1 Introduction

There is a large literature on the factors explaining new job creation at the micro-level. There is specific attention to the differential contribution of small and medium sized firms (SMEs) vis-à-vis large firms. Several micro-econometric works show that the former are responsible for the largest share of new job creation. The impact of industry dynamics has received less attention (Haltiwanger 2012); the relationship between job creation and destruction and firms’ entry and exit dynamics is similar to Schumpeter’s concept of creative destruction, and the evolutionary idea of the survival of the fittest. While newborn firms are more likely to create new jobs, their post-entry performance is also relevant since poorly performing firms may produce job losses due to firm contraction or failure.

This paper contributes to this literature by investigating the relationship between industry dynamics and job creation in Tunisia over the period 1996–2010, extending previous work (Rijkers et al. 2014) by focusing on the subcategory of newborn firms to investigate whether regulation and fiscal policies can induce better outcomes for small firms. Tunisia is a small open economy, located in North Africa, with a population of around 11 million. It was the first country in the MENA (Middle East and North Africa) region to experience political change after the uprisings in 2011, which in part were triggered by the high share of—especially skilled workers’—unemployment in the region. Previously, Tunisia appeared to be a stable country with a growing economy (GDP increased by approximately 5% in the period 2000–2010). However, in Tunisian and many other MENA countries, this growth was not inclusive, neither did it lead to much job creation (World Bank 2013).

This makes Tunisia an excellent context to explore the issue of firm dynamics and job creation in Developing Countries (DCs). On the one hand, the largest share of small newborn firms in Tunisia was a result of self-employment dynamics, making the Tunisian economy appropriate to explore the link between entry/exit and firm size in relation to job creation (Rijkers et al. 2014). On the other hand, the Tunisian government has been pursuing active industrial and labor policies to promote new firm formation and to help firms to grow, and to perform better in terms of job creation. A key element in this respect is a specific fiscal scheme for small firms, the régime forfaitaire, which allows small firms with particular requirements to pay their taxes as a lump-sum.

The paper makes two contributions to the literature. First, having data covering the whole population of firms provides a rare opportunity to investigate these subjects based not on a sample of firms but on the entire economy. Second, in contrast to previous studies, we exploit information on the fiscal scheme to which firms are assigned. This allows us to investigate the extent to which the entry and exit dynamics in Tunisia are related to the specific regulation and tax regime, regime forfaitaire, which is aimed at simplifying and reducing the tax imposition on small new born firms.

Our paper provides preliminary evidence on the impact of the government’s introduction of the regime forfaitaire, aimed at increasing job creation and the persistence of newborn enterprises by lowering tax levels and the simplifying the tax payment process. Although these results are exploratory, they have important implications for policy by highlighting the interplay between industrial and territorial policies in developing economies.

The paper is organized as follows: Sect. 2 reviews the literature; Sect. 3 presents the data and describes the database; and Sect. 4 presents the descriptive analysis. Section 5 discusses the econometric model and presents the findings and results. Section 6 concludes.

2 Literature review

New firm formation, industry dynamicsm and their role in the creation of new jobs are the subjects of major ongoing debate, and particularly in the context of developing countries (DCs) where the stability of political regimes depends heavily on economic performance, especially with respect to employment. The literature on these topics can be grouped into three main strands.

The first strand focuses on the determinants of entrepreneurship and firm creation. Several contributions examine the extent to which the decision to start a new firm is driven by individual motivations such as comparison between current income in the labor market and expected income from the creation of a new company, level of education, emotional intelligence, etc. Most newborn firms in DCs stem from self-employment and the profile of the entrepreneur is a key factor. Successful entry depends on individual features such as motivations, expected profit and revenue, previous experience, human capital, social capital, and macro-economic characteristics such as regulation, business cycle, local market dynamics, and availability of lending and capital. However, the role of location is also emphasized. Entrepreneurs are very sensitive to sector and location inertia. The initial composition of the economy can have an impact on entry/exit dynamics, and also on entrepreneurs’ quality (Vivarelli 1991; Arrighetti and Vivarelli 1999; Stam 2007). While this literature is useful for understanding the entry decision it provides no information on outcomes and fitness of newborn firms.

The second stream of literature investigates the economic and social impacts of small and newborn firms on the economy. Most studies suggest they play a prominent role in transforming the economy, fostering innovation and creating employment. They find that small and new firms are mainly responsible for new job creation (Beck et al. 2005). For example, in the case of the United States, during the period 1980–2011 (Decker et al. (2014), show that while the rate of new entry and exit of firms fell, and despite the fact that most new firms fail, a large proportion of net job creation was created by the minority of new firms that succeeded and grew. Conditional on their survival, these young firms account for a significant share of net job growth (Criscuolo et al. 2014). The picture varies from country to country, and with national levels of development. In rich countries entrepreneurship may drive economic growth but it has a negative impact in DCs (van Stel et al. 2005; Stam and van Stel 2011). However, although the evidence is scarce, it seem that in DCs newborn firms can play an important role in fostering innovation, upgrading industries and increasing efficiency (Goedhuys and Sleuwaegen 2010; Szirmai et al. 2011; Lorenz 2015). At the same time, they play an important social role by providing opportunities for millions of people to work in self-employment. In the case of Tunisia, Rijkers et al. (2014) find that the firm size distribution is skewed towards small firms because of stagnation among incumbents, and entrants starting small, typically as one-person firms. They found also that post-entry, small firms even if they survive, are the worst performers in terms of job creation. In addition, the association between productivity, profitability and job creation is weak suggesting similar weaknesses in the re-allocative process. The study by Bottini and Gasiorek (2009) shows the results of parametric and non-parametric analysis of employment growth rates, suggesting that while these are typically low there is considerable firm level heterogeneity whereby employment seems to grow more in exporting, small and young firms. Overall, while most of the literature points to a turbulence effect related to small and newborn firms, little attention is paid to their societal role in maintaining social cohesion.

The third literature strand focuses on the quality of entrepreneurs and small firms’ growth patterns. While small firms are responsible for the largest part of industry dynamics (entry and exit) in both developed countries and DCs, they face barriers to growth to develop into big firms. The economics literature suggests the co-existence of opportunity entrepreneurs and innovator entrepreneurs (Quatraro and Vivarelli 2015; Banerjee and Duflo 2007). It is also well known that new firms in DCs are mostly imitator firmswhich promote persistent turbulence and an unstable economic climate. Most DCs lack innovators; Schumpeterian innovator firms can transform the economy and increase productivity and efficiency. Innovative firms are supposed to grow more rapidly but also to die more rapidly if they do not achieve success. For example, compared to Europe, the United States traditionally has been more dynamic. In particular, successful firms in the United States grow more quickly than in Europe, and unsuccessful firms die more quickly in the United States than in Europe (Bravo-Biosca 2010; Criscuolo et al. 2014). The growth of small and newborn firms is an important issue for DCs, and empirical evidence in this area is scarce.

3 Data

This paper aims at investigating the job-creation effects of newborn firms in Tunisia. The analyses draw on two datasets. First, the Tunisian registry of firms, the Répertoire National des Entreprises (RNE), covering the period 1996–2010, which is collected by the Tunisian Institut National de la Statistique (INS). The RNE collects information from a host of constituent administrative databases including from the social security fund (Caisse Nationale de la Sécurité Sociale—CNSS) which is the source of employment data, as well as the Tunisian Customs, the Tunisian Ministry of Finance, and the Tunisian Investment Promotion Agency (l’Agence de Promotion de l’Industrie et de l’Innovation—APII). These sources provide data on all firms registered with the tax authorities (see INS (2013) for detailed information on the construction of the database).

Second, the Tunisian Statistical Business Register (TSBR) contains information on firm identification, economic activity, legal form, nationality, regime (onshore or offshore), age, employment, turnover, export turnover, profit, wage, and export and import values. It covers all registered private non-agricultural firms including self-employment units which account for the bulk of all enterprises, and covers the period 1996–2014. Note that information on turnover and benefits in TSBR is available only for enterprises submitting accounts during the period 2000–2014. To complete this information for the units without accounts, we integrated the responses to small enterprises surveys administered in 2007 and 2012 with the TSBR data. To identify firms that are registered but no longer economically active, referred to as “false active units”, we developed a statistical methodology using the results of economic surveys and activity signal variables (declaration of turnover, import, export, wages). For each enterprise in the SBR, we calculate a probability of “unit is active” (15% of one person units are false active units).

4 Descriptive analysis

It seems that very few studies focus on the social and economic impacts of firm dynamics on the labor market in the context of the MENA region.

Small firms in our analysis are firms with fewer than six employees (including the owner). Figure 1 shows the distribution of firms and employment by firm size. The prominence of small firms is clear; they account for 97% of the total number of firms in Tunisia, and are responsible for 42% of employment in the private sector (excluding agricultural jobs). The next biggest group contributes the most to employment (27%) and is represented by large firms which however, account for only 0.1% of the total population of firms.

Fig. 1
figure 1

Distribution of firms and employees, by firm size class, year 2012

Full size image

4.1 Entry and exit dynamics by firm size

Since 2004, entries and exits are confined mainly to small firms. Table 1 presents the entry and exit dynamics of Tunisian firms by size distribution. We observe that 97% of active firms are small enterprises. This share is higher if we look at the size distribution of newly founded firms where 99% of newly created firms are small firms. In terms of entry rates, i.e., the ratio between entry and active firms, the difference is around 11% for small firms, 2% for medium-sized firms, and 1.5 and 1% for large and very large firms, respectively. Small firms also have the biggest representation in net entry, where we see some stagnation in net creation among medium and large firms. Small firms are responsible for net firm creation.

Table 1 Number of entering, incumbent and exiting firms in Tunisia (yearly averages during the period 2000–2013)
Full size table

Figure 2 plots the time series for the entry dynamics (evolution of entry rates) in Tunisia (2000–2013) by firm size groups. The data are normalized with respect to the base year 2000. Figure 2 shows that small firms account for the largest share of entries in the economy, and their dynamics seem to be relatively unaffected by cyclical behaviors. The evidence is different for the other size groups. The most evident swings can be seen among the largest firms, with the other two size groups characterized by less pronounced cyclical behaviors.

Fig. 2
figure 2

Entry dynamics in Tunisia by firms’ size (during the period 2000–2013)

Full size image

The plot in Fig. 3 also depicts the different entry dynamics characterizing the size groups. Small firm entry is characterized by a smooth increasing trend, while the other size groups show steep decreasing trends. Overall, the contribution of small firms to entry dynamics is crucial and increases over time, while the shares of other size groups are lower, and their weights appear to decrease over time.

Fig. 3
figure 3

Survival rate for firms started in 2000 by size

Full size image

The analysis of entry and exit dynamics shows that the Tunisian economic environment is characterized by moderate dynamism, driven mostly by small firms. Table 2 presents figures for the survival rates of firms born in 2000. Consistent with the previous evidence, 98.9% of the firms created in 2000 are small firms, accounting for about 73% of the country’s total employment. If we look at survival rates, we find that 69% of small firms survived for five years, and 58% survived for 10 years. These shares are lower for medium-sized and very large firms, while survival rates for medium-large firms were about 73% in 2005 and 50% in 2010. These figures suggest that in the medium term, small firms have a higher chance of surviving than other firm sizes. Figure 3 confirms this result, showing that the survival rates curves of small and medium-large firms intersect between 2007 and 2008, showing that small firms are more persistent than other firms. In addition, still surviving small firms accounted for 78.5% of total employment in 2005, and 83.4% in 2010. This implies that survivor small firms were able to grow and create new jobs.

Table 2 Start-ups in 2000 and survivors at 2005 and 2010
Full size table

Table 3 presents net job creation rates by firm size and event (average values for 2000–2013). Overall, our calculations show that in the period observed the dynamics were positive, despite a quite large impact of exits on job destruction. The breakdown by size group suggests that the bulk of job creation comes from very large firms. However, if we look at the difference between incumbents and entrants, we find that the negative performance of small firms is due to job destruction by incumbents. If we compare the impact of entrants and exiting firms, the net effect for small firms is positive, while it is negative for very large firms.

Table 3 Net job creation by entering, incumbent and exiting firms in Tunisia (yearly averages during the period 2000–2013)
Full size table

This general overview of industrial dynamics in Tunisia culminates in Table 4 which presents data for 2012 on the breakdown of gross operating profits by size group, and their distribution. These figures are fairly reliable for all size group except small firms where the survey response rate was only around 21%. However, small firms show the lowest levels of gross operating profits, suggesting that although they have the highest social impact in terms of job creation, they do not seem to have a strong economic impact in terms of wealth creation.

Table 4 Gross operating profit in 2012
Full size table

4.2 Small firms in Tunisia: an overview

The detailed analysis of small firms is based on a survey administered by the Tunisian Statistical Office. The dataset coverage is very large and includes information on more than 90% of small firms. Small firms in Tunisia have several incentives not to become part of the informal economy. First, the cost of registration and firm declaration is low; second, the physical owner/employee can benefit from the social security and pension system; third, the fiscal administration in Tunisia is very efficient and there are several monitoring mechanisms which provide information on firm activity.

Within the group of small business, we can distinguish three sub-categories, based on legal status: legal person (personnes morales) or physical persons part of the real tax regime (PP_RR) which is composed mainly of liberal professions); physical persons belonging to the lump-sum regime (regime forfaitaire) (PP_RF)—who pay a fixed amount annually without the need to declare their revenue. Being able to assign small firms to these sub-categories allows us to investigate the extent to which firms that benefit from the régime forfaitaire show better survival and job creation rates.

Table 5 reports the breakdown of small firms’ survival rates by legal status. The bulk of the new firms created in 2000 benefit from the régime forfaitaire (60.8%), and account for about the 66% of total employment in small firms. Individual firms represent the lowest share (13.6) and account for only 4.6% of total employment.

Table 5 Small firms start-ups in 2000 and survivors at 2005 and 2010
Full size table

In relation to survival rates, about 69% of small firms were still active in 2005, and 58% were still active in 2010. The data show that while small firms in the régime forfaitaire show the lowest survival rates in both 2005 and 2010, they still account for the largest share of small firm employment as compared to those in the régime reel. At the time of their formation, small firms benefiting from the regime forfaitarie account for a larger share of job creation than the small firms labelled ‘personnes morales’. This is because the former are mostly individual firms. After 5 years, the evidence is opposite, as the number of employees of small firms ‘personnes morales’ is 1.6 times higher than that of small firms in the regime forfaitaire.

Table 6 presents figures for net job creation. Although the general evidence for small firms seems to be negative, in this table this is driven by the performance of ‘personnes morales’, i.e., enterprises owned by a legal entity. Among firms owned by a physical individual, those subject to the régime forfaitaire seem to show the best performance for job creation; the net effect is positive when we account for job losses from active firms, and is better than the performance of firms in the regime réel.

Table 6 Net job creation by small firms, by event
Full size table

Table 7 presents the distribution of small firms’ gross operating profits by legal status. Although the figures should be interpreted with caution, due to the low response rate of small firms to the questionnaire already referred to, they suggest that firms in the régime forfaitaire show the highest average values for gross operating profit, and the standard deviation is lower than for the other two types of small firms, pointing to a less skewed distribution.

Table 7 Gross operating profit of small firms in 2012
Full size table

In summary, we find that individual firms under the lump sum tax regime (regime forfaitaire) seem to be the main source of employment. At the same time, small businesses have the lowest revenues at slightly higher than the poverty level revenue.

One of the main findings of our study is that the survival (fitness) of small firms is more important in the long run than that of other types of firm. This survival rate has not been affected by the several economic crises that the Tunisian economy has experienced, or by the degradation of the economic climate. In the crisis years (2005, 2009 and 2011) in Tunisia the creation of small firms followed the same trend.

Section 4 presents the econometric model to explore the economic impact of firms on net job creation.

5 Econometric analysis

5.1 The model

We estimate employment-weighted firm-level regressions for net employment growth using the DHS (Davis et al. 1996) measure of firm-level employment growth, where g ist is the change in employment from year t-1 to year t, divided by average size: \(g_{ist} = 2\frac{{E_{ist} - E_{ist - 1} }}{{(E_{ist} + E_{ist - 1} )}}\), and where E ist denotes employment in firm i of type s at year t. This measure is symmetric, bounded between −2 and 2, and accounts for both entry and exit. By virtue of employment weighting, the mean of the dependent variable is equal to the appropriate employment weighted mean, and consequently coefficient estimates can be interpreted as employment weighted conditional means.

Our most general specification thus takes the form;

$$g_{ist} = \beta_{S} Size + \beta_{\tau } \tau + \beta_{I} I + e_{it}$$

where Size is a vector of the size dummies (previous year’s size), \(\tau\) is a vector of the time dummies and I is a vector of the industry dummies. This specification and the models it nests enable us to test a range of hypotheses. Essentially, we would expect the coefficient estimate of small firms to be larger than that of large firms β Ssmall  > β Slarge .

5.2 Results

The results of the econometric estimations are presented in Table 8 which highlights three complementary results: (1) we find strong support for the role of small firms in job creation; (2) the industry dynamics of small firms’ entry-exit has an impact on job creation (and applies to small firms in all economic sectors); (3) medium and large firms have no impact on net job creation in Tunisia.

Table 8 Econometric results
Full size table

5.2.1 Strong confirmation of the small firms’ dynamics in matter of net job creation

Our results show that net employment growth depends strongly on firm size. In other words, small firms are responsible for net job creation in Tunisia. This is true particular for small firms under the “forfaitaire” regime. This finding confirms previous studies of DCs and emerging countries. The novelty in our study is that this innovative regulation seems to be working well and is ensuring job creation. Small firms under a different fiscal regime seem to perform less well in relation to job creation. Therefore, fiscal system reforms related to new and small firms seem to be playing a prominent role in job creation. However, the introduction of a tax reform does not lead automatically to higher rates of job creation and Tunisia still suffers from high rates of unemployment, especially for skilled people.

5.2.2 Entry-exit dynamics of small firms affect job creation

Small firms under the “forfaitaire” regime are the most important category for job creation, and also for entry and exit. Most entrepreneurs in Tunisia prefer to enter the market as a small firm under this regime. Being subject to a fixed lump sum for taxes encourages the formation of new business and new activity. At the same time, our descriptive statistics show that Tunisia is suffering from scarce entry of big firms which could introduce new dynamics into the market. Our results do not allow us to explain whether this entry and exit dynamics create turbulence in the economy or not. The Tunisian economy is weak due to a lack of innovative firms, and only the entry of small firms under forfaitaire regime is having a positive impact on job creation. Most of these jobs are related to self-employment.

The literature shows that firm entry and exit are correlated in emerging countries (Bartelsman et al. 2004) and advanced countries (Bartelsman et al. 2005). Consistent with these studies, our results show that in Tunisia both entry and exit of small firms are correlated with job creation (respectively positively and negatively). They show also that only active small firms (especially those under the regime forfaitaire) have an impact on job creation. Active small firms under this regime are contributing positively to the creation of jobs and are improving the economic situation. The other categories of firms are having no impact on job creation.

The exit of small firms, especially those under the forfaitaire regime, is contributing negatively to job creation. Our results confirm the findings in the literature which suggests that small firms are the most dynamic category in relation to entry and exit. The evidence for Tunisia is in line with the literature.

5.2.3 Small firm dynamics apply to all economic sectors

Our findings show that whatever the sector of activity whether services, trade or industry, small firms under the régime forfaitaire are net contributors to job creation. However, in the case of services not the entire sample of small firms is correlated with job creation. Our findings show no substantial differences among sectors, and confirm that the fiscal regulation is important for more job creation across the entire economy. In other words, there is no specific regulation driving creation of jobs or entreprises in specific sectors. The economics literature studies the impact of this regulation by examining the differences among sectors (Tabarrok and Goldschlag 2014):

“Regulatory stringency does vary greatly by industry. More than 50 percent of industries, for example, received fewer than 354 regulatory restrictions between 1998 and 2012, while the top 5 percent of industries were subjected to more than 16,000 restrictions.” (Tabarrok and Goldschlag 2014: p. 3).

5.3 Discussion

Our findings seem to suggest that despite the capacity for job creation of small firms in Tunisia, the economy is suffering from structural unemployment and an inability to absorb the unemployed youth and especially those with tertiary education. Actually, since most small business in Tunisia are active in traditional sectors and are not technological start-ups, they might create jobs for unqualified rather than skilled people. Middle sized and big firms make no net contribution to job creation. Based on these findings, in Tunisia we observe that the returns to education are close to zero (World Bank 2008). The more highly educated the individual, the lower the chances of finding a job. Tunisia should encourage job creation in technological start-ups and promote innovators able to employ qualified people that is beyond the capacity of the tax regime. The structure of the economy, which is dominated by low-tech industry and low productivity services, the structural weaknesses and the low levels of competitiveness are limiting this dynamics. There is new regulation under discussion, which would allow technology based firms to be tax free for 10 years after entry.

The Tunisian government also provides huge subsidies for big and medium sized firms in Tunisia. It should impose grater conditionality on employment by those firms, and encourage employment of women. The rate of female participation in the job market in Tunisia is among the lowest in the world.

The quality of entrepreneurs is another issue. Most small firms under the regime forfaitaire are defensive and necessity entrepreneurs rather than Schumpeterian innovative entrepreneurs (Baumol 1990). This fiscal policy has ensured less informality by providing incentives for survival-driven, self-employment firms (Naudé 2010) to pay small amounts of tax and operate in the formal sector. Tunisia needs more opportunity entrepreneurs operating in innovative sectors to create more jobs.

6 Conclusions

This paper presented a simple economic model for the link between firm size and job creation in Tunisia (Davis et al. 1996). Our findings provide support to the literature on the specific role of small firms in job creation. Small firms contribute positively to job creation, while medium and large firms have no impact on jobs. Moreover, our model shows that the entry and exit dynamics of small firms have a deep impact on job creation. On the top of this, the innovative regulation, the regime forfaitaire, introduced in Tunisia has had an important impact, in that small firms benefiting from this particular scheme proved to have better performances in terms of net job creation.

Our results are not conclusive, and call for a deeper investigation of the role of composition effects in shaping the relationship between industrial dynamics and net job creation. Actually, Tunisia remains a fragile economy locked into mature specializations and a possible interpretation of our results may refer to the lack of technology-based firms able to absorb the high levels of unemployed skilled workers. Our results raise a basic question as to what extent the regime forfaitaire is more likely to provide an incentive for low-skilled workers but not skilled ones (and especially those with tertiary education). Future studies should therefore investigate the individual determinants of new firm formation in Tunisia, to ascertain whether low skilled people are more likely to create new firms under the as a form of self-employment, while skilled people prefer unemployment to entrepreneurship.