is especially important for entrepreneurship (Orobia et al., 2020). Tan and Li (2022) highlight that increasing access to information is an important component for entrepreneurs that has an impact on developing and exploiting business opportunities. This is also explained by Shi and Wang (2017) that the Internet can encourage entrepreneurs to find business opportunities and reduce market risk.

Apart from their important role in the economy, female entrepreneurs face dilemmas in developing their entrepreneurial activities (Bianchi et al., 2016). The literature identifies the barriers faced by most women with small businesses, namely, a lack of business skills and knowledge (Lange et al., 2000), poor marketing and distribution processes (Tejaningrum et al., 2016), and a lack of competence, particularly in the use of internet technology (Mack et al., 2016).

One of the efforts to improve the performance of female entrepreneurs is to increase their skills in utilizing digital technology for business development (Sharafizad, 2016; Sharafizad & Coetzer, 2016). Another study conducted by Mack et al. (2016) showed that in the United States, internet and social media use increased the business competitiveness of female entrepreneurs, especially beginners, so it is necessary to hold a digital technology training program for female entrepreneurs.

Some studies have been conducted in Indonesia on the relationship between the internet and income women entrepreneurs, including Melissa et al. (2015), who analyzed the impact of the internet on empowering female entrepreneurs in five major cities in Indonesia. The results show that internet technology increases their self-actualization and economic benefits for their family.

These facts shows that technology and the internet have played an important role in supporting women in developing their businesses more efficiently. Digital technology plays a very important role for entrepreneurs involved in today’s business situation, particularly during the pandemic, and in places where most people rely on the internet for their needs, including women who have limitations due to their multiple domestic roles as wives and/or mothers, which becomes another obstacle for their business activities (Rehman & Roomi, 2012)

2.1. Research objectives

1. This study examines the effect of internet utilization on the participation of women entrepreneurs during a pandemic. Our hypothesis is that the utilization of the internet enhances the opportunities for women entrepreneurs’ participation.

2. This study examines the effect of internet utilization on the income of women entrepreneurs during the pandemic. Since some lockdown and limited offline activities occurred during the pandemic, we also expect women with internet utilization to receive higher earnings.

3. Research method

The effect of internet utilization on the probability of women working as entrepreneurs will be estimated using probit model. Probit regression is used to model binary (1 and 0) or dichotomous dependent variables using a maximum likelihood method. Therefore, it is suitable to answer the first objective, whereas the dependent variable is a dummy variable consisting of two categories of female workers, namely, entrepreneurial work and nonentrepreneurial work. Compared to other method such as logit regression, it basically provides similar result, while using Linear Probability Model (LPM) which uses the Ordinary Least Square concept will result in the emergence of abnormal error distribution problems and heteroscedasticity violations. The other reason of using probit is that we are trying to estimate the earning equation in the second model because of selection bias issue and probit analysis is a requirement for the usage of Heckman Selection biased model.

As mentioned above, in the probit model, the dependent variable is a dummy variable consisting of two categories of female workers, namely, females with entrepreneurial work and females with nonentrepreneurial work. Entrepreneurial work is proxied by self-employment categories in the National Labour Force Survey, while nonentrepreneurial work includes wage employment, casual employment, and unpaid family work. The main independent variable is a dummy variable representing whether participants use the internet in their working activities. The control variables include several individual and household characteristics as follows:

1. Personal characteristics of female workers, including age (and age squared), marital status, area of residence, which is divided into urban and rural areas, and number of family members.

2. Level of education, which contains the highest education level completed and is categorized into three groups, namely, university levels, senior high schools, and junior high schools and below (as the reference).

3. Sectors of employment is a dummy variable consisting of five main sectors of categories, namely, agriculture, industry, trade, food and drink accommodation, and education services (whereas the service sector is the reference variable).

4. Experience of the respondents, whether women have previous entrepreneurial experience.

5. Training program, whether women participate in government training programs, i.e., the pre-employment card program, particularly during the pandemic.

6. In addition, the study adds the interaction variables of internet utilization with all of the control variables to cover the potential interaction effect of the variables on female entrepreneurs.

Following the method used by Comola and De Mello 2010 and Manning and Pratomo (2013), the study examines the relationship between internet utilization and earnings among female entrepreneurs in Indonesia as the second objective. In practice, however, individuals might select themselves into their preferred work-status category, depending on the level of earnings on offer. This implies that unobserved factors that affect the choice among types of work (and hence work status) are also likely to be correlated with the unobserved factors in the earnings equation, suggesting a potential sample-selection bias in the ordinary least squares (OLS) estimator. To control for this potential sample selection bias, we used Heckman’s selection-biased corrections, based on the binary (probit) equation from the previous estimate, when estimating the earnings equation.

In the Heckman’s model, the dependent variable in this estimate is the log of monthly earnings, while the explanatory variables are broadly the same as in the previous section. At least one explanatory variable likely to affect female entrepreneurs (in the first stage of estimation) but unlikely to affect the outcome variable (earnings) is needed to identify the selection term(s)—otherwise the selection-biased corrections will provide similar results to the OLS estimates. The number of household members and marital status are used as an identifying variable, to tackle selection bias (in the first-stage of the estimation). They are likely to affect work status (female entrepreneurs), but it should not directly affect the outcome variable (earnings). The sample selection term (λ) is also significant, providing evidence of a selection bias in the absence of a suitable correction process.

The data used in this study are individual data sourced from the 2020 National Labour Force Survey (Sakernas). The year 2020 were chosen to show the conditions during the COVID-19 pandemic when the mobilities are limited, there are social distancing and lockdowns in some areas. Sakernas is a regular cross-sectional labour force survey in Indonesia conducted by the National Central Bureau of Statistics (BPS) annually and/or quarterly since 1986. Therefore, Sakernas has been used as the basis for employment data in Indonesia for quite a long time. The main objective of Sakernas is to estimate and monitor the labour force statistics and characteristics in a vast archipelagic nation of Indonesia. This survey provides a rich source of the cross-sectional labour force data, covering about 160,000 respondents (about 0.1% of population) each year. The information included in Sakernas includes sociodemographic information, education, earnings and employment.

The sampling technique used by Sakernas is the Stratified Random Sampling based on census blocks in each district/city in Indonesia. The labour force is divided into groups (strata) based on characteristics such as geographical location, occupation, and education level. The purpose is to ensure that the sample represents the existing variations within the population. Following the stratification, sample units are randomly chosen from each stratum, in a gradual and systematic manner. Once the sample units are selected, data is then collected through a survey. In accordance with this chapter’s objective, this survey is powerful enough to capture most of the changes in the labour force characteristics across provinces.

Table 1 presents summary statistics for the main variables of female entrepreneurs. As presented in Table 1, 36.5 percent of women work as entrepreneurs. This means that more women are entrepreneurs than are employed at other jobs, such as wage employment, casual workers, and family workers. The average number of women who use the internet for their primary work activity is 25.5 percent. This number is relatively low because internet access is still difficult, especially for people in eastern Indonesia (Agahari, 2018). For urban areas, the number of female entrepreneurs is 43.5 percent, which is slightly lower than that for rural areas because urban areas are more represented by wage employment activities. The education level of female entrepreneurs who have graduated from SMA/SMK and university is 20 percent and 10 percent, respectively. A total of 65.3 percent of female entrepreneurs are married due to additional demands in meeting the economic needs of their families. The sector that contributes the most to entrepreneurship for women is the trade and agricultural sectors.

Table 1. Summary statistics

	Mean	SD
Dependent Variables
Women Entrepreneur	0.365	0.481
Log Earnings	13.905	1.004
Independent Variables
Internet	0.255	0.436
Urban	0.435	0.495
Training (Kartu Prakerja)	0.057	0.232
Senior High Academic & Vocational	0.261	0.439
Diploma & University	0.107	0.309
Age	40.669	16.540
Age squared	1927.631	1483.418
Experience	0.365	0.481
Married	0.653	0.475
Number of Families	4.043	1.687
Trade Sector	0.115	0.319
Food & Beverage Accommodation.	0.044	0.206
Agricultural Sector	0.182	0.386
Industry Sector	0.062	0.242
Services Sector	0.109	0.312
Interaction Variable
Internet * Age	9.244	16.782
Internet * Age Squared	367.109	766.352
Internet * Diploma & University	0.119	0.324
Internet * Senior High	0.091	0.288
Internet * Urban	0.160	0.367
Internet * Experience	0.093	0.291
Internet * Training (Kartu Prakerja)	0.037	0.189
Internet * Trade Sector	0.060	0.239
Internet * Food & Beverage Accomm.	0.018	0.134
Internet * Agricultural Sector	0.006	0.081
Internet * Industry Sector	0.026	0.161
Internet * Services Sector	0.134	0.340
Internet * Married	0.173	0.379
Internet * Number of HH. Member	1.046	1.952

4. Results and discussion

4.1. The effects of internet on Indonesian women entrepreneurs

The results of the probit regression are presented in Table 2. Based on the results, internet utilization generally increases the probability of women becoming entrepreneurs. It shows female workers that use the internet have a 33.8 percent greater chance of becoming entrepreneurs than women who do not use the internet. This is consistent with the prediction that the internet makes it easier for women to carry out their entrepreneurial activities. The finding is also similar to the studies conducted by Dettling (2017) who found that the availability and utilization of the internet is one of the factors for the increasing number of female entrepreneurs. In line with the statement above if women use the internet and have sufficient proficiency in internet skills, their propensity to become entrepreneurs will increase significantly and become women entrepreneurs (Shukla et al., 2021) The power of the internet and social media is seen as a solution to the dilemmas faced by women in managing and balancing activities between career and domestic life (Manisha & Tripti, 2015). In practice, the increase in digital-based activities has led to changes in behavior that also allow for forms of business that do not require a special physical place, for example, online sales that do not require a physical store, since the existence of stores can be transferred at home via the internet. In addition, some studies have also proven that the existence of the internet makes it easier for female entrepreneurs to continue to run their businesses even in the midst of the COVID-19 pandemic (see, for example, Wahid & Asiati, 2021).

Table 2. Probit regression results with interaction variables

Variable Dependent: 1 : Women Entrepreneur 0 : Non-Entrepreneur	Probit Without Interaction Effect		Probit With Interaction Effect
	Marginal Effect	P value	Marginal Effect	P-value
Internet Utilization	0.338	0.00	0.020	0.00
Urban Areas	−0.036	0.00	−0.024	0.00
Training (Kartu Prakerja)	0.041	0.00	0.073	0.00
Senior High Academic & Vocational	−0.015	0.01	−0.039	0.00
Diploma & University	−0.053	0.00	−0.112	0.00
Age	0.030	0.00	0.021	0.00
Age squared	−0.000	0.00	−0.000	0.00
Experience	0.014	0.01	0.041	0.00
Married	0.022	0.00	0.030	0.00
Number of Families	−0.004	0.02	−0.002	0.03
Sector of Employment
Trade Sector	0.129	0.00	0.180	0.00
Food & Beverage Accommodation	0.143	0.00	0.163	0.00
Agricultural Sector	−0.100	0.00	−0.073	0.00
Industry Sector	−0.031	0.18	0.025	0.04
Services Sector	−0.232	0.00	−0.243	0.00
Interaction Variable
Internet * Age			−0.018	0.00
Internet * Age Squared			0.000	0.00
Internet * Diploma & University			−0.095	0.00
Internet * Senior High			−0.056	0.00
Internet * Urban			0.023	0.00
Internet * Experience			0.048	0.00
Internet * Training (Kartu Prakerja)			0.039	0.03
Internet *Trade Sector			0.052	0.10
Internet * Food & Beverage Accomm.			−0.000	0.99
Internet * Agricultural Sector			0.080	0.05
Internet * Industry Sector			0.113	0.00
Internet * Services Sector			−0.000	0.97
Internet * Married			0.021	0.02
Internet * Number of HH. Member			0.002	0.25
No. observation		47566		47566
Pseudo R square		0.25		0.26

Comparing urban and rural areas, interestingly, the difference shows that the percentage of female entrepreneurs in urban areas is lower than in rural areas. This is potentially related to job opportunities in rural areas, which are generally much more limited than in urban areas, making entrepreneurship a potential choice in the labour market, while in urban areas, women have more opportunities to work as wage workers or paid employees in formal sectors. A similar result can also be seen in the study of Samantroy and Tomar (2018); Daymard (2015) which stated that female workers who were interested in becoming entrepreneurs were mostly living in rural areas. This is related to the greater confidence of rural residents in the skills of the company.

Then, the challenge is that, as mentioned by Irjayanti et al. (2019), female entrepreneurs in rural Indonesia are usually not familiar with internet-based information technology, and only a small number of them use it as a supporting instrument for their business development. This is in line with the interaction effect in Table 2 when women who are living in urban areas and using the internet have a greater probability of becoming entrepreneurs.

Government support during the pandemic, which is measured by the preemployment training program (Kartu Prakerja), shows that those who have participated in training have a greater chance of becoming female entrepreneurs, indicated by the positive marginal effects (the marginal effect is 0.041) compared to those who did not have training. Thus far, it seems that the program is successful in supporting females, particularly those who are just starting their business, and contributing to their entrepreneurial abilities. This is in line with the research conducted by Parnami and Bisawa (2015), and Tambunan (2017), where the training programs on start-ups or new businesses are important in supporting female entrepreneurs because most female entrepreneurs may not be familiar with the process of starting a business. Further studies have suggested that government should support women through micro-finance, education and training in the development of their skills to give more entrepreneurship opportunities (Sinha & Sinha, 2013).

Women with higher education levels are less likely to become entrepreneurs. The results show that females with tertiary and senior high school education are less likely to become entrepreneurs (with marginal effects of 0.053 and 0.015, respectively). The result suggests that women with higher education tend to be employed as wage employees. The link between education and entrepreneurship is often ambiguous. In developing countries, it is lower education is widely recognized as a barrier to entry into the formal labour market. It will encourage women to create their own business as a way out of response (Minniti & Naudé, 2010), this explains why female entrepreneurship is often higher in developing countries than in developed countries. Daymard (2015) have suggested that workers with lower education will vote as entrepreneurship as a way of escaping from the environment of salaried workers. However, different findings were found in a study conducted by Priya and Bose (2020) in India, which showed that the majority of female entrepreneurs had a higher education at the undergraduate and postgraduate levels compared to a smaller number of female entrepreneurs with a high school education.

The age variable shows positive results, meaning that when women get older, the probability of participating as entrepreneurs becomes greater. However, the aged-square variable shows a negative effect, which means that after women approach a certain age, it tends to reduce their probability of becoming entrepreneurs, possibly due to a decline in their productivity levels. Similar findings were also found in a study conducted by Priya and Bose (2020), where women in -the 20–30 year age group were more likely to start their online business than female in the 30–40 year age group, while women aged 40–50 years and over 50 years were less supportive of starting their own online business.

Previous experience is important. Women who have entrepreneurial experience activities have a 1.4 percent greater chance of becoming entrepreneurs in 2020. Previous experiences seem to be important, particularly when encountering various obstacles, such as asset and skill constraints (Gautam & Mishra, 2016). Similar findings were also found in research conducted by Priya and Bose (2020) and Mack et al. (2016), who support that previous experiences are important for opening a business or becoming entrepreneurs. The interaction effect is also positive and significant, suggesting that experiences combined with internet ability are an asset in becoming entrepreneurs for Indonesian women.

Furthermore, married females have a greater chance than singles of becoming entrepreneurs, as indicated by the positive marginal effects (the marginal effect is 0.022). This is potentially related to household income needs during the pandemic that forced them to do business and try to obtain additional income for the needs of their families. Although the result is supported by Ekpe (2011), who shows that the probability of married women becoming entrepreneurs is greater, Priya and Bose (2020) found that women were more likely to start their online business before getting married.

Women with a large number of family members have a lower chance of becoming entrepreneurs. Although the percentage is relatively small, this may be related to women’s domestic duties because when there are more family members, there are more domestic tasks at home that need to be done. This is consistent with a study by Tambunan (2017) that Indonesian women are still expected to be primarily responsible for taking care of the household and of the family. Therefore, they choose to become family workers rather than entrepreneurs.

Finally, looking at the sector of employment, women employed in trade and accommodation and restaurant sectors are more likely to be categorized as entrepreneurs, while women in agriculture, manufacturing, and services are less likely to be categorized as entrepreneurs. This figure is supported by Prasetyani et al. (2016), who show that female entrepreneurs make the greatest contribution to the trade sector compared to other sectors. The e-commerce survey by Badan Pusat Statistik (BPS), 2019) also shows that female entrepreneurs are likely to be found in the food and beverage accommodation sectors. However, with the assistance of the internet, there are potential entrepreneurial activities for women in agriculture and manufacturing as well, as shown by the interaction effect between internet utilization and sector of employment.

The first column of Table 3 presents the results of the earnings equation using Heckman’s selection-biased corrections model. The selection result from the Heckman model basically explained by probit model in Table 2. For the purpose of comparison, the OLS results (without the selection term) are presented in the second column. The lambda in Heckman’s model is significant, suggesting the proper model to tackle the potential selection-biased issue.

Table 3. Earnings equation for women entrepreneurs

Dependent Variable : Ln Income	Heckman		OLS
	Coef.	P>z	Coef.	P>z
Internet	0.723	0.00	0.327	0.00
Urban	0.133	0.00	0.215	0.00
Training (Kartu Prakerja)	−0.057	0.47	−0.162	0.00
Education
Senior High Academic & Vocational	0.055	0.15	0.364	0.00
Diploma & University	0.156	0.01	0.746	0.00
Ages
Age	0.102	0.00	0.031	0.00
Age squared	−0.001	0.00	−0.000	0.00
Experience	−0.136	0.00	−0.182	0.00
Sector of Employment
Trade Sector	0.924	0.00	−0.074	0.31
Food & Beverage Accommodation	0.912	0.00	0.077	0.06
Agricultural Sector	−0.031	0.82	−0.387	0.00
Industry Sector	0.247	0.10	0.043	0.55
Services Sector	0.032	0.82	−0.343	0.00
Interaction Variable
Internet * Age
Internet * Age Squared
Internet * Diploma & University	−0.070	0.36	0.033	0.39
Internet * Senior High	−0.106	0.07	0.046	0.18
Internet * Urban	0.010	0.84	0.124	0.00
Internet * Experience	0.075	0.12	0.071	0.00
Internet * Training (Kartu Prakerja)	0.061	0.53	−0.260	0.00
Internet * Trade Sector	−0.643	0.00	−0.322	0.00
Internet * Food & Beverage Accomm.	−0.595	0.00	−0.239	0.00
Internet * Agricultural Sector	0.009	0.97	0.311	0.00
Internet * Industry Sector	−0.336	0.04	−0.272	0.00
Internet * Services Sector	−0.359	0.05	−0.053	0.49
Internet * Married
Internet * Number of HH. Member
Constant	10.000	0.00	12.750	0.00
Lambda	0.562	0.00
No. observation		42673		34660
R square				0.235

Using the Heckman method, the use of the internet by female entrepreneurs increased their income by 72,3 percent. This is in line with the study conducted by Gustina et al. (2020). In other words, the use of the internet provides an opportunity to earn higher profits than that without using the internet. Similar results are shown by Mivehchi (2019), & Sasakawa Peace Foundation (SPF) (2017). The findings show that the internet and technology play an important and vital role both in facilitating job access and in increasing the marketing and sales of their products, which has an impact on increasing their income. The use of the internet helps entrepreneurs in market penetration, reduces business costs, improves business services, and adapts to the dynamics of business changes that lead to business efficiency and increased income. Indeed, the internet is available for women this is expected to help facilitate their business activities and increase their income (Asrofi et al., 2023). Further research from Olsson and Bernhard (2021) finding the state that in order to remain competitive and generate business growth, women entrepreneurs constantly have to learn new skills to capture the potential of digitalization especially regarding the knowledge and use of social media. The estimation results of the OLS model also show the same result, which is significantly positive with a percentage of 32.7 percent.

Female entrepreneurs who are living in urban areas have earnings that are 13.3 percent greater than those who are living in rural areas. This is mainly supported by the technological gap between rural and urban areas. The interaction effect suggests that technological developments are generally faster in urban areas, which will encourage entrepreneurs living in urban areas to have higher profits than those living in rural areas (see also Gustina et al. (2020).

Women entrepreneurs with higher levels of education (university and senior high school levels) have higher incomes than women with lower levels of education. This is supported by research conducted by Huarng et al. (2012), & International Finance Corporation (IFC) (2011), where female entrepreneurs with higher education tend to have higher incomes than those with lower education. The reason is that educated women tend to have more and better knowledge about how to make a profit in their business. Similar results were also found in the OLS results but with a smaller coefficient.

The training variable, which was measured by the preemployment training program supported by the government, showed a nonsignificant result in the Heckman model but a significant result in the OLS model. It seems that the implementation of the training has not shown a direct impact on earnings in the short term. However, it is likely that a significant impact will be shown in the long term, especially after the pandemic ends and when the economy is growing faster.

In the age variable, the results show a positive and significant effect, suggesting that increasing age will increase the earnings of female entrepreneurs. However, the age-squared shows negative results, which illustrates that after certain ages, income tends to be reduced, indicating a nonlinear relationship between age and income, which may occur due to increasing age and physical conditions that are no longer productive.

In the employment sectors, the trade and food and beverage sectors provide lower earnings for female entrepreneurs, while agriculture, manufacturing, and services are not significant. However, internet utilization seems to not yet be optimal in supporting the earnings of the main employment sectors because most of the interaction effects are not significant, and some provide negative effects for the trade and food and beverage accommodations sector because these sectors are vulnerable to the COVID-19 pandemic (Rahman et al., 2020). Therefore, there is some room for government support in progressing digital technology to increase the earnings of female entrepreneurs.

5. Conclusions

This study aims to examine the influence of internet use on the participation of female entrepreneurs in Indonesia during the COVID-19 pandemic as well as its implications for the income of female entrepreneurs. Using data from the 2020 National Labour Force Survey (Sakernas), two main findings have emerged from this study. First, in general, this study shows that women who use the internet have a greater chance of becoming entrepreneurs than women who do not use the internet. However, the impact was different for women living in rural areas and cities. The participation of female entrepreneurs in rural areas tends to be higher than in urban areas, and the internet tends to increase the participation of female entrepreneurs who live in cities.

Second, this study shows that female entrepreneurs who use the internet are likely to have higher incomes. Several factors that can be considered in increasing the income of female entrepreneurs are age, living in urban areas, and improvement in education level. The sectors that have the potential to increase the income of female entrepreneurs during the pandemic are trade, food and beverage, and industry.

The results of this study have important implication for policymakers. Various government programs to support improvements and increase digitalization need to be prioritized, especially in efforts to improve welfare as well as women’s participation in the labour market. The other factor is the need of infrastructure development of telecommunications infrastructure in remote areas. The geographical condition of Indonesia as an archipelagic country makes it difficult to provide digital infrastructure that is evenly distributed throughout Indonesia. All of them can also be encouraged through public-private partnerships. Additionally, this study suggests the ongoing exploration of digital literacy training and entrepreneurship training to ensure the functioning and sustainability of women-owned businesses.

However, there are some limitations to this research. This study does not identify specifically what digital media used by the females. In fact, some media may be more effective than other media in supporting employment and earnings. Additionally, the study does not examine the effects of internet utilization for the unemployed women. All of them will be considered as the further studies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study is funded by Direktorat Jenderal Pendidikan Tinggi, Riset, dan Teknology through the grant Pendidikan Magister Menuju Doktor Untuk Sarjana Unggul (NKB-0267/E5/AK.04/2022).

Notes on contributors

Dien Amalina Nur Asrofi

Dien Amalina Nur Asrofi is doctoral candidate and researcher in the Department of Economics, Brawijaya University- Indonesia. Her research interest in labor economics, and economics development. Focusing on female labour, female entrepreneur, and children education.

Devanto Shasta Pratomo

Devanto Shasta Pratomo is a professor in labour economics at Brawijaya University-Indonesia. His research focuses on labour economics, education and migration.

Farah Wulandari Pangestuty

Farah Wulandari Pangestuty is a lecturer in Department of Economics at Brawijaya University, Indonesia. Her expertise is Economic Development, Nutrition and Food Economics, Food Security, and Rural Development.