Out with the Old or Out with the New: The Uncertain Role of Entrepreneurship in Developing Countries in the Post-COVID-19 Context

The promotion of policies and initiatives that support job creation through entrepreneurship in developing countries is at the heart of the 2030 sustainable development goals (SDG) agenda, and, in particular, SDG 8 and 9. Yet, entrepreneurship remains an abstract concept difficult to define and com-prehend, and the literature on entrepreneurship has been developed mostly in a developed country context. In the current global context, dominated by the dramatic impact of the COVID-19 pandemic, the relevant policies to be adopted by developing countries to face the socioeconomic consequences of the pandemic must be identified and analyzed. One of the main concerns of these policies is unemployment. The simultaneous supply and demand shocks caused by the pandemic have raised calls for an unprecedented level of public intervention in both developing and developed countries to prevent massive unemployment and chains of bankruptcies. The objective of this paper is to study the role of entrepreneurship in job preservation or job creation at the macroeconomic level. After reviewing different taxonomies of entrepreneurship and discussing motivations of entrepreneurs in a developing country context, the fundamental relationship between entrepreneurship, job creation, and macroeconomic performance in developing countries in both the pre- and post-COVID-19 contexts is analyzed. Using panel data for a sample of 24 countries members of the Organization of Islamic Cooperation (OIC) across the period 2009-2018, the paper demonstrates a U-shaped relationship between entrepreneurship and the level of economic development of nations, which translates into a negative correlation between economic growth and entrepreneurial dynamism for developing countries with per capita income below a certain limit. This proves that different phases of the economic cycle require different policies and promotion of entrepreneurship should therefore be adjusted to the economic conditions of each country.


Abstract
The promotion of policies and initiatives that support job creation through entrepreneurship in developing countries is at the heart of the 2030 sustainable development goals (SDG) agenda, and, in particular, SDG 8 and 9. Yet, entrepreneurship remains an abstract concept difficult to define and comprehend, and the literature on entrepreneurship has been developed mostly in a developed country context. In the current global context, dominated by the dramatic impact of the COVID-19 pandemic, the relevant policies to be adopted by developing countries to face the socioeconomic consequences of the pandemic must be identified and analyzed. One of the main concerns of these policies is unemployment. The simultaneous supply and demand shocks caused by the pandemic have raised calls for an unprecedented level of public intervention in both developing and developed countries to prevent massive unemployment and chains of bankruptcies. The objective of this paper is to study the role of entrepreneurship in job preservation or job creation at the macroeconomic level. After reviewing different taxonomies of entrepreneurship and discussing motivations of entrepreneurs in a developing country context, the fundamental relationship between entrepreneurship, job creation, and macroeconomic performance in developing countries in both the pre-and post-COVID-19 contexts is analyzed. Using panel data for a sample of 24 countries members of the Organization of Islamic Cooperation (OIC) across the period 2009-2018, the paper demonstrates a U-shaped relationship between entrepreneurship and the level of economic development of nations, which translates into a negative correlation between economic

Introduction
The promotion of policies and initiatives that support job creation through entrepreneurship in developing countries is at the heart of the 2030 Sustainable Development Goals (SDGs) agenda, and, in particular, SDGs 8 and 9. Yet, entrepreneurship remains an abstract concept difficult to define and comprehend. Entrepreneurship is frequently confused with self-employment, an important source of employment in developing nations. Perry et al. (2007) relate that self-employment may constitute up to 30% of the total employment in Latin America, for example. In a more comprehensive study across 74 developing nations, Gindling and Newhouse (2014) relate statistics for mainly low and middle-income groupings averaging 51.6% and 33.5%, respectively, according to World Bank definitions. Owner-managers of small enterprises are therefore a critical category of entrepreneurs that must be considered in any attempt to provide and preserve jobs for all those in line with the SDG targets. However, the individual self-employed is not the only category of entrepreneurs in the economy. As Hagedoorn (1996) puts it, entrepreneurship can be broadly classified into two streams for research: one that looks on entrepreneurship at the level of individuals and the other that looks at corporate entrepreneurship and "intrapreneurship" especially at the level of large firms. Both categories are affected at times of crisis and this paper is specifically concerned about the drivers of successes and failures of entrepreneurship policies at a macroeconomic level in developing countries, with a focus on the post-COVID-19 context. COVID-19 has unleashed an unprecedented economic crisis expected to affect the developing prospects of all nations across the globe for years to come. The "swift and massive shock of the coronavirus pandemic" since its advent in early 2020 is expected to be the harbinger of the deepest recession since the Second World War (Felsenthal, 2020). Entrepreneurial activities in all sectors have been affected by the simultaneous supply and demand shocks caused by the pandemic, and the World Economic Forum (WEF) estimates, for example, that the pandemic has forced more than 70% of start-ups to terminate full-time employee contracts. Yet, the degree and extent of the impacts of the recession on entrepreneurial activities are not well-understood. One of the reasons for that gap is the lack of methodological tools to link the aggregate macroeconomic ac- This has an important consequence when looking at the macroeconomic impacts of the COVID-19 recession on both entrepreneurial activity and employment in developing nations, and this is discussed in sections 6 and 7 with a brief theoretical review of the main drivers of the relationship between wages and employment levels. The policy options for developing nations with limited fiscal space and growing unemployment levels in the context of such a major recession are very limited, calling for great selectivity and a detailed understanding of the most efficient mechanisms to achieve job preservation or job creation targets despite challenges. The conclusion will hence propose some areas for further research and some candidate policy approaches worth considering given the research results presented in the previous sections.

The Drivers of Entrepreneurship
Despite the consensus on the role of entrepreneurship in economic growth, the interactions between entrepreneurs and their environment are very complex and difficult to comprehend. These interactions reflect the bidirectional relationship between entrepreneurs and their environment and the way they are also affected by several socio-economic factors in the economies where they operate. Over the years, economic theory has gradually incorporated additional factors in analyzing entrepreneurship as a driver of economic activity. Walia and Chetty (2020) differentiate between three main types of theories of entrepreneurship, namely the classical theories, the neo-classical theories, and the Austrian market process as illustrated in Figure 1. While the classical and neo-classical approaches focus on the actions undertaken by entrepreneurs for the production and distribution of goods and services, the Austrian approach is based on the knowledge that entrepreneurs possess about the economy. Casson and Wadeson (2007) consider four approaches to define the entrepreneur based on his function, role, personal characteristics, and behavior. Function has to do with the action undertaken through innovation and risk taking, while role is related to the action as an investor owning capital and exploiting production factors. Personal characteristics and behavior are more related to attitudes and decision making. In this context, the entrepreneurial function determines his impact on the macroeconomic performance.  As an investment activity, entrepreneurship is based on the knowledge of the entrepreneur and on his ability to take a call regarding existing opportunities to overcome the challenges resulting from his decision. At the same time, opportunities are the results of changes in several socio-economic conditions influencing the demand and supply conditions. For Casson and Wadeson (2007), the fact that the environment is volatile implies that opportunities are never exhausted, and what is limited is just the stock of projects available to entrepreneurs. In studying the determinants of entrepreneurship, this leads to distinction between the internal factors related to the entrepreneur's skills and capabilities and the external factors related to his environment and the way the environment affects his decisions.
The question of how the environment affects entrepreneurship has been a subject of divergence between economists and has triggered many empirical studies to identify the critical factors for success and failure in different contexts. del Olmo-García et al. (2020) distinguish between economic environment and institutional environment and their contribution to the failure of self-employed entrepreneurs in the eurozone. Their results demonstrate that in addition to economic factors, the culture and perceptions of society about entrepreneurship are significant factors in reducing entrepreneurship failure. Furthermore, their study shows that the quality of formal institutions (including protection of property rights, the quality of the legal system etc.) plays a key role in reducing the failure of entrepreneurship. Interestingly, Devece et al. (2016) show that in Spain, while innovation remains the key success factor, opportunity-driven entrepreneurship tends to perform much better than necessity-driven entrepreneurship in periods of recession.
Overall, studying entrepreneurship is complex in the sense that it implies analyzing the economic, social, and institutional factors responsible for the success and failure of any entrepreneurial activity. Recently, an increasing number of studies have been interested in the level of entrepreneurship in different economic phases. In that sense, in addition to the multidisciplinary aspect of entrepreneurship, the results of any study are conditioned by the context of the study. For example, while several studies have been able to assess the situation in the context of developed countries (Maritz et al., 2020;Galindo-Martín et al., 2021;Devece et al., 2016;del Olmo-García et al., 2020), studies in the context of developing countries have been very limited until now (Mendoza et al., 2021;Omri, 2020). Lastly, the nature of entrepreneurship is very important to assess the sustainability of its impact on economic growth. In fact, opportunity entrepreneurship is driven more by innovation and the opportunities it creates in the economy. Hence, this kind of entrepreneurship, which is more predominant in developed countries, can leave a lasting impact on economic growth by creating added value and more jobs, as well as triggering other innovations through competition. However, necessity entrepreneurship is in general driven by the lack of alternatives and by the low level of employment created by the economic activity. In this context, this kind of entrepreneurship can be terminated as soon as some opportunities are created and become available to the entrepreneurs.
In addition to the distinction between developed and developing countries, some recent researches have studied the dynamics of entrepreneurship in periods of crisis (Maritz et al., 2020;Galindo-Martín et al., 2021;Devece et al., 2016). Similarly, the current crisis created by the COVID-19 pandemic has triggered some questions about the potential impacts on entrepreneurship, as well as the potential opportunities this crisis has created. For example, according to the World Economic Forum (WEF), the pandemic has forced more than 70% of start-ups to terminate full-time employee contracts. At the same time, the WEF considers that the pandemic has also created new opportunities through the creation of new needs and the opening of new markets related to the attempts to cope with the challenges posed by the new modes of life.

Relationship between National Income and Entrepreneurship Levels
Since 2004 has partially addressed some of the caveats of the previous GCI framework, which was deficient at several levels, including issues of methodology, consistent bias, qualitative vagueness, incorrectness, or redundancy (Lall, 2001), but methodologically, the framework of the GCI still suffers from multiple methodological issues including, for example, obvious issues of endogeneity and causality (Workie and Hekelová, 2016). Yet, the GCI has now established itself as a global indicator of the ability of countries to provide prosperity to their citizens and is constantly adopted by policymakers and development institutions to assess the successful macroeconomic performance of nations. As a side effect to this convergence of the global discourse on development prospects of nations, the paradigm shift from managerial to entrepreneurial capitalism, which once characterized the US economy (Baumol et al., 2007;Baumol, 2009), has now extended to how "good" or "bad capitalism" can be defined for all nations (Baumol, 2009).
The economic literature on the relationship between entrepreneurship, economic development, and growth is expansive. As Hagedoorn (1996) puts it, this literature can be broadly classified into two streams of research: one looking at entrepreneurship at the level of individuals and the other looking at corporate entrepreneurship and "intrapreneurship" especially at the level of large firms. Hagedoorn (1996) also explains how this dichotomy is created in reference to early and late works of Schumpeter on innovation. Quadrini (2009)

Stylized Facts on Entrepreneurial Activity and Outcome Levels in Developing Nations
Even in the absence of relevant quantitative models and theoretical frameworks, several stylized facts can be derived from data on developing nations and the levels of self-employment or entrepreneurial activity. First of all, self-employment is an important source of employment in developing nations. Perry et al. (2007) relate that it may constitute up to 30% of total employment in Latin America, for example. In more comprehensive studies across 74 developing nations, Gindling and Newhouse (2014) relate statistics for mainly low and middle-income groupings averaging 51.6% and 33.5%, respectively, according to World Bank definitions (see Table 1 below).
The fact that one out of two workers or one out of three workers, respectively, in low income or lower middle-income countries is self-employed is in itself a major finding. Gindling and Newhouse (2014) analyze the type and quality of jobs that these self-employed workers occupy and their relative success. For example, the proportion of self-employed workers is significantly higher in agricultural jobs, reaching 60% or 47%, in the sample countries of the study for low income or lower middle-income groups, respectively. More importantly, the authors are able to demonstrate a clear "order" in such countries between employment categories, and wage and salary employees are revealed to be better off in particular than own-account workers.
This leads to the second stylized fact regarding self-employment and entrepreneurial activity in developing nations: the heterogeneity of the trajectories of entrepreneurs with respect to their growth potential. Gindling and Newhouse (2014) refer to two main groups among the self-employed workers: one group of self-employed with "limited growth prospects who either are self-employed by necessity, due to the lack of wage employment opportunities, or have voluntarily chosen self-employment over wage employment" and, in contrast, a higher tier of self-employed consisting of "innovative, successful entrepreneurs with greater potential and ambition for growth". This echoes the necessity-motivated entrepreneurship (NME) and the opportunity-motivated entrepreneurship (OME) in Bennett and Nikolaev (2019). A key conclusion of Gindling and Newhouse (2014) along these lines is that only a small proportion of the entrepreneurs in the first group (NME) are successful. Using different definitions of success (job creation or per capita consumption in the household), less than a third of the entrepreneurs are considered successful in the sample low income and lower middle-income countries. In this paper, this will be referred to as the "entrepreneurship trap" whereby self-employed individuals driven by necessity are not able to actually grow, accumulate capital or exit poverty and deprivation.
R. Abdelkafi, S. Refas A third major fact relates to the massive exclusion of individual entrepreneurs from the formal economy in low income and lower middle-income countries. The mechanisms are described in Auriol (2013). Starting from the well-known fact that the shadow economy (or informal sector) has a significant size in low income and lower-income countries, representing for example as reported by (Enste and Schneider, 2003) up to 39% of the economy across a sample of 76 developing nations, Auriol (2013) explains how several administrative, economic or cultural factors prevent entrepreneurs from entering the formal economy, even when successful. This implies various adverse consequences among which two are of significance in the context of this study: entrepreneurs in the informal sectors are exposed to weakened property rights and heightened risks, and at the same time, due to the heavy social costs incurred on them by the forced mutual help they have to extend to family and relatives in replacement of social safety nets, there is a net loss in job and wealth creation due to the inefficient allocation of resources.
A fourth major fact, especially in the context of development policies aiming at fostering employment, is what Schoar (2009) calls the divide between subsistence and transformational entrepreneurship. After reminding of the renewed expectations in recent years that entrepreneurship will spur the development of poor nations, Schoar (2009) rightfully observes that "most policy makers as well as economic researchers treat entrepreneurs as a homogeneous group of actors that are uniformly affected by economic conditions or policy interventions" and in doing so they fall to keep in mind a significant characteristic of low income and lower-middle income countries: the very low prevalence of medium-sized companies. The evidence is scanty to show why a very large number of tiny firms are unable to graduate into medium-sized firms which in the context of developed nations are the primary employer and contributor to economic success, but the facts are harsh and clear: the size distribution of firms in low-income and lower middle-income countries is bimodal with many tiny firms and few large firms. In this context, the characteristics of transformational entrepreneurs and the levers to support them are not well-identified, and they are not rightfully stimulated to boost the economy. Another dimension of this problem refers to the largely reported issue of skill mismatch to support successful entrepreneurs in their growth stage as discussed for example in Lee (2013). The following Table 2 summarizes these four main facts reported in the literature about entrepreneurial activity and income levels.

Quantitative Model and Empirical Results
In this section, an empirical analysis is conducted to test the relationship be- Following Omri (2020), we used the new business density as an indicator of entrepreneurship, to estimate the following model with 1, , 24 i =  representing countries and 1, ,10 T =  for years included in the sample: where it Y represents the dependent variable, in our case entrepreneurship. it X is a K-dimensional vector of explanatory variables, β a (K × 1) vector of coefficients independent of i and t and it ε the error term varying over i and t. -Government effectiveness is included to reflect the role of institutions in developing entrepreneurship (Galindo-Martín et al., 2021;Omri, 2020). This variable captures perceptions of the quality of public services, the quality of the civil service, the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies 1 . Theoretical Economics Letters

ENTRE
LGDPPC LGDPPC In order to confer more homogeneity on our sample, we chose to test the model on the sample excluding rich GCC countries. Table 4  Another interesting way to look at the relationship between entrepreneurship and economic development is to study the causality between the two variables.
As mentioned earlier, one of the main complexities in entrepreneurship lies in the fact that this variable participates in economic growth, and at the same time it is also affected by several economic factors related to economic development.
Based on that, we tried to perceive the causal relationship between entrepreneurship and economic growth by applying a Granger causality test on the variables ENTRE and LGDPPC on both samples. The results are as shown in Table   5 below:  LGDPPC Note: The t value is bracketed, and *** and **indicate significance at the 1% and 5% levels, respectively. The results presented in Table 5 prove that in both samples, the causality is only from the GDP per capita to entrepreneurship. The causality from entrepreneurship to growth is insignificant. This could be explained by the quality of entrepreneurship created in developing countries. While this entrepreneurship can be created following an increase in economic growth, it fails to drive economic growth in these countries. As mentioned above, necessity-based entrepreneurship is dominating the economies of developing countries. Lack of opportunities and innovation in developing countries makes entrepreneurship less effective in developing countries in comparison to developed countries.
Before concluding, the next sections will now discuss some theoretical con-

What Are the Possible Impacts of Recession on Employment Levels and Wages?
In veloping nations. However, this is further heightened in factor-driven labor-rich economies, which naturally supply an even higher number of job seekers in the market. In this context, one may be tempted to simplistically describe the dilemma that decision makers are faced with: either jobs of current workers should be saved on a priority basis with the limited levers available for government intervention or the onus should be on supporting the economy to rapidly transform to be able to create jobs for both the job seekers entering the market and the current workers who will inevitably lose their jobs as a result of the crisis.
The two dynamics are simplistically described in the graphs below. In both graphs, equilibrium employment level in the economy is defined as the intersection of a price-setting curve, which depicts the aggregate demand for jobs from the economy at the production frontier, and the wage-setting curve, which is the occupational demand for wage-jobs by job seekers (against the available alternatives of unemployment or other activities). Since the first effect of the crisis is the immediate impact on the revenues of firms and their ability to pay salaries, the price-setting curve is immediately shifted downwards and the equilibrium employment level is shifted from wh to wl: a net amount of jobs (wh − wl) is immediately lost in the economy. The dashed area 1 in Figure 2 below represents the net income lost in the immediate term at an aggregate level by the working class due to these job losses.
The secondary impact of the recession subsequent to the crisis is to shift the wage-setting curve downwards, especially for low-pay workers. This effect is particularly marked in developing countries with weak social protection structures (Arranz et al., 2020). Interestingly, and as pictured in Figure 3 at an aggregate level, this could generate a positive employment outcome since the lower wages due to lower expectations from job seekers as well as the entry of new job seekers in the market may lead to a higher equilibrium employment level wg. In practice, achieving this new maximum employment level will pose a challenge for the economy requiring transformation of jobs and recovery measures. The net amount of jobs created (wg -wl) will possibly recover partly or fully the net amount of jobs (wh -wl) that were immediately lost. At the same time, the aggregate income impact is pictured by the net difference between the diamond-filled area 2, which is the additional aggregate income due to job creation, and the dashed area 3, which is the aggregate income lost due to salary reduction.  This analytical framework must however be complemented to account for the role of entrepreneurship and self-employment when managers or workers do not simply enter the job market as price-takers of the wages offered by the available opportunities but in fact create their own jobs out of necessity or opportunity (Bennett and Nikolaev, 2019). Three populations are especially concerned in the graphs above: 1) in the immediate term, the population which lost their jobs as a consequence of the recession (wh − wl) in case there is no relief policy to immediately preserve their jobs or provide them with temporary wage support (necessity-motivated entrepreneurs) as pictured by area 1 in Figure 2; 2) in the medium term, the job seekers at the right of the equilibrium level wg,, who do not occupy the jobs at the wages offered by the market, because it is below the minimum expected wage for the concerned jobs (necessity-motivated entrepreneurs); and 3) in the medium term or immediate term, any employed individual who has a higher expectation from an entrepreneurial venture than the equilibrium wage offered by the market (opportunity-motivated entrepreneurs). An analytical challenge posed by the uncertainties related to wage-demand functions of the entrepreneurs, who intrinsically take risks by becoming entrepreneurs but have high-reward expectations, is to reveal their wage (or profit) expectations. Consequently, among the three populations just described the quantity of necessity-motivated or opportunity-motivated entrepreneurs with higher expectations that equilibrium wages cannot be pictured in Figure 3. The next sections will discuss however how different trajectories can be expected for these three populations of entrepreneurs, and how this will in turn lead to differentiated aggregate outcome at the macroeconomic level.

The Difficult Timing of Policy Interventions
The question of the relevant policy interventions for preservation and creation of jobs in times of recession at a macroeconomic level has two main dimensions: an effectiveness dimension and a time dimension. Due to the specifics of the agricultural jobs in a developing country context, these policies should be ideally treated differentially for agricultural and non-agricultural jobs. However, for the sake of simplification, this paper will present the candidate measures for job preservation and creation along two timelines only: short-term relief and longer-term recovery measures. The proposals are derived from Loayza et al. (2009), who discuss the time dimension of the policy interventions of developing nations to limit the detrimental effects of the economic crisis caused by the COVID-19 pandemic. Table 6 summarizes the most relevant recommendations along two dimensions: labour supply and consumption or, in other words, policy interventions on the supply and demand side of the economy. The "swift and massive shock of the coronavirus pandemic" since its advent in early 2020 is expected to lead to the deepest recession since the Second World War (Felsenthal, 2020). This is partly because the simultaneous demand and supply shocks from the pandemic  The challenges pertaining to the timing of job preservation and creation policies, therefore, become extremely challenging in the context of this crisis for developing countries with limited fiscal space, weak financial markets, small fiscal multipliers, and massive demographic pressure.

Conclusion
This paper is primarily about the relationship between entrepreneurship, job creation or preservation, and macroeconomic performance in developing countries. This complex relationship, fundamental in the context of the sustainable development agenda, is even more important in the context of massive supply and demand shocks caused by the COVID-19 pandemic. The primary conclusion of this paper is that in developing countries, the relationship between entrepreneurship and economic growth is determined by the current stage of economic development: due to multiple challenges, the relationship is weak in the lower income economies while in higher income economies GDP growth translates successfully into more entrepreneurial activity.
Other findings of the study include the observed prevalence of necessity-motivated entrepreneurship in developing countries, which starkly contrasts with the corporate or individual entrepreneurial activity in high-income countries, predominantly opportunity-driven. In addition, the trajectories of entrepreneurs with respect to their growth potential are found to be very variable and a majority of entrepreneurs in developing countries context are constrained in their growth potential due to administrative, economic or cultural factors, which translates into lower aggregate economic performance at a macroeconomic level.
The observed exclusion of individual entrepreneurs from the formal economy in The empirical study demonstrates that a non-linear model including a quadratic term for the GDP per capita is most adapted to map the relationship between entrepreneurship and economic growth at a macro-economic level in a developing country context. The model is calibrated and tested using panel data for a sample of OIC member countries. The model successfully demonstrates the quadratic relationship and the sample also shows a unidirectional causality going from the GDP per capita to entrepreneurship. This demonstrates that the low-quality entrepreneurship created in developing countries, in most cases necessity driven, is hardly a strong pillar of economic growth or recovery post-recession. In the context of an immediate and significant impact of the COVID-19 crisis on employment in developing countries, the findings of this paper would question therefore a one-size-fits-all policy-making approach which leverages on entrepreneurship or innovation to sustain or create jobs for economies in recession. The dilemma between protecting the existing fragile jobs and adopting entrepreneurship development policies is very acute and the right policy approach highly depends on the stage of development of the country. In this context, a menu of candidate policy interventions most promising to address the employment challenges in developing countries is briefly discussed in the last section.
This paper can constitute a starting point for further research on the same subject and in particular on the relevant approaches to fight unemployment through targeted and multi-cycle entrepreneurship policies in a developing country context. One of the main limitations of this research is the absence of disaggregated information to characterize the demand functions at a micro-level of necessity-motivated or opportunity-motivated entrepreneurs in a pre-crisis or post-crisis context. Understanding how the expectations of entrepreneurs are affected by the aggregate unemployment levels in particular is critical to understand whether the appetite for entrepreneurship grows or reduces in a context of employment crisis for wage jobs. Further research efforts are also required to characterize with more depth the relationship between entrepreneurial activity and economic growth and reveal in particular the mechanism by which GDP growth effectively leads to more successful entrepreneurship for higher income countries.