Capital, Digitalization, and Formality: Chilean Micro-Enterprises During COVID-19

Abstract

Small businesses, particularly micro-entrepreneurships, are a vital economic engine in many developing nations, providing essential income and employment. This study analyzes the post-shock trajectory of Chilean micro-enterprises through the lens of the digitalization–formality trade-off during and after the pandemic. During the crisis, micro-enterprises relied on a short-run substitution mechanism: those with greater capital and intensive Internet use saw a notable increase in sales, demonstrating that digital channels were effectively substituting for the growth benefits typically conferred by formal status. Interestingly, formal business registration or permits did not directly translate into higher sales during this period. Looking at the medium-run pattern, the initial surge in necessity-driven businesses was followed by renewed incentives for opportunity-seeking entrepreneurs as the economy recovered. However, the crisis created a lasting disincentive: both men and women were less likely to formalize their businesses after the pandemic, indicating that the high cost or low benefit of formality persisted, further entrenching the reliance on informal, digitally enabled operations.

1. Introduction

A micro-entrepreneur is an individual who starts and operates a very small business venture, typically with limited resources, capital, and staffing (e.g., Alshebami, 2025b). These businesses, known as micro-enterprises or micro-businesses, are generally characterized by having fewer than ten employees, often just one, and requiring minimal initial investment. Examples of micro-entrepreneurs include freelancers, home-based businesses, small-scale retailers, and rideshare drivers, among others (e.g., Tamvada, 2021, Chapter 1).

Micro-entrepreneurs play a vital role in economic development, especially in developing countries, by generating income, creating jobs, and fostering innovation at the grassroots level (e.g., Ruiz-Martínez & Quiroz-Rojas, 2022; Jayachandran, 2021; Ruiz-Martínez et al., 2021; Meressa, 2020). In that sense, micro-entrepreneurship is considered a vehicle for empowering individuals and communities, which involves the intersection of gender, social capital, and entrepreneurship (Crittenden & Crittenden, 2021; Tamvada, 2021, Chapter 2). In recent years, digital platforms and mobile technologies have reshaped how micro-entrepreneurs operate, enabling broader market access and lowering barriers to entry (e.g., Pawełoszeka et al., 2023).

The challenges small businesses face during difficult times, such as economic downturns or crises, have often been overlooked in academic research, with most studies focusing on how large corporations respond to critical situations (e.g., Fairlie, 2020; Childs et al., 2022; Fuming et al., 2022; Chapman Cook & Karau, 2023; Fairlie et al., 2023; Matikonis & Graham, 2024; Scapini & Vergara, 2024; Alshebami, 2025a, 2025b). In this sense, small businesses frequently encounter different forms of market uncertainty, and studies show that a proactive approach is key to adapting. By developing resilience, embracing innovation and agility, and building strong connections with customers and partners, small businesses can do more than just endure; they can convert adversity into advantage (e.g., Chapman Cook & Karau, 2023; Alshebami, 2025a, 2025b).

Filling this significant gap in the literature, this study offers new evidence on how small businesses adapted to the COVID-19 pandemic by analyzing its impact on micro-entrepreneurial activities in Chile. In doing so, this study makes several key contributions. First, this research provides valuable insights into the Chilean context, where micro-enterprises play a significant role in the country’s economy, contributing to job creation and economic growth. Indeed, according to the Chile Internal Revenue Service, micro-firms accounted for 72% of all registered companies and 8% of permanent jobs in 2023. By focusing on Chilean micro-entrepreneurs, the study offers a geographically specific understanding of crisis adaptation.

Second, the novelty of this analysis lies in its comprehensive approach, considering a range of economic and social factors impacting micro-entrepreneurial activities. This includes crucial elements such as capital endowment, time allocated to domestic and caregiving duties, business location, and business formalization—factors often overlooked in broader studies.

Furthermore, by utilizing data from the 6th and 7th Micro-entrepreneurship Surveys (EME) conducted before and after the pandemic, this research offers a unique longitudinal perspective. This allows for an examination of trends before, during, and after the crisis, contributing to a deeper understanding of the long-term implications of global crises on micro-entrepreneurship.

Finally, the study’s findings offer practical implications by underscoring the importance of specific policy interventions. The research highlights the need for policies that address barriers to business formalization, gendered constraints, and support opportunity-driven entrepreneurship in the post-crisis economic recovery phase, among others.

Findings can be classified into three aspects. (i) Pandemic effects and digital substitution: During the pandemic, the study found a digitalization-formality trade-off in achieving sales growth. Specifically, sales were positively associated with intensive Internet usage, indicating that digital channels served as a low-cost substitute for formal compliance when administrative capacity was strained. Conversely, business registration and municipal permits did not necessarily contribute to more sales, suggesting that the benefits of formality were outweighed by its costs during the acute phase of the crisis. (ii) Formality’s enduring drivers and post-shock decline: The research isolates the factors that typically drive formality in this post-shock environment. Business formality was positively influenced by capital endowment and opportunity-driven entrepreneurship, confirming that formality remained complementary to growth strategies requiring investment. However, formality was negatively associated with time spent on unpaid work, highlighting a gendered constraint mechanism. Critically, formality levels decreased for both men and women post-pandemic, underscoring a lasting reluctance or inability to reintegrate into the formal system after the shock. (iii) Entrepreneurial motivation and trajectory: The analysis maps the trajectory of entrepreneurial motivation, noting that the crisis fueled a wave of necessity-driven businesses. As the economy recovered, the post-pandemic period saw renewed incentives for opportunity-driven entrepreneurs, whose growth orientation and higher capital drove the demand for the re-emerging complementarity of formal institutions.

This article is organized as follows. Section 2 presents a conceptual framework for research hypothesis development. Section 3 presents the data and refers to methodological aspects. Section 4 and Section 5 present and discuss the empirical findings, respectively. Section 6 closes by summarizing the main conclusions, presenting policy implications, and suggesting future research topics.

2. Conceptual Framework

2.1. Domestic Context: Chile’s COVID-19 Pandemic Strategy

Chile’s response to the COVID-19 pandemic evolved through several phases, reflecting the challenges of balancing public health, economic stability, and social equity (Aguilera et al., 2022). Chile reported its first confirmed case on 3 March 2020. The government implemented early measures like closing borders, suspending schools, and imposing quarantines in high-risk areas. Instead of nationwide lockdowns, Chile adopted a dynamic quarantine strategy, targeting specific regions and adjusting restrictions based on infection rates. Chile quickly ramped up its testing capacity and contact tracing efforts, becoming one of the leaders in Latin America for testing per capita. In this sense, Chile became a global leader in vaccination rollouts. The vaccination campaign began in December 2020, prioritizing healthcare workers and high-risk groups. By mid-2021, a significant portion of the population was fully vaccinated. Chile was among the first countries to implement widespread booster doses to counter the waning immunity of early vaccines.

2.2. Research Hypotheses

2.2.1. Business Startups

During the COVID-19 pandemic, which significantly disrupted global economies and labor markets, many women faced unique pressures that pushed them toward entrepreneurship as a means of financial stability. Indeed, industries like retail, hospitality, and education—where women are disproportionately represented—were hit hardest by pandemic-related closures. This pushed many women to seek alternative income sources, often through starting their own businesses (e.g., Torres et al., 2021). Furthermore, school closures and childcare challenges led many women to leave traditional employment. Entrepreneurship offered the flexibility to balance work with caregiving responsibilities. While some entrepreneurship is driven by innovation and opportunity, necessity-based entrepreneurship—starting a business due to lack of other viable income sources—became more common during the pandemic, particularly among women facing financial pressures (e.g., Uddin & Barua, 2024). Therefore, the first research hypothesis is:

H₁: 

During COVID-19, women were more likely to start businesses out of economic necessity.

2.2.2. Flexible Businesses

According to the IMF, the informal economy involves activities that have market value and would add to tax revenue and GDP if they were recorded. Alternatively, the Women in Informal Employment Globalizing and Organizing (WIEGO) refers to the informal economy as a diversified set of economic activities, enterprises, jobs and workers that are not regulated or protected by the state. Such activities do not include illegal ones, such as drug running or human trafficking. According to the IMF, around two billion workers are employed in informal jobs, and four out of every five businesses are not formally registered (https://www.weforum.org/stories/2024/06/what-is-the-informal-economy/ (accessed on 10 November 2024).

New and unregistered businesses often operate with less bureaucracy and overhead, allowing them to quickly pivot to meet emerging demands (e.g., Estrin et al., 2024). For example, businesses that shifted to producing essential items like masks, sanitizers, or home delivery services often saw an increase in sales. In this regard, the Global Entrepreneurship Monitor (2022) highlights how informal entrepreneurship surged during the pandemic, particularly in emerging markets, where economic necessity led many to start businesses in underserved sectors.

During the COVID-19 pandemic, many micro-entrepreneurs turned to online sales channels, leveraging social media, e-commerce platforms, and delivery services. In addition, some micro-entrepreneurs expanded their product lines to meet changing consumer needs or take advantage of new opportunities (e.g., Parker et al., 2023; Sagala & Őri, 2024). These low-cost tools provided immediate access to customers; especially as traditional retail channels faced closure. Unregistered businesses often cater to hyper-local markets. With supply chain disruptions affecting larger competitors, smaller businesses had opportunities to meet local demands more effectively. Therefore, the second research hypothesis is:

H₂: 

Unregistered and new micro-businesses, which used the Internet more intensively, were more likely to increase their sales during the pandemic.

2.2.3. Family- and Opportunity-Driven Entrepreneurship

Opportunity-driven entrepreneurs often start businesses to exploit a gap in the market, driven by innovation or growth potential. This proactive approach aligns with formal registration to access broader markets, legal protections, and financing opportunities (e.g., Huang et al., 2023). In turn, family-driven micro-entrepreneurs, who typically inherit or start businesses to sustain family legacy, often benefit from pre-existing resources, such as networks, assets, and knowledge (e.g., Fernandez, 2023), encouraging formalization.

Formal registration facilitates access to government support, loans, and markets (e.g., Rand & Torm, 2012), creating a strong incentive for family and opportunity-driven entrepreneurs to formalize their businesses. Necessity-driven entrepreneurs, conversely, are more likely to remain informal due to resource constraints or the urgent need to generate income. Opportunity-driven entrepreneurs often envision scalability, internationalization, or product innovation (e.g., Estrin et al., 2024), all of which benefit from formal business structures. Similarly, family businesses often prioritize longevity and reputational integrity (e.g., Miller & Le Bretton-Miller, 2021), making formalization a logical step.

Despite the advantages, the costs and complexities of formalization (e.g., tax compliance, legal registrations) can deter some family and opportunity-driven entrepreneurs from developing economies from registering. The likelihood of registration also depends on local norms, legal frameworks, and incentives (e.g., Ulyssea, 2018; Trinajstić et al., 2022). For instance, in regions where informal businesses dominate, even well-resourced entrepreneurs may opt to operate informally due to less rigid enforcement or cultural acceptance of informal enterprises. Hence,

H_3a: 

Family- and opportunity-driven micro-entrepreneurship is more likely to be associated with registered businesses and higher levels of capital endowments.

On the other hand, as discussed in Section 2.2.1, job losses, reduced working hours, and economic uncertainty forced people to find alternative ways to generate income and support their families. These ventures were often based on immediate needs and readily available skills or resources. As the pandemic’s immediate economic shock subsided and job markets began to recover, the urgent necessity driving these ventures would likely diminish for some. People might return to formal employment or find more stable income sources, leading to the closure of some necessity-driven micro-enterprises and to shifting motivations and priorities.

Specifically, the pandemic may have reinforced the importance of family time, work–life balance, and the desire to build a business that aligns with family values and provides flexibility. As the immediate crisis eased, individuals might have been more inclined to pursue ventures that involve family members, build a legacy, or provide a better quality of life for their families, rather than solely focusing on immediate survival (e.g., Alguera Kleine et al., 2024). The pandemic also created new needs, gaps, and opportunities in the market. These could be related to changes in consumer behavior, technological advancements, or shifts in societal priorities (e.g., Al-Omoush et al., 2020). As the economy stabilized and new trends became clearer, individuals with a more strategic and forward-looking mindset were better positioned to identify and capitalize on these emerging opportunities, leading to a rise in opportunity-driven ventures. Hence,

H_3b: 

Family- and opportunity-driven micro-entrepreneurship gained momentum after the pandemic.

2.2.4. Business Formalization

There are two broad views about business formalization (Jayachandran, 2021). The first one states that small businesses generally do not need or want the benefits that come with formalization, such as access to credit or broader markets, because they operate in a different sphere than larger firms. In this sense, this dual economy theory states that informal businesses tend to have low productivity, so they would not thrive in the formal sector. Consequently, encouraging them to register will not promote economic growth (La Porta & Shleifer, 2014). The second view is that informality is a rational response to the excessive bureaucracy, high costs, and complex regulations of the formal economy. Assets that exist in the informal sector but lack legal recognition, such as land or businesses without formal titles or registration, are denominated dead capital. These assets cannot be leveraged to secure loans, attract investment, or scale operations because they are not formally documented (e.g., de Soto, 2001).

Overall, empirical evidence shows that formalization in emerging economies does not necessarily trigger micro-enterprise growth. For example, a study conducted by de Mel et al. (2013) for Sri Lanka, informal firms were offered incentives to formalize (e.g., covering registration costs, financial grants, or business services). However, most businesses did not see a significant increase in profits or access to credit after formalizing. Hence, formalization alone did not guarantee better outcomes if complementary factors (e.g., access to markets, infrastructure, or enforcement of legal rights) were missing. Campos et al. (2018) in turn concluded that formalization increased the likelihood of Malawi businesses applying for loans but did not significantly improve their ability to secure credit or grow. Therefore, formal registration may be insufficient if financial institutions remain risk-averse or disconnected from small firms.

Examples of studies that have found a positive impact of formalization in emerging economies are McKenzie and Sakho (2010), Fajnzylber et al. (2011), and Besley and Persson (2013). Specifically, McKenzie and Sakho (2010) concluded that formal firms in Bolivia gained access to larger markets and government contracts, which contributed to an increase in profitability compared to their informal counterparts. Fajnzylber et al. (2011) studied a Brazil’s tax simplification program for small businesses. They concluded that formalized businesses experienced higher revenues and growth in the medium term, particularly when they also gained access to formal financial services. Besley and Persson (2013) in turn concluded that formalization policies, when paired with effective tax incentives, led to increased tax compliance and higher government revenues, which could be reinvested in public goods and services. Thus, a well-designed tax system can promote formalization and boost national development.

A study by Ruiz-Martínez and Quiroz-Rojas (2022), aimed at identifying the factors that affected formalization of Chilean micro-entrepreneurs in the pre-pandemic period, revealed that accounting usage, income level, and intellectual capital were positively associated with the likelihood of formalization of men and women. Conversely, the probability of formalization decreased as the number of hours of non-remunerated work increased. This factor affected women more severely.

During the pandemic, activities such as home-based food production, online reselling, and freelancing saw growth. Some entrepreneurs avoided formal registration to bypass licensing fees, taxes, or health compliance costs, which were seen as burdensome (e.g., Estrin et al., 2024). This phenomenon was especially pronounced in developing countries and among vulnerable populations (International Labor Organization—ILO Brief, 2020). It is likely these barriers remained significant deterrents even after the pandemic ended:

H₄: 

Business registration did not resume after the pandemic.

2.2.5. An Institutional-Complementarities Perspective

Figure 1 presents a model representation for the research hypotheses of this study. The institutional-complementarities perspective suggests that the relationship between formal institutions (like business registration and regulation) and firm behavior (like using digital channels and achieving growth) is dynamic (e.g., Fernandez, 2025; Simarasl et al., 2024; Aparicio et al., 2021; Godley et al., 2021; Urbano et al., 2019). The hypotheses can be framed within this perspective by relating them to the substitutability of digital channels for formal compliance when the formal system was strained, and the potential complementarity between formality and growth when institutions stabilized.

H₁ (Necessity entrepreneurship): 

Women starting businesses out of economic necessity during the pandemic implies a surge in informal or new micro-businesses where formal compliance may be deferred due to urgency and high costs. The necessity-driven nature points to immediate income generation, often relying on quick-start, low-cost options like digital platforms instead of navigating slow or closed administrative channels.

H₂ (Digital channels and sales growth): 

Unregistered and new micro-businesses (i.e., those with low/no formal compliance) found that intensive use of the Internet (the low-cost digital channel) was a key driver for sales growth during the pandemic. This shows that the digital channel substituted for the growth-enabling functions that are typically associated with formal registration during a period of institutional stress.

H_3a (Formality and endowments): 

Opportunity-driven micro-entrepreneurship (a more growth-oriented goal) is associated with registered businesses and higher capital endowments (which formal status can help secure, for example, bank loans). This is the expected state when the institutional environment is stable, and formality complements growth.

H_3b (Post-pandemic momentum): 

The gaining of momentum by family- and opportunity-driven micro-entrepreneurship after the pandemic suggests a shift back toward stability. Opportunity-driven businesses are those most likely to benefit from the re-emerging complementarity of formality with growth, indicating a return to an environment where formal status is again valued for scaling and accessing resources.

H₄ (Registration did not resume): 

This hypothesis represents a potential breakdown or delay in the institutional stabilization process. If business registration did not resume post-pandemic, it indicates that the high costs or strain on administrative capacity persisted, preventing the full re-emergence of the formality-growth complementarity. The informal, digitally reliant substitution patterns may have become entrenched, or the government’s administrative functions remained impaired, keeping the cost of formality high and discouraging a return to the formal system.

3. Data and Methodological Aspects

3.1. Data

This study is based on the 6th and 7th waves of the Micro-entrepreneurship Survey (EME), which are freely available at http://www.economia.gob.cl/category/estudios-encuestas/encuestas-y-bases-de-datos (accessed on 10 November 2024). The EME has been jointly prepared by the Chilean Ministry of Economy, Development and Tourism and the National Institute of Statistics (INE) since 2013. To date, it is the main instrument for characterizing formal and informal micro-enterprises in the country, providing relevant data for the development and monitoring of public policies in this area. The EME is based on the National Employment Survey (ENE), is nationally and regionally representative, and analyses individuals who are self-employed or own a micro-enterprise with up to 10 employees. It is important to note that the EME does not follow the same group of individuals over time as a panel survey would.

Fieldwork for the sixth and seventh editions of the EME was conducted in May–August 2019 and May–August 2022, respectively. According to the 6th EME, in 2019 there were 2,057,903 micro-entrepreneurs, of which 53.1% were informal and 46.9% formal. Reasons frequently given for not carrying out the formalization procedures with Chile’s Internal Revenue Service (SII) were that the business was too small, or the activity was infrequent, and registration was not essential for the operation of the business. In 2019, 61.4% of respondents were men and 38.6% were women. Women had a higher prevalence of self-employment (89.9%) than men (81.3%).

In turn, in 2022 there were 1,977,426 micro-entrepreneurs, of which 41.7% were women and 59.3% men. About 41.7% of micro-entrepreneurs were formal, a figure slightly lower than that of 2019. Similarly to 2019, the prevalence of self-employment among women was higher, reaching 92.6% compared with 86.3% of men. As a result of COVID-19, there were 410,955 (20.8%) micro-entrepreneurs who started their activity. Of these, 207,586 were women and 203,369 were men, representing 25.8% and 17.3% of micro-enterprises owned by women and men, respectively.

Study variables are listed and defined in

Table A1. List of variables in alphabetical order.

Variable	Description	Type	Source
Accounting	=1 if he/she keeps separate accounting for business.	Binary	EME6 cuentas_ue; EME 7 e1
Breadwinner	=1 if he/she is the main economic provider of the household.	Binary	EME7 proveedor
Business start	=1 if business or self-employment activity was started due to the COVID-19 pandemic.	Binary	EME7 b1
Capital wealth	Monetary value of capital goods—computers, equipment, tools, utensils, vehicles—recorded into thirds.	Categorical	EME6/7 h1
Education	=1 if he/she has tertiary education (i.e., technical higher education, university education, master’s degree, doctorate).	Binary	EME6 cine_eme; EME7 cine_eme_re
Entrepreneurial exp.	=1 if, before starting business, he/she had another activity as a self-employed worker.	Binary	EME7 b8
Family	=1 if motivation for business is a family tradition.	Binary	EME7 motivacion
Foreign	=1 if he/she has foreign citizenship.	Binary	EME 7 nacionalidad
Formal	=1 if he/she registered entrepreneurial activities at Chile’s Internal Revenue Service (SII).	Binary	EME 6/7 Informalidad
Health plan	=1 if he/she contributed to a private or state health plan (Isapre/Fonasa) during the last 12 months.	Binary	EME6 e10_1; EME 7 e11_1
Home-based workplace	=1 if the workplace is a home with or without facilities.	Binary	EME 7 lugar_trabajo
Internet	=1 if he/she uses the Internet for business or self-employment activity.	Binary	EME7 i1
Increased Internet usage	=1 if Internet usage for business purposes increased due to the COVID-19 pandemic.	Binary	EME7 i8
Increased sales	=1 sales increased during the COVID-19 pandemic.	Binary	EME7 d8
Male	=1 if male.	Binary	EME6/7 sexo
Married	=1 if married.	Binary	EME7 est_conyugal
Metropolitan Region	=1 if he/she lives in Región Metropolitana.	Binary	EME7 region
Municipal permit	=1 if he/she got a municipal patent or permit to operate business or self-employment activity.	Binary	EME7 e10
Necessity	=1 if motivation for business was an economic necessity.	Binary	EME7 motivacion
Old business	=1 if business is more than 10 years old.	Binary	EME 6 b2; EME7 b3
Opportunity	=1 if motivation for business was a market opportunity.	Binary	EME7 motivacion
Paid working hours	Daily hours of paid work.	Continuous	EME 7 trabajo_remunerado
Pension fund	=1 if he/she contributed to private or state pension funds (AFP/INP) during the last 12 months.	Binary	EME6 e10_2; EME 7 e11_2
Personnel	=1 if he/she had workers employed or hired for at least one hour a week, including unpaid family members and working partners, during the previous month.	Binary	EME 6/7 f1
Sales	Income from sales of goods/products or provision of services on a monthly basis.	Continuous	EME 7 d5
Services	=1 business or self-employment activity offers services, where the baseline is selling goods/products.	Binary	EME7 c1a
Trained	=1 he/she received any type of training for the economic activity carried out.	Binary	EME6/7 capacitacion
Under 45	=1 if he/she is under 45 years old.	Binary	EME 7 Tramo_etario; EME 6 grupo_edad
Unpaid working hours	Number of daily hours devoted to domestic and care work.	Continuous	EME 7 trabajo_no_remunerado; EME 6 hr_tnr

of the Appendix A.

3.2. Descriptive Statistics

Figure 2a,b depict figures from the merged EMEs 6 and 7, while Figure 3a–c depicts those from COVID-19-related questions contained in EME 7. As shown in Figure 2a, men and women were concentrated in Commerce (1,187,689 micro-entrepreneurs in total) and Services (1,165,535 micro-entrepreneurs in total). As for business registration by sector, Figure 2b shows that the highest rate of registration was in Services (57.3%) while the lowest rate was in Agriculture & Fishing (23.5%).

Figure 3a in turn shows that around 21% of businesses were started due to COVID-19. Among entrepreneurs who started for this reason, Figure 3b shows that the most frequent motivation was necessity (70.8%) followed by opportunity (23.7%). Figure 3c shows the main sources of financing for business expenses (i.e., raw materials, salaries, bills, among others) recorded by EME 7. As can be seen, personal savings and business income were the most important sources for men (16% and 79%, respectively) and women (19% and 76%, respectively). Other sources of financing available during the pandemic, such as emergency family income (IFE), government programs, and pension fund withdrawals, represented individually a negligible share (0.41%, 0.45%, and 0.58%, respectively, for the group of men and women).

3.3. Methodological Aspects

This study uses binary/multinomial logistic and quantile regression models and treatment-effects estimation for observational data—regression adjustment (RA) and nearest-neighbor matching (NNM). Treatment-effects estimation refers to the process of quantifying the causal impact of a treatment, intervention, or exposure on an outcome of interest in observational (non-randomized) data. Unlike randomized controlled trials, where treatment assignments are random, observational data often suffers from confounding and selection bias, making causal inference challenging (e.g., Stuart, 2010).

RA controls covariates’ influence by fitting separate outcome regressions for each treatment level. The average treatment effect (ATE) is then estimated as the difference between the averages of the predicted outcomes across all data points (e.g., Perraillon et al., 2020). NNM in turn estimates the missing potential outcome for each subject by averaging the outcomes of similar individuals who received the opposite treatment. Similarity is determined by a weighted function of the covariates. The treatment effect is the average difference between the observed and imputed potential outcomes for all individuals (Stuart, 2010).

In essence, both RA and matching are powerful, indispensable tools for controlling observed confounding. They enhance the validity of causal inferences in observational studies by making treatment groups as comparable as possible on measurable factors, such as education and prior experience in the entrepreneurial context (EME), though they cannot address unobserved heterogeneity directly.

4. Empirical Findings

This section is organized into three distinct subsections. Section 4.1 is dedicated to testing hypotheses H₁ and H₂, utilizing binary logistic and quantile regression analyses on the COVID-19-related questions in EME7. Subsequently, Section 4.2 and Section 4.3 address the testing of H_3a, H_3b, and H₄, employing binary and multinomial logistic regression models, as well as RA and NNM methodologies, based on data from EMEs 6 and 7. All statistical models, which are fitted with Stata 19, consider EME expansion factors.

4.1. Business Entry and Increased Sales Due to COVID-19: Evidence from EME 7

4.1.1. Entry

Table 1. Logistic regressions for the impact of COVID-19 on business start: EME7.

	(1)	(2)	(3)
	Men	Women	All
	Start	Start	Start
Independent variables	odds ratio	odds ratio	odds ratio
Opportunity driven	0.472 ***	0.476 ***	0.470 ***
	(0.003)	(0.003)	(0.002)
Male			0.827 ***
			(0.004)
Control variables
Entrepreneurial exp.	0.813 ***	1.000	0.899 ***
	(0.005)	(0.007)	(0.004)
Breadwinner	0.724 ***	0.793 ***	0.789 ***
	(0.004)	(0.005)	(0.003)
Married	0.902 ***	0.771 ***	0.840 ***
	(0.006)	(0.006)	(0.004)
Under 45	1.780 ***	2.543 ***	2.126 ***
	(0.011)	(0.016)	(0.009)
Education	1.651 ***	1.496 ***	1.586 ***
	(0.010)	(0.010)	(0.007)
Foreign	3.020 ***	1.638 ***	2.318 ***
	(0.028)	(0.018)	(0.016)
Unpaid/paid working hrs.	1.277 ***	1.126 ***	1.154 ***
	(0.008)	(0.003)	(0.003)
Trained	0.814 ***	0.672 ***	0.736 ***
	(0.006)	(0.004)	(0.003)
Health plan	0.868 ***	1.057 ***	0.949 ***
	(0.007)	(0.009)	(0.005)
Pension plan	0.575 ***	0.733 ***	0.638 ***
	(0.005)	(0.007)	(0.004)
Middle 33rd pctl. capital wealth	0.816 ***	0.649 ***	0.725 ***
	(0.006)	(0.004)	(0.003)
Upper 33rd pctl. capital wealth	0.750 ***	0.575 ***	0.663 ***
	(0.005)	(0.005)	(0.004)
Metropolitan Region	1.097 ***	1.047 ***	1.073 ***
	(0.006)	(0.006)	(0.004)
Observations	1,045,833	660,853	1,706,686
Pseudo R2	0.095	0.087	0.090

Notes: (1) Robust standard errors in parenthesis. (2) *** p < 0.01.

shows binary logistic regression models for the decision to start a business due to the pandemic for men, women, and both groups. The independent variables of interest are entrepreneurial motivation and gender. Control variables include demographics—age, education, breadwinner status, marital status, foreign citizenship, and capital wealth percentile; entrepreneurial experience, training for current economic activity, the ratio of unpaid to paid working hours, health and pension plans, and geographic region (e.g., Ruiz-Martínez & Quiroz-Rojas, 2022). Parameter estimates are expressed as odds ratios so that a given predictor variable is positively (negatively) associated with the dependent variable if its odds ratio is greater (less) than 1.

Regression results for men and women—columns (1) and (2)—show that new entrepreneurs were most likely young, single, educated, untrained, foreign-born, necessity driven, endowed with low levels of capital wealth, and non-breadwinners. Results for the full sample reported in column (3) also indicate that new entrepreneurs were less likely to be men. Indeed, the likelihood of starting a new business for a man was about 3 percentage points lower. This lends support to H₁. From columns (1) through (3), it is apparent that individuals with a relatively higher share of unpaid working hours were pushed into entrepreneurship.

4.1.2. Sales

Table 2. Logistic regressions for the impact of COVID-19 on increased sales: EME7.

	(1)	(2)	(3)
	Men	Women	All
	Increased sales	Increased sales	Increased sales
Independent variables	odds ratio	odds ratio	odds ratio
Increased Internet usage	1.155 ***	1.613 ***	1.282 ***
	(0.006)	(0.011)	(0.005)
Formal	0.986 *	0.920 ***	0.925 ***
	(0.007)	(0.008)	(0.005)
Old business	0.896 ***	0.577 ***	0.781 ***
	(0.006)	(0.005)	(0.004)
Control variables
Male			1.199 ***
			(0.006)
Entrepreneurial exp.	0.676 ***	0.746 ***	0.722 ***
	(0.004)	(0.006)	(0.004)
Breadwinner	1.083 ***	0.870 ***	0.968 ***
	(0.006)	(0.006)	(0.004)
Married	0.820 ***	1.179 ***	0.971 ***
	(0.005)	(0.009)	(0.004)
Under 45	1.376 ***	1.206 ***	1.320 ***
	(0.008)	(0.009)	(0.006)
Education	1.419 ***	1.580 ***	1.449 ***
	(0.008)	(0.012)	(0.007)
Foreign	1.285 ***	1.065 ***	1.186 ***
	(0.013)	(0.014)	(0.009)
Unpaid/paid working hrs.	0.761 ***	0.900 ***	0.899 ***
	(0.005)	(0.003)	(0.003)
Opportunity driven	0.939 ***	1.055 ***	0.989 ***
	(0.005)	(0.007)	(0.004)
Trained	1.073 ***	1.048 ***	1.058 ***
	(0.006)	(0.008)	(0.005)
Health plan	0.994	0.776 ***	0.895 ***
	(0.008)	(0.007)	(0.005)
Pension plan	1.219 ***	1.059 ***	1.211 ***
	(0.010)	(0.011)	(0.008)
Accounting	1.008	1.261 ***	1.103 ***
	(0.007)	(0.010)	(0.006)
Municipal permit	0.893 ***	0.773 ***	0.847 ***
	(0.006)	(0.007)	(0.005)
Home-based workplace	1.467 ***	1.383 ***	1.455 ***
	(0.009)	(0.009)	(0.006)
Middle 33rd pctl. capital wealth	1.340 ***	1.029 ***	1.126 ***
	(0.010)	(0.008)	(0.006)
Upper 33rd pctl. capital wealth	1.533 ***	1.269 ***	1.348 ***
	(0.012)	(0.012)	(0.008)
Personnel	1.210 ***	0.827 ***	1.056 ***
	(0.008)	(0.007)	(0.005)
Metropolitan Region	0.924 ***	0.801 ***	0.871 ***
	(0.005)	(0.005)	(0.004)
Services	0.943 ***	0.601 ***	0.777 ***
	(0.006)	(0.005)	(0.004)
Observations	687,459	466,648	1,154,107
Pseudo R2	0.038	0.059	0.038

Notes: (1) Robust standard errors in parenthesis. (2) *** p < 0.01, * p < 0.1.

reports logistic regression models for the likelihood that a business increased its sales due to the pandemic. The independent variables of interest are increased Internet usage, formal (registered) business, and age of the business. In addition to the control variables of Table 1, business-related variables are considered: accounting usage, municipal permit, location (home-based), existence of personnel, and economic sector—services, where the baseline is selling goods/products. In this regard, EME treats goods and products interchangeably as one single category.

It is worth noting that the regression models of Table 2 incorporate controls for education, training, and prior entrepreneurial experience, which measure entrepreneurial abilities. These abilities are in turn associated with capital accumulation and proficiency in Internet use. Taking these variables into consideration aims to mitigate potential omitted variable bias stemming from observable abilities. However, the influence of unobservable abilities remains a potential source of bias that is difficult to quantify.

As shown, more intensive Internet usage contributed to more sales. In the full sample, the likelihood of more sales for those businesses using the Internet more intensively increased by 5.5 percentage points.1 Sales also increased with higher levels of capital wealth, help of staff members, accounting usage (men and full sample regressions), and working from home. In contrast, registering one’s business or running an established business reduced the odds of increasing sales (with marginal impacts of −1.7 and −5.5 percentage points, respectively, in the full sample); similarly, for obtaining a municipal permit (a marginal impact of −3.7 percentage points in the full sample). This evidence lends support to H₂. Furthermore, businesses offering services were less likely to increase their sales than those offering goods or products.

In terms of entrepreneurial characteristics, individuals with education and training, not necessarily with entrepreneurial experience, foreign-born and young were more likely to see their sales increase during the pandemic. In contrast, those entrepreneurs who spent relatively more hours on housework and caregiving saw their chances of increased sales decrease. The regression model for the full sample suggests that men were more likely to increase their sales. In the case of marital and breadwinner status, their impact depended on the gender of the entrepreneur. For example, married women were more likely to increase their sales during the pandemic.

Table 3. Quantile regression for log(sales): EME7.

	Men			Women			All
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	Log(sales)	Log(sales)	Log(sales)	Log(sales)	Log(sales)	Log(sales)	Log(sales)	Log(sales)	Log(sales)
Independent var.	q25	q50	q75	q25	q50	q75	q25	q50	q75
Formal	0.245 ***	0.279 ***	0.376 ***	0.245 ***	0.107	0.039	0.225 ***	0.262 ***	0.250 ***
	(0.039)	(0.054)	(0.069)	(0.074)	(0.073)	(0.066)	(0.050)	(0.054)	(0.039)
Control variables
Male							−0.091 **	−0.082 **	−0.178 ***
							(0.041)	(0.036)	(0.037)
Married	−0.053	−0.039	−0.038	0.157 **	0.136 ***	0.191 ***	0.005	0.013	0.060
	(0.048)	(0.044)	(0.061)	(0.069)	(0.039)	(0.054)	(0.036)	(0.038)	(0.059)
Under 45	0.086 *	−0.013	−0.039	0.239 ***	−0.010	−0.014	0.133 ***	0.017	−0.010
	(0.052)	(0.046)	(0.061)	(0.065)	(0.060)	(0.051)	(0.037)	(0.031)	(0.048)
Entrep. Exp.	−0.013	−0.084	−0.044	−0.025	0.032	0.173 **	−0.018	−0.047	0.054
	(0.061)	(0.061)	(0.057)	(0.054)	(0.060)	(0.079)	(0.032)	(0.036)	(0.045)
Breadwinner	0.121 **	0.091	0.112 *	0.185 ***	0.053	0.109	0.128 ***	0.058	0.114 **
	(0.048)	(0.060)	(0.060)	(0.066)	(0.063)	(0.087)	(0.040)	(0.039)	(0.045)
Trained	−0.077	−0.036	−0.056	−0.008	0.010	0.001	−0.031	−0.008	−0.023
	(0.067)	(0.062)	(0.073)	(0.055)	(0.053)	(0.072)	(0.050)	(0.041)	(0.054)
Education	0.158 **	0.161 ***	0.231 ***	0.018	0.071	0.011	0.104 *	0.121 ***	0.196 ***
	(0.065)	(0.057)	(0.077)	(0.068)	(0.053)	(0.092)	(0.060)	(0.044)	(0.052)
Opportunity	0.190 ***	0.147 ***	0.101 *	0.184 ***	0.100	0.086	0.168 ***	0.129 ***	0.112 ***
	(0.040)	(0.043)	(0.053)	(0.059)	(0.066)	(0.058)	(0.027)	(0.038)	(0.041)
Log (Unpaid hrs.)	−0.047 *	−0.002	0.001	−0.093 **	−0.067	−0.075	−0.049 *	−0.027	−0.011
	(0.027)	(0.015)	(0.033)	(0.044)	(0.042)	(0.055)	(0.025)	(0.020)	(0.029)
Middle 33rd pctl. cap	0.196 ***	0.175 ***	0.108	0.206 ***	0.223 ***	0.203 **	0.205 ***	0.153 ***	0.150 **
	(0.043)	(0.049)	(0.079)	(0.062)	(0.059)	(0.086)	(0.045)	(0.053)	(0.059)
Upper 33rd pctl. cap	0.365 ***	0.414 ***	0.390 ***	0.533 ***	0.588 ***	0.483 ***	0.451 ***	0.427 ***	0.434 ***
	(0.064)	(0.047)	(0.077)	(0.076)	(0.078)	(0.070)	(0.043)	(0.057)	(0.074)
Old business	−0.032	−0.112 **	−0.133 *	−0.123	−0.120 *	−0.040	−0.071 *	−0.115 ***	−0.120 ***
	(0.066)	(0.051)	(0.076)	(0.088)	(0.064)	(0.065)	(0.042)	(0.037)	(0.042)
Internet	0.234 ***	0.247 ***	0.254 ***	0.213 **	0.254 ***	0.125 **	0.240 ***	0.256 ***	0.206 ***
	(0.052)	(0.044)	(0.085)	(0.092)	(0.082)	(0.060)	(0.037)	(0.044)	(0.050)
Personnel	0.546 ***	0.582 ***	0.685 ***	0.447 ***	0.599 ***	0.600 ***	0.523 ***	0.576 ***	0.666 ***
	(0.059)	(0.072)	(0.067)	(0.095)	(0.080)	(0.083)	(0.058)	(0.049)	(0.071)
Municipal permit	0.353 ***	0.367 ***	0.408 ***	0.207 **	0.208 **	0.197 **	0.331 ***	0.288 ***	0.292 ***
	(0.071)	(0.077)	(0.115)	(0.092)	(0.088)	(0.092)	(0.045)	(0.054)	(0.056)
Home-based wp	−0.023	−0.097	−0.139	−0.332 ***	−0.415 ***	−0.430 ***	−0.244 ***	−0.257 ***	−0.244 ***
	(0.069)	(0.063)	(0.095)	(0.070)	(0.062)	(0.061)	(0.038)	(0.040)	(0.045)
Region/Sector fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Pseudo R2	0.145	0.158	0.192	0.122	0.122	0.126	0.147	0.158	0.192
Observations	945,540	945,540	945,540	707,610	707,610	707,610	1,653,150	1,653,150	1,653,150

Notes: (1) Robust standard errors in parenthesis. (2) *** p < 0.01, ** p < 0.05, * p < 0.1.

complements the above evidence by reporting quantile regressions for logarithmic sales of men, women, and both groups. The advantage of quantile regressions is that they allow quantifying how the impact of covariates changes when moving from low to high sales levels (e.g., Hao & Naiman, 2007, Chapter 3). The quantiles considered are the 25th, 50th, and 75th. As reported in columns (1) through (3), business registration had a positive and increasing impact on men’s sales. This evidence goes in line with what was discussed in Section 2.2.4, as to the potential benefits of business registration. However, this was not the case for women. For example, for the 50th and 75th quantiles, registration had no discernible impact on women’s log(sales).

In turn, the impact of personnel on log(sales) was consistently positive and increasing for men, women, and both groups. Interestingly, unpaid work hours negatively affected only the 25th quantile of log(sales) for men and women. Meanwhile, being a breadwinner encouraged sales, but only in the 25th quantile when men and women are analyzed separately, and in the 25th and 75th quantiles for the entire sample. Other covariates, such as municipal permit and Internet usage, were positively associated with log(sales), but their numerical magnitudes were not necessarily increasing with the log(sales) quantile.

4.2. Family-, Necessity- and Opportunity-Driven Micro-Entrepreneurs: Evidence from EME 6 & 7

Table 4. Multinomial logistic regression for entrepreneurial motivations: EME6 & 7.

	(1)	(2)
	Family	Opportunity
Independent variables	Relative-risk ratio	Relative-risk ratio
Formal	2.832 ***	1.964 ***
	(0.027)	(0.008)
Middle 33rd pctl. capital wealth	0.814 ***	1.177 ***
	(0.009)	(0.005)
Upper 33rd pctl. capital wealth	1.113 ***	1.691 ***
	(0.013)	(0.008)
Year 2022	0.864 ***	1.135 ***
	(0.007)	(0.004)
Control variables
Male	1.382 ***	1.389 ***
	(0.014)	(0.006)
Married	0.438 ***	0.771 ***
	(0.004)	(0.003)
Under 45	1.332 ***	1.206 ***
	(0.012)	(0.004)
Foreign	0.837 ***	1.145 ***
	(0.017)	(0.008)
Education	0.818 ***	1.127 ***
	(0.008)	(0.004)
Pension fund	1.738 ***	1.104 ***
	(0.023)	(0.006)
Health plan	0.672 ***	1.145 ***
	(0.008)	(0.005)
Log (Unpaid working hrs.)	0.842 ***	0.768 ***
	(0.004)	(0.002)
Personnel	2.118 ***	1.330 ***
	(0.019)	(0.005)
Regional fixed effects	Yes	Yes
Sector fixed effects	Yes	Yes
Number of observations	1,776,832
Pseudo R2	0.086

Notes: (1) Robust standard errors in parenthesis. (2) *** p < 0.01. (3) The baseline motivation is necessity. (4) The sample comprises entrepreneurs that have been in business for 12 years or less.

presents a multinomial logistic regression for entrepreneurial motivations—family tradition, necessity (baseline), and opportunity—for the combined surveys of 2019 and 2022. The sample of interest is relatively new entrepreneurs, who have been in business for 12 years or less. It is important to note that by combining both surveys, it is not possible to isolate a group of new micro-entrepreneurs only. Indeed, EME 6 allows the identification of micro-entrepreneurs who had been in business for nine years or less at the time of the survey, while EME 7 allows the identification of those who had been in business for two years or less, or 12 years or less at the time of the survey. These three groups are therefore considered to be relatively new micro-entrepreneurs.

The independent variables of interest are formal business, capital wealth, and a time fixed effect for the post-pandemic year of 2022 (where the baseline is 2019). Control variables include entrepreneur characteristics (gender, age, foreign citizenship, education), personnel, health and pension plans, time dedicated to domestic work and caregiving, regional and sector fixed effects. Parameter estimates are expressed as relative risk ratios, which can be interpreted as the change in the odds or relative risk of being in a particular category (e.g., motivation), relative to the baseline, for a one-unit change in a predictor variable when holding other predictor variables constant. As in a binary logistic model, a positive (negative) regression coefficient in a multinomial logistic model corresponds to a relative-risk (odds) ratio greater (less) than 1.

As shown in columns (1) and (2), family and opportunity-driven entrepreneurs were more likely to run formal (and staffed) businesses, relative to the baseline of necessity-driven entrepreneurs. (For example, the marginal impact of a formal business in the likelihood of observing a family- or opportunity-driven entrepreneur was 2.6 or 12.9 percentage points, respectively). Opportunity-driven entrepreneurs were likely to belong to the upper 33-percentile of capital wealth, and to be male, single, foreign-born, educated, with health and pension plans. Family-driven entrepreneurs shared most of these characteristics, except that they were more likely to be domestic citizens, less educated, and without access to a health plan. This evidence lends support to H_3a.

The estimation results also suggest that the odds of opportunity-driven entrepreneurship increased in 2022, relative to necessity-driven entrepreneurship (with corresponding marginal impacts of 3.0 and −2.3 percentage points). This result is somehow expected because quarantine periods ended in 2021, so Chile’s economy fully opened the following year. This brought new business opportunities, as economic sectors took their path to recovery. Nevertheless, H_3b receives partial support, as family entrepreneurship did not gain more share than necessity entrepreneurship after the pandemic. It is possible that, while family entrepreneurship is typically associated with long-term goals and, potentially greater resources, continued economic instability may have increased risk aversion among individuals and families. Furthermore, the pandemic could have depleted family savings or limited their resources, making it difficult to launch ventures that often require greater initial investment and planning.

4.3. Evolution of Formalization

Table 5. Logistic regressions for formal entrepreneurial activity: EME6 & 7.

	(1)	(2)	(3)
	Formal	Formal	Formal
	Men	Women	All
Independent variable	odds ratio	odds ratio	odds ratio
Year 2022	0.714 ***	0.787 ***	0.739 ***
	(0.003)	(0.004)	(0.002)
Control variables
Male			1.095 ***
			(0.004)
Married	1.490 ***	1.679 ***	1.525 ***
	(0.006)	(0.008)	(0.005)
Under 45	0.683 ***	0.776 ***	0.695 ***
	(0.003)	(0.004)	(0.002)
Foreign	0.926 ***	1.815 ***	1.232 ***
	(0.009)	(0.021)	(0.009)
Education	2.549 ***	2.266 ***	2.372 ***
	(0.012)	(0.012)	(0.008)
Middle 33rd pctl. capital wealth	1.588 ***	3.083 ***	2.083 ***
	(0.008)	(0.016)	(0.007)
Upper 33rd pctl. capital wealth	4.059 ***	4.799 ***	4.603 ***
	(0.022)	(0.032)	(0.018)
Trained	1.540 ***	1.080 ***	1.294 ***
	(0.007)	(0.005)	(0.004)
Opportunity driven	1.594 ***	1.751 ***	1.670 ***
	(0.006)	(0.008)	(0.005)
Old business	0.661 ***	0.491 ***	0.597 ***
	(0.003)	(0.003)	(0.002)
Internet	2.539 ***	1.490 ***	2.096 ***
	(0.011)	(0.008)	(0.007)
Pension plan	1.879 ***	2.020 ***	1.953 ***
	(0.011)	(0.013)	(0.008)
Health plan	1.877 ***	1.628 ***	1.753 ***
	(0.009)	(0.009)	(0.006)
Log (Unpaid working hrs.)	0.932 ***	0.728 ***	0.897 ***
	(0.002)	(0.003)	(0.002)
Personnel	2.224 ***	2.386 ***	2.365 ***
	(0.010)	(0.013)	(0.008)
Region fixed effects	Yes	Yes	Yes
Sector fixed effects	Yes	Yes	Yes
Observations	1,905,486	1,273,220	3,180,355
Pseudo R2	0.294	0.319	0.283

Notes: (1) Robust standard errors in parenthesis. (2) *** p < 0.01.

, columns (1)–(3), presents logistic regression models for the likelihood of business registration for men, women, and both groups for the combined surveys of 2019 and 2022. As shown at the top of the table, formalization became less likely for men, women, and both groups in the post-pandemic period (with corresponding marginal impacts of −3.2, −2.4, and −2.9 percentage points). For example, the odds for women of registering their businesses fell by 21.3% in 2022 relative to 2019. These odds fell even more for men (28.6%). This evidence lends support to H₄.

From the three columns, it is consistently the case that business registration was positively associated with high levels of capital wealth, marriage, senior age, education, training, opportunity, Internet usage, pension/health plan access, and having staff, while it was inversely associated with business age and time devoted by their owners/managers to unpaid work. This evidence agrees with the findings of Ruiz-Martínez and Quiroz-Rojas (2022).

Table 6. Formalization in 2019 versus 2022.

(a) Regression adjustment estimates of formalization rates.
(i) Full sample
Year	Rate	Robust s.e	z	p > z	95% conf. interval
2019	0.446	0.006	73.82	0.00	0.434	0.458
2022	0.429	0.007	61.91	0.00	0.415	0.442
(ii) Men
Year	Rate	Robust s.e	z	p > z	95% conf. interval
2019	0.456	0.008	58.37	0.00	0.441	0.471
2022	0.441	0.009	46.61	0.00	0.422	0.459
(iii) Women
Year	Rate	Robust s.e	z	p > z	95% conf. interval
2019	0.432	0.009	45.75	0.00	0.414	0.451
2022	0.410	0.010	40.95	0.00	0.391	0.430
(b) Nearest-neighbor matching average treatment effects
(i) Full sample
Year	ATE	Robust s.e	z	p > |z|	95% conf. interval
2022	−0.042	0.009	−4.55	0.000	−0.060	−0.024
(ii) Men
Year	ATE	Robust s.e	z	p > |z|	95% conf. interval
2022	−0.046	0.012	−3.70	0.000	−0.070	−0.021
(iii) Women
Year	ATE	Robust s.e	z	p > |z|	95% conf. interval
2022	−0.050	0.014	−3.59	0.000	−0.077	−0.022

Notes: (1) Treatment corresponds to the post-pandemic period. Regression adjustment models for the control and treatment groups are based on the specification of Table 5. (2) Nearest-neighbor matching estimates are based on the Mahalanobis metric.

presents further evidence in this regard. Specifically, panel (a), (i) through (iii), reports RA-based estimates of business registration rates for 2019 and 2022. In this case, the treatment corresponds to the post-pandemic period. Regression models for the control and treatment groups are based on the specification of Table 5. From (i) through (iii), one sees that business registration fell by 1.7 percentage points in the full sample, and by 1.5 and 2.2 percentage points in the subgroups of men and women, respectively. For example, the estimated registration rate in the full sample for 2019 was 44.6%, with a 95% confidence interval of (43.4%, 45.8%), while that for 2022 was 42.9%, with a 95% confidence interval of (41.5%, 44.82%).

Panel (b) complements this evidence by presenting average treatment effects (ATEs) based on NNM. The covariates used to match individuals in the treatment and control groups are the same as those of the regression adjustment. As shown, the ATEs suggest that registration rates fell in the post-pandemic period by a greater extent than what regression adjustment predicted: 4.2, 4.6, and 5.0 percentage points in the full sample, men subgroup, and women subgroup, respectively. These estimates are closer to the figures observed in EME 6 & 7. For example, among women, business registration rate fell by 4.6 percentage points, from 41.32% to 36.75%, while that of men fell by 4.4 percentage points, from 45.39% to 40.96%, when comparing the samples of 2022 and 2019.

It should be noted that the estimates from both the RA and NNM approaches rely on the strong assumption of selection on observables. This means they only adjust for differences between the pre- and post-pandemic periods (treatment and control groups) that are captured by the covariates included in the models (i.e., the observable characteristics). Consequently, these methods do not account for potential differences due to unobservable factors (such as inherent risk tolerance, psychological impacts of the pandemic, or pre-existing unmeasured skills) that might simultaneously influence both the likelihood of being in the treatment period and the business registration rate. Therefore, the reported estimates should be interpreted as measures of association or conditional average effects, not necessarily as the true causal effect of the post-pandemic period if unobservable confounders are present.

In summary, H₁ (economic necessity spurs female entrepreneurship), H₂ (Internet use boosts sales for new micro-businesses), H_3a (capital-endowed and registered businesses driven by family and opportunity), and H₄ (post-pandemic decrease in new business registration) were empirically supported while H_3b (post-pandemic rise in family- and opportunity-driven entrepreneurship) was only partially supported (because of no evidence of post-pandemic rise in family-driven entrepreneurship).

5. Discussion

The pandemic-induced crisis triggered a short-run substitution mechanism, where the high fixed and variable costs of formality (like registration and permits) became binding. This was evidenced by new micro-entrepreneurs being predominantly necessity-driven (Brixiová et al., 2020), characterized by low capital wealth and untrained status (Estrin et al., 2024). These businesses achieved success by substituting slow formal processes with intensive Internet usage, which was a significant driver of sales growth (Pawełoszeka et al., 2023; Sagala & Őri, 2024; Alshebami, 2025b). This survivalist nature (Alshebami, 2025a) resulted in the counterintuitive finding that formal business registration was inversely associated with increased sales for the overall cohort, aligning with evidence from some emerging economies (Jayachandran, 2021). The positive impact of capital wealth, staff, and accounting usage on sales, however, underscores the enduring value of traditional resources and managerial practices.

As economies transitioned into recovery, a distinct medium-run pattern in entrepreneurial motivation emerged. Opportunity-driven entrepreneurship, which is associated with better-resourced individuals—male, educated, higher capital wealth (Estrin et al., 2024), gained share following the end of lockdowns. This rise signals the normalization of economic activity and the exploitation of new opportunities. However, this growth was not mirrored by family-driven entrepreneurship, which did not gain more share than necessity-driven ventures, thus challenging the notion that familial-resource-based entrepreneurship would necessarily thrive in the post-crisis environment (Calabrò et al., 2021). Finally, the overall decline in business formalization in the post-pandemic period (Alshebami, 2025a) suggests that many entrepreneurs preferred to retain the flexibility and lower costs of informality, indicating sustained uncertainty or insufficient incentives for formalization. A visual representation of these findings is presented in Figure 4.

The above discussion highlights a clear gendered constraint mechanism where unpaid care time depressed formalization and lower-quantile sales. The necessity-driven profile was disproportionately female and burdened by a relatively higher share of unpaid working hours—a global trend exacerbated during the pandemic (Ruiz-Martínez & Quiroz-Rojas, 2022; Kan et al., 2022). This domestic resource-drain limited business growth, evidenced by the negative impact of unpaid work hours on sales at the 25th percentile for both genders. The resulting gender disparity showed that registration had a positive impact on sales only for men at higher quantiles, while having no discernible impact for women, who were likely concentrated in more informal, local business models tied to their constrained resources and domestic responsibilities (Ruiz-Martínez & Quiroz-Rojas, 2022; Brieger et al., 2023).

Nevertheless, one should bear in mind that this study is confined to two distinct periods: the short-run substitution mechanism triggered by the pandemic crisis and the medium-run pattern that emerged during the economic recovery. This temporal limitation means the findings may not apply to the long-term, post-recovery entrepreneurial landscape. The sustained effects of the shift to informality or the ultimate success rates of necessity- vs. opportunity-driven ventures over a longer horizon remain unaddressed. Furthermore, the study may be limited in its ability to disaggregate or prove the direct, structural mechanisms by which domestic burden translates into lower business formalization and depressed sales, beyond the simple correlation with unpaid hours.

6. Conclusions

Micro-entrepreneurship serves as a foundational element of the Chilean economy, fostering employment and contributing significantly to economic development. This article explored the influence of the COVID-19 pandemic on micro-entrepreneurial activity within Chile, with particular emphasis on critical determinants such as business formality, capital endowment, educational attainment, and skills. The integration of a pre-pandemic analytical period enriched the study’s depth, facilitating a temporal comparison of entrepreneurial motivations and formalization trends. This methodological approach proved instrumental in isolating pandemic-specific effects from antecedent patterns.

Concurrently, evidence from other economies indicates that small- and medium-sized enterprises encountered substantial financial strain during the pandemic, thereby curtailing their capacity for investment in innovative initiatives (e.g., Bartik et al., 2020; Fairlie, 2020; Kuckertz et al., 2020; Fairlie et al., 2023; Alshebami, 2025b). Notably, U.S. data shows that small businesses were more likely to permanently close than larger enterprises during the nascent stages of the pandemic (Fairlie et al., 2023). In turn a comprehensive study of small businesses for eight Latin American countries showed that the pandemic had large negative impacts on employment and beliefs regarding the future, and that government policy had limited impact for small and informal firms (Guerrero-Amezaga et al., 2022).

6.1. Main Findings and Contributions

During the pandemic, increased sales correlated with higher levels of capital wealth, help from staff members, more intensive Internet usage, accounting usage, and working from home. In contrast, registering, obtaining a municipal permit, or running an established business reduced the odds of increasing sales. Furthermore, businesses offering services were less likely to increase their sales than those offering goods or products.

On the other hand, the role of domestic and caregiving responsibilities, especially during the pandemic, underscored the gendered nature of business constraints. For example, the number of unpaid hours had a greater negative impact on the likelihood of formalization for businesses led by women than for those led by men. Moreover, the decline in business formality after the pandemic, observed among both men and women, suggests a possible shift towards informal economic activities as people sought faster and less bureaucratic ways to adapt to the crisis. This trend may have been driven by increased barriers to formalization during the pandemic (e.g., administrative closures, financial constraints) and the need to generate immediate income.

In summary, this study contributed to existing literature by filling a notable research gap. Instead of focusing on large, established businesses, it provided an in-depth, longitudinal analysis of micro-entrepreneurs, a segment of the economy that is often overlooked but crucial for economic vitality in emerging economies. The research is particularly valuable because it offers geographically specific insights from Chile, a context that is not only understudied but also serves as a compelling case study for understanding how smaller economies respond to global crises.

By employing a novel and comprehensive analytical approach, the study moved beyond traditional economic models. It not only quantified the impact of various business strategies during the pandemic but also uncovered counterintuitive findings, such as the fact that business formality could hinder, rather than help, sales growth. This methodological design allowed the research to provide a more nuanced understanding of the factors that truly contribute to a micro-enterprise’s resilience. The focus on both informal and formal businesses, and its ability to track their evolution over time, provided a rich, data-driven narrative that challenged conventional wisdom and offered new directions for policy and future research.

6.2. Policy Implications

By implementing a combination of the following policies, governments can aim to foster a more resilient, inclusive, and growth-oriented micro-entrepreneurial sector, addressing both the immediate challenges highlighted by the pandemic and the longer-term structural issues.

(i)

Policies to support sales growth and digitalization

• Digitalization support programs: Implement initiatives that provide financial and technical assistance for micro-businesses to enhance their online presence. This could include subsidies for website development, e-commerce platform integration, digital marketing training, and access to affordable Internet services.
• Capital injection and access to finance: Create accessible micro-loan or grant programs with simplified application processes to provide crucial capital for inventory, technology upgrades, and marketing efforts.
• Promote the adoption of basic business tools: Offer training and subsidized access to user-friendly accounting software and digital management tools. This can help businesses better track finances, manage inventory, and make informed decisions.
• Facilitate business networks and collaboration: Encourage the formation of online and offline networks that allow micro-businesses to share knowledge, resources, and potentially collaborate on marketing or distribution.

(ii)

Policies addressing barriers to formalization

• Streamlined registration and permitting processes: Simplify the processes for business registration and municipal permits by reducing bureaucratic hurdles and associated costs.
• Temporary suspension or reduction in formalization fees: Consider temporary measures to waive or significantly reduce registration and permit fees, particularly for micro-businesses, to incentivize formalization during economic recovery.
• Support for established businesses: Implement measures to alleviate the challenges faced by established businesses during crises. This could include tax breaks, access to credit lines, or tailored advisory services to help them adapt and avoid informalization.

(iii)

Policies addressing gendered constraints

• Caregiving support and flexible work arrangements: Explore policies that support working parents and caregivers, particularly women entrepreneurs. This could include subsidized childcare options, promotion of flexible work arrangements, and awareness campaigns to encourage a more equitable distribution of domestic responsibilities.
• Targeted financial and training programs for women: Design specific programs that provide financial resources, business training, and mentorship tailored to the unique challenges faced by women entrepreneurs.

In that sense, Chile government’s National Support and Care Policy 2025–2030 aims at improving the well-being of caregivers and those in need of care. This is achieved through a 2025–2026 Action Plan with 100 measures developed jointly by various public agencies. In turn, the universal nursery project seeks to guarantee access to a nursery for all workers with children less than two years of age, regardless of their employment status.

(iv)

Overall policy considerations

• Data collection and monitoring: Implement robust data collection mechanisms to track the performance and challenges of micro-businesses, disaggregated by gender and formality status.
• Coordination across government agencies: Ensure effective coordination between different government agencies involved in business registration, support programs, and social welfare to create a cohesive and supportive ecosystem for micro-entrepreneurs.

6.3. Future Research

A follow-up analysis could explore whether the decline in formality persisted after the pandemic or whether entrepreneurs returned to formality. Such an analysis could also investigate how the pandemic affected micro-enterprises in different industries, highlighting the different recovery pathways and challenges. Another aspect to consider would be the long-term effect of the pandemic on micro-enterprise activity, focusing on whether businesses that started out of necessity turned into opportunity-driven ventures.