1 Introduction

The self-employment of immigrants could lead to economic benefits for the host country. Studies note significant variations in the self-employment of immigrants. In the year 2019, self-employment in Europe was approximately 14.7%, which exceeded the USA (6.3%), Australia (9.6%) and Canada (8.3%). Nearly 12% of immigrants employed in 2020 in the European Union were self-employed (OECD 2021). For most developed countries with relatively higher immigrant self-employment rates than the developing countries, the self-employment numbers of immigrants were higher than natives (Hunt 2011).

Along with other determinants, self-employment also largely depends on the potential of becoming self-employed, implying that some individuals have certain personality traits that help to become self-employed (Mueller and Thomas 2000). Personality traits are defined as natural dispositions in individuals such as eliciting certain responses to certain situations which may be unique to each individual. Self-employment as a result of personality traits in the context of immigrants is scarcely covered in the existing literature. Therefore, our study fills this gap in the literature. We use entrepreneurship and self-employment interchangeably throughout our study.

While the existing literature reports higher self-employment rates among immigrants than natives (Andersson and Hammarstedt 2010), individuals who migrate may sometimes face more significant barriers to self-employment due to difficulties in integration into the host countries. These include language barriers, restricted access to capital, and a limited understanding of regulatory laws compared to the natives. At the same time, some studies, such as by Borjas (1986), find that many immigrants in the twentieth century were pushed into self-employment due to necessity and financial requirements. Other studies report that immigrants are relatively more skilled and educated in the present times and, therefore, more likely to utilize their education and experience to start businesses in both traditional and non-traditional sectors (Hunt 2011). Barriers to self-employment entry, however, may be less problematic for immigrants who have resided for longer durations in a host country, such as second-generation immigrants. While first-generation immigrants are not born in the host country, second-generation immigrants identify as natives, where either one or both of their parents are first-generation immigrants (Dinesen and Hooghe 2010).

Personality traits that were originally a part of psychological research have also become relevant for economics as a potential means to explain the employment status of individuals. There exists extensive empirical evidence supporting the notion that personality traits are important in explaining self-employment decisions. For example, Zhao and Seibert (2006) showed that the personality structure of entrepreneurs differs from that of managers. By analyzing how personality traits relate to self-employment decisions amongst natives, first-generation and second-generation immigrants, our paper aims to contribute to the different shades of the literature focusing on the relation of personality traits with self-employment. We formally test two hypotheses: (1) hypothesis linking personality traits to self-employment among natives and immigrants, and (2) hypothesis pertaining to differences in self-employment amongst natives and immigrants as a result of education.

Our main objective is to analyze the multidimensional factors such as personality traits and human capital affecting the probability of becoming self-employed among natives, first-generation, and second-generation immigrants. In addition, we aim to provide a recent regional perspective on self-employment determinants for countries in the OECD. We look at specific personality traits instead of the commonly used broad traits (such as the big five) as the former may lead to a wider and comprehensive research direction (Dudley et al. 2006). We look at specific personality traits (including: need for achievement, internal locus of control and, innovation and risk-taking) which are theoretically expected to positively affect running certain businesses and thus may increase the likelihood of self-employment. We focus on natives, first-generation and second-generation immigrants with an aim to fill a gap in the literature by examining personality traits and a range of other determinants on self-employment for these three groups. Our study also examines the strength of this relationship across the types of entrepreneurs, that is, those with and without employees. Individuals who are ‘self-employed with employees’ may aid in job creation due to setting-up relatively larger organizations. Therefore, such a self-employment may be conducive to the socio-economic development of a country (Schoar 2010). Finally, we will look at how belonging to an OECD country impacts self-employment rates amongst all three groups. To the best of our knowledge, our study is the first to use the IVS dataset for such analyses.

We adopt the probit methodology and utilize the unique Integrated Values Survey (IVS) dataset from 2010 to 2021, constructed using the European Values Study (EVS) Trend File and the World Value Survey (WVS) Trend File covering 115 countries (Haerpfer et al., 2020; EVS 2021). We include other determinants such as socio-demographic characteristics and also assess how immigrant networks may contribute to the self-employment of immigrants.

2 Brief theoretical perspective

We briefly shed light on three theories in the entrepreneurship literature, where two are relevant for immigrants; the disadvantage theory and human capital theory, and one for personality traits; the personality theory. The disadvantage theory suggests that immigrants may pursue self-employment as that may be the only way they are expected to survive and succeed in the host nations. The theory argues that drawbacks such as challenges of speaking a new language, prejudice, and unemployment may cause some groups to favor particular individuals over others in the labor market. Thus, by decreasing the returns to immigrants in paid work (including full-time or part-time), such disadvantages may push them towards self-employment (Hammarstedt 2001). On the other hand, the human capital theory has been widely discussed in the immigrant entrepreneurship literature. According to Sanders and Nee (1996), human capital implies education, experience, knowledge, and other characteristics, including motivational incentives and leadership. This human capital assists immigrants in pursuing self-employment primarily through better education and language proficiency and may account for self-employment differences amongst generations of immigrants.

Concerning personality traits, we argue that individuals may be more inclined towards self-employment as a result of certain personality traits. We draw inspiration from the personality theory which suggests that the effect of personality trait variables on self-employment decisions is mediated by the goals and strategies of individuals (refer to Baum and Locke 2004). Brockhaus (1982) mentioned the traits consistently associated with self-employment: the need for achievement, risk-taking behaviors and innovation, and an internal locus of control. Individuals who possess a need for achievement are more likely to be engaged in creativity (Collins, 2004). Kihlstrom and Laffont (1979) illustrate that individuals with a risk-averse personality were more likely to become employees while those possessing higher risk-tolerance were more likely to become entrepreneurs or self-employed. Rotter (1966) introduced expectations pertaining to internal, as well as external control of reinforcement. Individuals who use their own actions to define their future outcomes are believed to possess an internal locus of control. Self-employed individuals are continuously expected to make fast decisions. Therefore, an internal locus of control may be very important in the choice to become self-employed. Our study examines the finer micro personality traits as these may be more highly linked to self-employment.

3 Literature review

3.1 Entrepreneurial trait research

Former literature on entrepreneurship underlines the importance of certain aptitudes and traits required for individuals to be self-employed (Schumpeter 1934). Schumpeter proposed that an entrepreneur must be a leader and must possess an autonomous drive to fight. Individuals considering self-employment must possess the ability to identify opportunities and be wary of every decision they take in terms of production, marketing, and investments. They need to understand the competition they face and acknowledge the high possibility of risk, uncertainty, and loss they may face as an outcome (Schumpeter 1934). Therefore, while human capital and experience influence the decision to be self-employed, several specific skills shaped through different personality traits may also play a determining role in self-employment.

Frese (2009) described how personality traits impact entrepreneurial success as well as moderate the relationship between active performance and entrepreneurship. Active performance includes entrepreneurial orientation, networking, active goals, and so on (see Frese 2009 for more detail). Therefore, personality traits may impact active performance which indirectly translates to entrepreneurial success as well as positively contribute to self-employment directly through innovation and other similar traits. Their study also stressed the role of the environment which defines different contexts (market conditions, networks, and economic development) through which self-employment may be affected. Figure S.1 in the supplementary files, which is an adaptation of the model introduced by Frese (2009), sums up the preceding discussion.

While earlier studies have focused on the big five personality traits (e.g. Zhao and Seibert 2006), our study examines the more micro personality traits as these may be more deeply linked to self-employment. Brockhaus (1982) narrowed down the traits consistently associated with self-employment: the need for achievement, risk-taking behaviors and innovation, and an internal locus of control. Individuals who possess a need for achievement are more likely to be engaged in creative activities that entail future planning and responsibility (Collins, 2004). Thus, individuals who have a high need for achievement are more likely to be attracted towards and be self-employed. To succeed in a competitive market, self-employed individuals need to be able to sell goods and services which involve innovation. Innovation or innovative thinking has been proven to be a prerequisite for self-employment as an individual must possess the ability to create new products or processes, particularly those related to technological products (Schumpeter 1934). In addition, the self-employment literature has also highlighted the importance of the willingness to take risks in self-employment decisions. To be accepting of incurring losses on each investment, self-employed individuals are required to possess a certain level of risk-tolerance. This link was first introduced by Kihlstrom and Laffont (1979). Kihlstrom and Laffont (1979) suggested that individuals with a risk-averse personality were more likely to become employees while those possessing higher risk-tolerance were more likely to become entrepreneurs. Finally, the concept of the locus of control was first introduced by Rotter (1966) to indicate expectations pertaining to internal, as well as external control of reinforcement. In this study, we focus on internal locus of control. Individuals who rely on their own actions to define their future outcomes are believed to possess an internal locus of control. Self-employed individuals are constantly required to make quick decisions, and thus an internal locus of control may be highly significant in the choice to become self-employed.

Summing up, we propose the following hypotheses in terms of personality traits.

H1

Indicators of need for achievement are expected to be positively related to the likelihood of being self-employed.

H2

Indicators of innovation and risk-taking are expected to be positively related to the likelihood of being self-employed.

H3

Indicators of internal locus of control are expected to be positively related to the likelihood of being self-employed.

3.2 Immigrant-native differences in self-employment

Studies show that self-employment among immigrants is increasing (Aldén et al. 2021), where immigrants find self-employment to be more lucrative due to difficulties in entering and succeeding in the labor market, as suggested by the disadvantage theory. Factors affecting the decisions of immigrants to become self-employed may be divided into two categories, namely, push factors and pull factors. Possessing the autonomy to make business decisions, migrant networks, and flexibility in work timings and control (also known as pull factors) may act positively in helping immigrants to become self-employed. Likewise, troubles in landing a job and discrimination in the labor market (push factors) also attract immigrants towards self-employment (Lazaridis and Koumandraki 2003).

Second-generation immigrants, while residents of the same community as their parents, can differentiate themselves from their previous generation in terms of their greater human capital, such as higher education, and better language proficiency. These factors may allow second-generation immigrants to transition towards diverse, white-collared jobs from family-owned businesses (Iyer and Shapiro 1999). Furthermore, second-generation immigrants may receive intergenerational transfers and resources from self-employed parents in the form of human capital, knowledge and contacts, and also generational wealth. Andersson and Hammarstedt (2010) analyzed self-employment across several generations and found that self-employment persisted across generations, mostly for immigrants.

Former studies noted that individuals who are highly educated have a greater tendency to be engaged in jobs in comparison to being self-employed. Simoes et al. (2016) in their study suggested that individuals possessing a higher education in comparison to lower and middle levels of education have a greater chance of being employed in good occupations that pay exceptional wages (which usually may be the case for natives). Thus, they do not feel the need to transition to self-employment. Alternatively, previous literature proposes that immigrants who are highly educated may transition towards self-employment due to having greater knowledge. For instance, Peroni et al. (2016) documented that immigrants with higher education had a higher probability of being self-employed. While immigrants with higher education may have better job opportunities in the labor market in comparison to less educated immigrants, they are also more equipped to recognize self-employment opportunities (Simoes et al. 2016). Further, while earlier studies observed that immigrants were inclined towards self-employment as a necessity and due to financial constraints (Borjas 1986), recent literature suggests that immigrants mostly choose to become self-employed as they view self-employment as an opportunity for a better future. Thus, while immigrants in the past may have set up businesses as a result of necessity, in recent times immigrants have used their education and experience to set up innovative and successful businesses (Hunt 2011). This is complemented by entrepreneurial jobs recently transitioning to those requiring more creativity and technicality, attracting highly educated individuals. Thus, immigrants possessing a higher education have higher chances of being self-employed. We propose the following hypothesis for education.

H4

Natives with lesser education in comparison to higher education are less likely to be self-employed while immigrants with a higher education are more likely to be self-employed.

A large proportion of the literature examines the big 5 personality traits in relation to self-employment (e.g. Zhao and Seibert 2006). Our study alternatively looks at other narrower personality traits as determinants of self-employment. Secondly, despite the relatively large amount of literature focusing on differences in self-employment between immigrants and natives, little is known about the link between personality traits and self-employment among them. In our paper, we aim to fill this knowledge gap, alongside looking at different generations of immigrants. Since the share of self-employed immigrants is shown to be on the rise, where self-employment has also been looked at as an escape for immigrants who face barriers to entry in the labor market, it is crucial to study the determinants of self-employment amongst immigrants in detail.

4 Materials and methods

Our study uses the Integrated Values Survey (IVS) which is constructed using the European Values Study (EVS) Trend File and the World Values Survey (WVS) Trend File for the period 2010 to 2021 (Haerpfer et al., 2020; EVS 2021), which are both repeated cross-sectional longitudinal datasets. We ensure that the data for native, first-generation, and second-generation immigrants was available uniformly across these years. The WVS and EVS are global research projects initiated by renowned scholars who study the socio-demographic characteristics of individuals, as well as their beliefs and attitudes over time. The surveys have been conducted since 1981 (Haerpfer et al., 2020; EVS 2021). Further information on the construction of the IVS dataset and other details can be found in the supplementary files in Supplementary Section S.1 towards the end of the Supplementary Tables document.

Below we briefly discuss our dependent and some key independent variables from the IVS dataset used in our analysis. Table S.1 in the supplementary files provides the detailed variable description.

4.1 Dependent variables

4.1.1 Self-employment

Individuals who identify as being self-employed when asked about their occupation status in the IVS dataset. This variable was selected as our main dependent variable to capture self-employment in our study.

4.1.2 Self-employment with employees and self-employment without employees

We include ‘self-employment with employees’ and ‘self-employment without employees’ as two additional dependent variables. Literature pertaining to self-employment in certain contexts such as freelancing states that not all individuals who are self-employed prefer hiring employees (Bögenhold and Klinglmair 2016). Employers without employees may be owners of very small businesses. Some of the previous studies have noted a U-shaped relationship between both types of self-employment with age (Simoes et al. 2016). Studies also note a positive association between formal and higher education with being a job creator (self-employed with employees), implying that individuals with more knowledge on efficiently managing and organizing businesses tend to be self-employed with employees (Dvoulety 2018). While looking at migration history and types of self-employment, studies suggest that immigrants are less likely to be job creators (Dvoulety 2018), that is, to be ‘self-employed with employees’. This could be because most immigrants may be working out of necessity, and therefore, may be unable to hire others while trying to fulfill their own needs. Further, immigrants may be more risk-averse to invest in large businesses. Therefore, it is important to assess the determinants of self-employment with employees and without employees, which may also help to attain relevant policy insights. Figure 1 shows the percentage of natives, and first- and second-generation immigrants within the different self-employment categories in our sample.

Fig. 1
figure 1

Self-employment categories for natives, first-generation and second-generation immigrants

Note: Self-employment expressed in percentages

Data limitation: No observations were available for second-generation immigrants for ‘self-employed with and self-employed without employees’

Full size image

4.1.3 Immigrants

Those individuals born outside a country are termed as first-generation immigrants in our study, while individuals born in a country are termed as natives. Additionally, we use the term second-generation immigrants as those who are natives themselves but either have an immigrant mother or father or both of their parents are immigrants, following Dinesen and Hooghe (2010).

4.2 Independent variables

4.2.1 Personality traits

For the purpose of testing personality traits, we selected a number of variables from the IVS dataset. The variables consisted of the importance of certain values to be instilled in children, attitudes towards social issues; preference for income equality versus individual effort, preference for government in comparison to private ownership of business, responsibility for welfare by state versus individual themselves, opinions towards rights of unemployed individuals to refuse jobs, and stances towards competition.

Individuals were asked to specify which values they considered important for their children. Values such as ‘independence’, ‘obedience’, and ‘imagination’ indicated traits like innovation. ‘Thrift’ was included to indicate an internal locus of control as savings generated from thrifting can be used for investments. ‘Hard work’ and ‘determination’ were selected to capture the need for achievement. The preceding variables were coded as dummies. Individuals who worked harder were expected to be more driven to pursue success in their entrepreneurial ventures. Alternatively, if individuals perceived success to be led by luck, they tended to be lazy and carefree (Tabellini 2010). Additionally, a few variables were selected to measure respondents’ attitudes toward social issues. For all these variables, respondents were asked to answer on a scale ranging from 1 to 10. Details of each variable are indicated in the supplementary Table S.1. Amongst these variables, only the ‘incentive for individual effort’ indicates the need for achievement whereas ‘government ownership of businesses’, ‘government to take more responsibility’, ‘unemployed have a right to refuse job’, and ‘competition is harmful’ refer to risk-taking attitudes among individuals. The preceding discussion provides the motivation for the selection of these independent variables. Higher values of all these variables indicate the risk-averse nature of respondents. The study also includes variables ‘job priority given to nationals’ and ‘immigrants being a strain on the welfare system’ to measure the attitudes of natives towards immigrants.

4.2.2 Individual characteristics

Our study uses a total of eleven control variables that may potentially affect our results. Age and age-squared variables reflect the age of the participant and its quadratic term. A few studies pertaining to entrepreneurship observe an inverted U-shaped association between age and self-employment, while some others find mixed evidence for this association (Simoes et al. 2016). Our study includes gender as a binary variable, which is coded as 1 if the respondent is a female and 0 otherwise. Self-employment amongst females is found to be lower owing to factors like the risk-averse nature of females and a greater burden of household and childcare responsibilities (Simoes et al. 2016). We also controlled for the education of a respondent by including three dummies for lower and middle-level education, where higher education is the base category. Respondents were said to have a lower-level education if they had early childhood education, primary education, or lower secondary education. Further, middle-level education denotes that that respondent possessed upper secondary education or post-secondary non-tertiary education. Finally, a respondent with higher-level education had either one of the following: short-cycle tertiary education, Bachelor, Master, Doctoral, or equivalent (Haerpfer et al., 2020).

Further, household size and its quadratic term are also included. We also examine the marital status of respondents and whether they have children and its association with self-employment. Another possible determinant of self-employment is the nature of the employment of the spouse. The human capital theory suggests that spouses may benefit from the sharing of knowledge, experiences, resources, and social contacts, thus positively influencing decisions to become self-employed (Parker 2008). Lastly, we include happiness which measures the emotional quality of life of the respondents (see Table S.1).

4.2.3 Social capital

We also include immigrant networks as independent variables. The first-generation immigrant network is the percentage of all the foreign-born respondents in each country following Brieger and Gielnik (2021). Former studies exploring the relationship between social capital and self-employment find mixed evidence (Doh and Zolnik 2011). We essentially use immigrant networks as a proxy for social capital. Immigrants tend to rely on migrant networks for resources, specifically in the earlier stages of starting a business (Sommer and Gamper 2018). Immigrant networks are believed to affect self-employment through shared customs of the migrants from the same country. Limited knowledge of labor markets in the host country may affect immigrant integration negatively. Thus, a wide immigrant network could help them to assimilate through similar languages, traditions, and social norms (Barnard and Pendock 2013). However, a few other studies argue that a migrant network may potentially discourage self-employment by placing obligations on individuals and preventing them from reaching out to broader business networks that are not part of the migrant network (Sommer and Gamper 2018). Thus, while a higher social capital may indicate higher trust and sharing of knowledge and ideas, it may also be linked to the “exclusion of outsiders, … restrictions on individual freedoms…” (Portes 1998). Hence the overall evidence is mixed for the relationship between immigrant networks and self-employment.

Additionally, we also use being a part of an organization as a proxy for social capital. The variable member of an organization is “1” when the respondent is an active or inactive member of any organization (as shown in the WVS questionnaire), and “0” in the condition that they are not. Being associated with an organization has been shown to have a positive influence on the decision to become self-employed (Jayawarna et al. 2015).

4.2.4 OECD

We include the Organization for Economic Co-operation and Development (OECD) in our analysis as a dummy variable. Pietrobelli et al. (2004) found that OECD countries have significantly lower self-employment rates, where one possible reason cited was the higher tax rates being imposed on self-employed individuals. Further, it was found that some economies in the OECD have made it more difficult to do business recently. For example, corporate income taxes grew in Latvia for the year 2018 (World Bank 2020). Figure S.2 in the supplementary files presents the number of countries included within the OECD.

4.3 Method

Due to the binary nature of our dependent variable, the theoretical measurement model used in this study is a discrete choice model (the Probit model).

The model used in our analysis is as follows

$$\:{\text{P}\text{r}(Y}_{i}\:\left|\:\right)={\Phi\:}({\alpha\:}\:\text{x}\:\:{Personality\:traits}_{i}+\:\beta\:X)\:\:\:$$
(1)

Where i is the individual, the dependent variable \(\:{Y}_{i}\) is a binary variable self-employment which equals 1 when an individual is engaged in self-employment, and 0 otherwise. \(\:{Personality\:traits}_{i}\) includes the personality traits in our model. \(\:X\:\) represents the other independent variables. As the equation presented above is not linear, the magnitude cannot be interpreted and only the sign can be examined. Therefore, we calculate the marginal effects for the main tables and report them in the supplementary tables.

Table 1 shows the results for self-employment amongst natives, first-generation and second-generation immigrants. Subsequently, Table 2 looks at determinants of self-employment with and without employees for natives and first-generation immigrants. Table S.2 in supplementary files includes the descriptive statistics for the full sample, for natives and immigrants. Tables S.3 and S.4 show the marginal effects for Tables 1 and 2, respectively. Table S.5 includes self-employment amongst native and first-generation immigrants for only the sample of OECD countries and Table S.6 shows its marginal effects. Tables S.3 to S.6 thus show the magnitude of the effects of the independent variables. Table S.7 displays self-employment amongst native and first-generation immigrants and includes the variable “member of an organization” for only the sample of OECD countries.

Table 1 Self-employment for natives, first-generation and second-generation immigrants
Full size table

5 Results

Using Probit modeling, we perform a number of regressions. Table 1 presents results for natives, first-generation, and second-generation immigrants separately. We include a dummy variable for OECD countries. Below we briefly shed light on the significant variables in our models and the discussion of these results is presented in the subsequent section.

In Table 1, in models (columns) 1, 3, and 5, the association of some of the basic individual-level characteristics with self-employment for natives, first-generation and second-generation immigrants is presented. While in models 2, 4, and 6, some additional variables indicating personality traits are added. Models 1 to 6 show that compared to males, females are less likely to be self-employed and the results are statistically significant for natives, first-generation and second-generation immigrants. The association between age and self-employment on the other hand is an inverted U-shaped relationship. This implies that the likelihood of self-employment is low at a very young age, then increases with age, and then starts to decline in the later stages of life. For spousal self-employment, natives, first-generation and second-generation immigrants are more likely to be self-employed if their spouse is also self-employed. Household size is positively linked with self-employment for natives but the sign reverses when considering the square of household size, indicating an inverse relationship. Lastly, married natives and first-generation immigrants are less likely to be self-employed.

One of the important correlates of self-employment is education, which is an indicator of human capital. Natives with lower and middle-level education in comparison to higher education are more likely to be self-employed. While first-generation and second-generation immigrants with a middle-level education are less likely to be self-employed than those with a higher education. Another important difference in native and first-generation immigrant self-employment is the presence of children. While the former are more likely to be self-employed if they have children, the opposite is true for first-generation immigrants.

Examining models 2, 4, and 6, which include happiness, we find that it is positively associated with self-employment among natives and first-generation immigrants but is insignificant for second-generation immigrants. With regards to important qualities to teach children, imagination is positively associated with self-employment for natives. Similarly, obedience is also negatively associated with self-employment amongst natives. Alternatively, independence is negatively associated with self-employment in second-generation immigrants. For attitudes toward social issues, incentive for individual effort is positively associated with self-employment for natives and first-generation immigrants. This means that natives and first-generation immigrants who believe that incentive for individual efforts should be promoted instead of income equality, are more prone to being self-employed. Besides, natives and immigrants who believe that the unemployed have a right to refuse jobs reflect a negative association with self-employment. Similarly, natives who prefer government ownership of business are less likely to be self-employed. To study attitudes towards immigrants, the variable ‘immigrants are a strain on the welfare system’ was included, which is negative for natives implying natives who agree with the statement are less likely to be self-employed. Lastly, belonging to OECD countries is negatively associated with self-employment for natives and second-generation immigrants.

To examine the determinants of being self-employed with and without employees, we present Table 2. Only first-generation immigrants were considered since there were insufficient observations for second-generation immigrants for these variables. Columns 1 to 3 show that individual-level variables such as being a female are negatively associated with self-employment with and without employees (for both natives and first-generation immigrants) and the effect is highly statistically significant. Age and self-employment without employees show an inverse U-shaped relationship for natives. Possessing middle-level education in comparison to higher education is positively associated with self-employment with and without employees for natives and the effect is statistically significant, while it is insignificant for first-generation immigrants. Similarly, natives and immigrants are more likely to be self-employed if their spouses are also self-employed but the magnitude is highest for first-generation immigrants who are more likely to be self-employed with employees. In terms of the important qualities to teach children, native parents who value independence are less likely to be self-employed without employees, whereas those who value imagination are more likely to be self-employed without employees. Similarly, obedience is negatively associated with self-employment without employees for natives.

Table 2 Self-employment categories: self-employment with and without employees for natives and first-generation immigrants
Full size table

On the other hand, first-generation immigrants who value hard work, imagination, and thrift are all less likely to be self-employed with employees. A different case can be observed for self-employment without employees for first-generation immigrants, but the results are insignificant and thus inconclusive. In terms of attitudes toward social issues, the incentive for individual effort shares a positive association with self-employment for both with and without employees for natives. However, for first-generation immigrants, the incentive for individual effort is statistically significant and positively associated with only self-employment without employees. Other preferences such as government ownership of business and that the unemployed have a right to refuse jobs are all less likely to be self-employed with and without employees for natives and first-generation immigrants. Natives who find competition harmful (implying a risk-averse nature) are also less likely to be self-employed with employees. Finally, natives who believe job priority should be given to nationals and that immigrants are a strain on the welfare system are less likely to be self-employed without employees. Residing within an OECD country is negatively associated with self-employment with and without employees in the case of natives and this result is statistically significant.

Table S.5 in the supplementary files shows the results for self-employment for OECD countries with the immigrant network percentage as an independent variable to examine the effect of immigrant networks on the likelihood of self-employment. We include results for native and first-generation immigrants only. The results show that a first-generation immigrant network reduces the likelihood of self-employment for first-generation immigrants and the coefficient is statistically significant at the 5% level in OECD countries. Interestingly, natives living in OECD countries with a high first-generation migrant network percentage are also less likely to be self-employed.

6 Discussion of results

6.1 Brief discussion of main results

Our study examined the multidimensional determinants of self-employment for first-generation and second-generation immigrants and natives while looking at socio-demographic variables, personality traits, and immigrant networks as the possible determinants.

In terms of the proposed hypotheses, we looked at various personality traits. Personality trait variables indicating a need for achievement included important values to teach children like hard work and determination and a preference towards incentive for individual effort by the respondent. Incentive for individual effort was positive and significantly linked with self-employment. This confirms our first hypothesis, suggesting that indicators of the need for achievement are expected to be positively related to the likelihood of being self-employed. Thrift as an important quality to teach children indicated an internal locus of control. While the variable was insignificant elsewhere, it shared a statistically significant negative relationship with self-employment with employees for first-generation immigrants. This result contradicts our second hypothesis which suggests indicators of the internal locus of control are expected to be positively related to the likelihood of being self-employed. This could be driven by a myriad of factors such as despite having a high internal locus of control, the high financial risk involved in self-employment with employees for first-generation immigrants drives them against self-employment. Therefore, in theoretical terms, immigrants who have a high internal locus of control are not as likely to be self-employed. This is in contradiction to the theory proposed by Rotter (1966). Variables exhibiting risk-taking propensity values like independence were negatively associated with self-employment without employees for natives, whereas imagination was positively associated with it. Kihlstrom and Laffont (1979) show that individuals with a risk-averse personality were more likely to become employees while those possessing higher risk-tolerance were more likely to become self-employed. Therefore, our results are in opposition to the theory proposed by Kihlstrom and Laffont (1979). Finally, variables including government ownership of businesses, the unemployed having a right to refuse jobs, and competition is harmful (which indicates lower risk-taking propensity) also shared a negative relationship with self-employment. This confirms our third hypothesis that the indicators of innovation and risk-taking are expected to be positively related to the likelihood of being self-employed.

For variables indicating differences in the likelihood of self-employment for immigrants and natives, an important one was education. While natives who have lower or middle-level education possessed a greater chance of self-employment than those with higher education, the reverse held true for first-generation immigrants. Similarly, in Table 2 natives were more likely to be self-employed if they possessed a middle-level education in comparison to higher education. Past literature suggests that individuals with a higher education have higher chances of being employed in good occupations that pay exceptional wages which may be the case for natives (Simoes et al. 2016). A different case is observed for immigrants where due to a shift towards more technical businesses, immigrants use their higher education and experience to recognize and explore self-employment opportunities (Hunt 2011). Therefore, immigrants possessing higher education have greater chances of being self-employed. This confirms our fourth hypothesis that natives with lower education in comparison to higher education are less likely to be self-employed, while immigrants with higher education are more likely to be self-employed.

However, studies such as those by Amac et al. (2019) have shown that migrant and refugee students in higher education face some obstacles such as language barriers, guidance issues, information asymmetries in navigating education and career pathways, missing documentation, financial problems, and inadequate academic support from their parents. Host countries need to address such challenges so that immigrants and refugees can obtain higher education and benefit from opportunities in the labor market or become self-employed. It is also essential that host countries have educational policies that allow immigrant students to learn inclusively and freely express their views regardless of race or religion at all levels of education to enable them to effectively assimilate into societies (Osler 2017).

6.2 Discussion of other key results

Our results show that individual characteristics of self-employment including age, gender, education, and spousal employment are significant determinants of self-employment. The findings reveal that older individuals are more likely to be self-employed and the likelihood of self-employment begins to decline for individuals close to retirement age, exhibiting an inverted U-shaped relationship. Moreover, all females possess a smaller probability of self-employment despite their immigrant status. This may be attributed to females having a larger share of household and childcare errands (Simoes et al. 2016). Further, spousal self-employment contributes to the provision of capital, resources, and networks which increases the likelihood of becoming self-employed. The same can be said about household size since a greater number of individuals in the household would likely contribute to the human capital needed for businesses. An important determinant accounting for native-immigrant differences in self-employment is the presence of children at home, where natives with children are more likely to be self-employed and the opposite is true for first-generation immigrants.

Immigrant networks may be viewed as a form of social capital, or as social relations which may provide access to employment or higher wages. Douglas Massey first suggested this perspective using social capital theory (Arango 2016). It has also been suggested that such a network might also include other intermediary institutions such as smuggling gangs or humanitarian organizations, however, such institutions might not fall under the ambit of social capital (Arango 2016). In our study, we find that first-generation immigrant networks in a country reduce the likelihood of self-employment for natives, as well as first-generation immigrants, suggesting that higher social capital in the form of immigrant networks reduces self-employment amongst immigrants. Andersson (2021) also noted a similar relationship where immigrant networks by themselves decreased the likelihood of self-employment. We have discussed the mixed evidence on the relationship between immigrant networks and self-employment earlier in our study. We have elaborated on further research areas regarding immigrant networks and self-employment in the Conclusion of our study.

Finally, our study also reports that belonging to an OECD country decreases the likelihood of being self-employed for natives as well as immigrants for both generations. This has been observed in previous studies (Pietrobelli et al. 2004) and may be linked to a theory proposed by Kuznets in 1996 that suggested a negative link between economic growth and self-employment (Deane 1967). Our study also illustrates that natives are less likely to be self-employed with employees in OECD countries.

7 Concluding remarks

A rise in immigration rates has encouraged research concerning the assimilation of immigrants and the factors contributing to it. Self-employment may be a positive contributor to immigrant assimilation in the face of multiple barriers like labor market discrimination and language difficulties amongst others. Our study contributed to the literature by examining factors influencing the self-employment of natives, first-generation and second-generation immigrants, while utilizing the unique and comprehensive Integrated Values Survey (IVS) dataset from 2010 to 2021 (covering 115 countries) for the first time for such analysis, to the best of our knowledge. Individual-level characteristics, as well as personality traits, and first-generation immigrant networks, were used as variables to understand self-employment amongst natives and immigrants from both first and second generations.

Our empirical analysis demonstrated that for individual-level factors, males were more likely to be self-employed than females, age had an inverse U-shaped relationship with self-employment, and higher education was positively related to the self-employment of immigrants from both generations, whereas the same did not hold for natives. Women had a lower likelihood of self-employment than men for all three groups, that is, natives, first-generation and second-generation immigrants. Furthermore, personality traits like achievement, motivation, and risk-taking propensity were positively associated with self-employment among natives. Moreover, the first-generation immigrant network reduced the likelihood of self-employment for first-generation immigrants in OECD countries. Thus our study confirmed the view that the influence of social capital in the form of immigrant networks, while mostly seen as positive in some previous studies in the context of finding employment and providing contacts (Doh and Zolnik 2011; Aguilera and Massey 2003), could be more multidimensional than previously observed and requires further research to be conducted to explore some of the drawbacks of networks as well as the policy pathways via which immigrant networks could be more helpful for immigrants in terms of knowledge-sharing.

Our study has highlighted the salient determinants of self-employment. There are a few limitations of our study. While several variables were included for personality traits, only one variable was included for internal locus of control due to limited data observations. Further, it must be noted that our study looks at the influence of immigrant networks on self-employment decisions without looking at geographically knitted networks. The latter may provide more details on the mechanism of the effect of immigrant networks on self-employment but could not be included in our study due to limitations posed by our dataset. Lastly, it is essential to acknowledge that economic crises and the resilience of countries to such crises, as well as advanced infrastructure facilities may also affect self-employment decisions, but we could not account for these factors due to the unavailability of such data in our sample. Moreover, we recommend that future scholars study the role of psychological aspects and intergenerational transmissions for self-employment decisions as well as the role of specific entrepreneurship skills, education, and training and the sub-sectors selected for self-employment if such detailed information becomes available in the future for the IVS dataset. With regards to the immigrant network, scholars may find it interesting to look at immigrant networks within similar ethnicities and their effect on self-employment. In some instances, there may be regulatory constraints affecting the potential of immigrants to become self-employed. Therefore, nations that intend to encourage self-employment should review such regulatory barriers. Additionally, it is important to ensure that immigrants and natives have the same access to credit, as well as the provision of secure property rights (UNHCR 2006). Future research may explore the effect of such constraints on immigrant self-employment.

Some salient policy insights emerge from our study. Since females have lower self-employment rates consistently, policymakers could focus on improving self-employment amongst them, especially by assisting in relieving their childcare responsibilities via state-sponsored childcare programs. For immigrants, there could be skills-based training programs organized on innovation and creativity to support less educated immigrants in starting new businesses since immigrants with greater human capital are more likely to be self-employed as well as policies to enable ease of access to financial capital. Also, since natives with a lower education are more likely to be self-employed, training must be conducted to prepare them against rising competition in the labor market and help them effectively transition towards more innovative businesses.