The role of foreign direct investment (FDI) on domestic entrepreneurship in South Africa

ABSTRACT

This study examines the effect of foreign direct investment on domestic entrepreneurship in South Africa. With the focus on inward capital flows, the study specifically employed stock data and the Global Entrepreneurship Monitor to measure the impact. The data set analysed is for the period 2000–18, and after testing a Threshold Vector Autoregressive model, it was established that there is a short-run and long-run nonlinear relationship between foreign direct investment and domestic entrepreneurship in South Africa. The key findings of the study were that foreign direct investment has a positive short-run and long-run influence on domestic entrepreneurship. The policy recommendations are for government to create an eco-system that supports entrepreneurship through the lowering of regulatory burden on new domestic firms and enact robust sector-specific localisation policies for big corporations.

KEYWORDS:

• Foreign direct investment
• entrepreneurship
• entrepreneurial activity
• firm creation
• global entrepreneurship monitor

1. Introduction

New firm creation is a critical factor in the development and growth of an economy. Studies have confirmed that the creation of new firms contributes more to job creation than existing businesses (Ayyagari et al., 2011). Many studies have argued that new business formation has an indirect effect in securing efficiency, speeding up structural change, driving innovation and providing consumers with greater product choice (Schumpeter, 1947; Acs & Audretsch, 1990; Markusen & Venables, 1999; Fritsch & Mueller, 2004). This results in economic growth and a competitive business environment, while the direct effect of business formation is employment due to the setting up of new capacities (Fritsch & Mueller, 2004). Levine & Renelt (1992) posit that one of the strong determinants of growth is capital investment. Their study followed the Solow growth model that concludes that capital formation determines economic growth. Some scholars argue that Foreign Direct Investment (FDI) increases economic growth in the host country more effectively than domestic investment (Borensztein et al., 1998; Mencinger, 2003).

Furthermore, Margeirsson (2015) postulates that FDI increases domestic competition, leads to technological advances and knowledge spillovers to domestic firms, increases human capital capabilities and encourages firm development in the host country. Therefore, FDI positively influences economic growth, provided the right policy framework exists in the host country (OECD, 2002). However, Mencinger (2003) points out that FDI can have a negative influence in the short term on the host country when Multi-National Enterprises (MNEs) crowd out domestic firms. Additionally, the repatriation of profits can negatively influence the Balance of Payments (BoP), which affects the economic growth of the host country in the long-run (Landsburg, 1979).

Due to a lack of alternative funding models for economic growth, most developing countries over the past few decades have placed much reliance on FDI. Furthermore, being an attractive destination for FDI has become a top priority for their domestic programmes. This translates to a country’s policies geared towards incentivizing the inward flow of FDI (Lautier & Moreaub, 2012). In the short term at least, foreign firms distort the supply of labour because they attract local labour to their firms. The main attraction emanates from foreign-owned firms being able to take advantage of their superior production technology that they can transfer by virtue of being multinationals which results in higher wages offered by the new entrants into the domestic market (de Backer & Sleuwaegen, 2003).

With the focus on job creation and poverty alleviation in South Africa, conscious of the challenge that South Africa has low levels of savings and brain drain, the focus has moved to creating policies that enable a positive environment for inclusive growth and entrepreneurship. This study investigates the influence of FDI on entrepreneurship in South Africa. The objectives of this research to further determine the short-run and long-run relationship between FDI and domestic entrepreneurship in South Africa by determining how domestic entrepreneurship responds/reacts when there is a fluctuation in FDI stock and to distil policy recommendations to support an environment conducive for domestic firms in an eco-system that has FDI.

2. Literature review

2.1. Theoretical framework

The literature is not conclusive as to the nature of the relationship between FDI and entrepreneurship (Jovanovic, 1994; Åslund, 1995; Matzner, 1995). One school of thought maintains that FDI is essentially economic colonialism, where ultimately there is a transfer of economic control and wealth from host economies to external powers (Landsburg, 1979; Ake, 1996; Heijdra & van der Ploeg, 2002). These dependency theorists maintain that FDI, by its nature, is imperialistic and exploitative; that emerging economies are the worse off because of exporting raw materials (without value addition) and systematic exploitation. They further argue that foreign firms crowd out domestic firms by controlling and increasing barriers to entry in key parts of an economy.

Conversely, free-market theorists are of the belief that there are extensive benefits of FDI for host economies (Matzner, 1995; Ugochokwe et al., 2013). The theory posits that from FDI, emerging economies benefit through the transfer of efficient production methods, human capital, western knowledge of business ethics and management traits (Danakol et al., 2014). In addition, due to the presence of FDI in the form of MNEs, advance foreign business practices will spill over to domestic entrepreneurs. Overall, the theory on the relationship between FDI and entrepreneurship attempts to address both the supply side and demand side linkages. Furthermore, FDI flows have varied over the longer horizon rather than the shorter term, implying that their effects are visible over time (Lim et al., 2014).

2.1.1. Positive spillovers

FDI transmits technological advancement to the host country and further enhances labour market skills transfer and knowledge spillovers, whereby individuals employed in MNEs will start their own enterprises using the skills and knowledge gained in their tenure at these multinationals (Fosfuri et al., 2001; Acs et al., 2008; Danakol et al., 2014). The attraction of FDI can also lead to increased levels of investment in the host country’s economy, such as an increase in employment opportunities and higher levels of output for domestic firms (Urata & Kawai, 2000; Geršl et al., 2007). Furthermore, Borensztein et al. (1998) concur that when looking at the determinants of entrepreneurship, FDI is a critical element in fostering economic growth, market development, and technology transfer. This invariably carries over to an overall direct influence on economic growth. Although multiple positive possible effects can be highlighted, the main positive and direct effects of FDI – besides investment on GDP – is the increase in modern technology, managerial skills transfer and better access to finance (Liu et al., 2000).

Caves (1974) postulates that the entrance of foreign competition into the domestic market encourages efficiency, as they tend to enter into markets that have high capital investment requirements. MNEs further boost the optimal distribution of services and goods, resulting in improved competition and more consumer preferences being met, thus breaking down monopolies in a host country industry (Waldkirch & Ofosu, 2010). Albulescu & Tămăşilă (2014) further suggest that another vertical spillover from FDI is the opportunity for domestic firms to become part of the value chain as suppliers and contractors to MNEs and by working with MNEs in joint ventures and other collaborations where the transfer of technology and skills is possible.

2.1.2. Negative spillovers

The theory also suggests that foreign firms may exploit their competitive advantage and market position by monopolising labour to the point that would-be entrepreneurs forgo starting their own enterprises for earning a competitive salary (Danakol et al., 2014). Waldkirch & Ofosu (2010) add that through higher wages, MNEs may attract more workers that are skilled. This leaves domestic firms with limited pickings due to their inability to offer competitive remuneration. MNEs may also crowd out entrepreneurs by raising product market standards and pricing to such an extent that local firms are unable to compete and are forced out of business (Grossman, 1984; Djankov & Hoekman, 2000).

Consequently, De Backer & Sleuwaegen (2003) postulate that, entrepreneurial and worker abilities considered, FDI not only changes the spread of would-be-entrepreneurs but it also lowers the overall quantity of entrepreneurs in the domestic economy. De Backer & Sleuwaegen (2003) based this on extending an existing model by Jovanovic (1994), which suggested that through the entrance of foreign competition, possible local entrepreneurs would choose employment over starting an enterprise.

2.1.3. FDI impact on economic growth

The theory suggests that FDI on economic growth is strongly tied to the presence of MNCs in the domestic economy, stemming from the technological spillovers from developed economies, thus resulting in efficiency and growth (Blomström & Wolf, 1994; Barrell & Pain, 1997; Ramirez, 2000; Fedderke, 2001). Mehic et al. (2013) agree with the technological spillover theory and further posit that FDI has a positive influence on the performance of an economy in that it contributes to capital formation and further augments domestic savings and other domestic investment.

However, literature on this relationship does not agree. Some scholars suggest that FDI has no significant influence on the performance of the host country (Carkovic & Levine, 2002; Mencinger, 2003). This emanates from the empirical findings, which have not necessarily considered the variance in technological development of home country to host country (Mehic et al., 2013). Furthermore, Mehic et al. (2013) suggest that the influence of FDI on growth is dependent on how it performs in relation to domestic capital. According to World Bank (2017), during the financial crisis of 2008–09, FDI inflows into South Africa declined by 1.61% of GDP, while GDP contracted by 1.8%. These statistics support the theory that suggests that reduced FDI results in a contraction when it comes to GDP (Akoto, 2016).

2.1.4. FDI effect on poverty and the role of policy

Borensztein et al. (1998), Hermes & Lensink (2003), and Lensink & Morrissey (2006) suggest that there are conditions that are conducive to FDI enhancing growth in the domestic economy. These conditions are both non-policy and policy conditions, like human capital development, local financial markets and the effect on domestic investment, which directly link to the effect on productivity (Nyawo & van Wyk, 2018). Therefore, there has been limited positive FDI influence in sub-Saharan Africa as there are underdeveloped capital markets, low labour force development, and general low domestic investment and productivity (Morrissey, 2012).

2.2. Empirical framework

2.2.1. Positive impact of FDI

There are many positive spillovers generated by investment from abroad. Dupasquier & Osakwe (2006) conclude that Africa is dependent on infrastructure development. Similar research (Botrić & Škuflić, 2006; Kersan-Skabic & Orlic, 2007) focused on developing and emerging economies, affirming that infrastructure development plays a vital role in the attraction of FDI inflow. On specific infrastructure type, Dutse (2008) isolated Information and Communication Technology (ICT) market infrastructure as the main component that is a positive inducer of FDI inflow.

Expanding on this, Ayyagari & Kosova (2010) analysed the effect of FDI on domestic firm entry in 245 industries in the Czech Republic between 1994 and 2000, using regression analysis. They found that a higher foreign presence induced the entry of domestic firms in the same industry and concluded that this indicated the existence of positive FDI horizontal spillovers. The study also found traces of significant vertical entry spillovers, further establishing that through linkages, FDI stimulates domestic firm entry. Doytch & Epperson (2012) investigated eight countries over a five-year period from 2004 to 2008 and found a positive relationship between FDI and domestic entrepreneurship but only for Middle Income Countries. The Generalised Methods of Moments (GMM) method was used to capture the endogeneity of the explanatory variables. Adegbite & Ayadi (2010), in their study of Nigeria using Ordinary Least Squares regression analysis for the period 1992–2007, also concluded that FDI had a positive influence on economic growth. They posit that the extent of the positive influence can be limited through human capital and macroeconomic environment.

Wach & Wojciechowski (2016) conducted a study that investigated the relationship between inward FDI and the entrepreneurship rate focusing on four Visegrad economies and the study was confined to the period between 2000 and 2012. Using the OLS regression methodology and vector error correction model (VECM), the study found the relationship between stock FDI and the entrepreneurship rate to be statistically positively significant.

2.2.2. Negative impact of FDI

Waldkirch & Ofosu (2010), using the System GMM on Ghana’s manufacturing sector, affirm the negative effect of foreign firms on the productivity of local firms in the short run, due to foreign firms having lower production costs, which enable them to be more competitive in terms of the market and more attractive in terms of the labour market. Using panel data for the period of 1992–98, they further found that the long-run influence of FDI on total factor productivity was also negative and that possible positive effects on domestic productivity would likely take over a decade to realise.

Grossman (1984) supports the notion that the entrance of import competition and FDI decreases the level of domestic entrepreneurship due to the lower reward in entrepreneurial income versus salary income and the further employment of would-be astute domestic entrepreneurs in foreign firms. Aitken & Harrison (1999), using panel data, found that in the Venezuelan context FDI had negative spillover effects. Danakol et al. (2014) support these findings. Their panel data study of 70 countries (inclusive of developed and developing economies) covered the period from 2000 to 2009 and found that the intra-industry relationship between FDI and domestic entrepreneurship was negative. This crowding out is mainly due to the competition effect, in that foreign firms are in direct competition with domestic firms.

2.2.3. U-curve impact of FDI

De Backer & Sleuwaegen (2003) concluded in their research focusing on Belgian manufacturing industries, that there is a positive relationship between FDI and entrepreneurship, due to the spillovers in research and development between foreign and domestic institutions, as well as demonstration and networking. However, through labour market selection and product market pricing, FDI can crowd out domestic entrepreneurs and this phenomenon can be reversed in the long term due to the positive relationship (De Backer & Sleuwaegen, 2003).

2.2.4. Zero impact of FDI

Djankov & Hoekman (2000) used firm-level data for the period 1992–96 in the Czech Republic. They found that joint ventures between FDI and local firms had negative spillovers, especially with local firms that did not have foreign partnerships. However, FDI on its own had zero effect in terms of spillovers. Their study used a random-effects model and the Heckman two-step procedure was utilised to correct any bias in the selected sample.

Konings (2000) in Romania, Bulgaria and Poland support these findings. Using panel data, FDI claims to have a negative effect in Romania and Bulgaria but zero effect in the Polish economy. Due to the absence of any spillover effects on domestic firms from foreign-owned companies, the study concluded that in the case of Poland this was a result of the negative effect of competition overpowering positive technological spillovers. The GMM method was utilised and to account for possible endogeneity in the data, the fixed effects model was utilised.

2.2.5. South African context

Akoto (2016), in his study of the nature of the causal relationship between FDI, GDP and exports in South Africa, found that in the long run FDI has an effect on increasing exports. He further determined that in the short run there is a causal relationship between GDP and exports. The study used quarterly data from 1960 to 2009 sourced from SARB and VECM Granger causality. In addition, Gossel & Biekpe (2013) found that growth in the South African economy is driven by fixed investment and trade, as opposed to capital flows. Using quarterly data for the period 1995–2011, they utilise TYDL methodologies and to overcome the shortcomings of this approach, they then use the augmented Dickey and Fuller root tests. They further found that integrated to the trade-led growth of the country were portfolio inflows and not FDI.

Magombeyi & Odhiambo (2018), investigating the relationship between FDI and poverty in South Africa, reviewed the period 1980–2014. Using an autoregressive distributed lag approach (ARDL), they found that FDI has a positive effect on reducing poverty in the long run and a negative short-run effect on poverty reduction (when poverty is proxied using infant mortality). However, no significant effect was found in the results when using life expectancy and household expenditure and consumption as a proxy (in both the short and long run).

3. Methodology

3.1. Study design

The study followed a deductive approach to answer the research objectives. The study employed the threshold vector error correction model (TVECM) and the nonlinear causal relationship to test the causality and cointegration between FDI and domestic entrepreneurship. This is similar to the TVECM utilised by Wang et al. (2009) to examine the relationship between FDI and domestic gross direct investment (GDI) in Taiwan over the period 1981–2006. They found that by utilising the TVECM that the long-run relationship between FDI and GDI was complementary (Wang et al. 2009). An extensive study of literature on the effect of FDI on entrepreneurship was undertaken and various contexts explored in arriving at a conclusive result for South Africa.

3.2. Data collection

The study uses time-series data over the period of 17 years 6 months, commencing from the year 2000 to 2018. Secondary data was utilised for FDI stock, as the independent variable and TEA as the dependent variable. The two data sets measuring FDI and entrepreneurship are obtained from South African Reserve Bank (2018) database and the GEM (2018) database, respectively. This data is also relevant and suited for the set objectives and significance of the study.

3.3. Data analysis

The study used econometric methods and time series data to achieve predetermined objectives regarding the relations between specific variables, as a result, quantitative methods are suited for the enquiry. The two data sets make a good experimental unit for the application of the TVECM for testing a short-run and long-run effect of FDI on domestic entrepreneurship in South Africa. This data is also relevant and suited for the set objectives and significance of the study.

3.4. Model diagnostics checks

3.4.1. Normality

The Jarque–Bera (JB) test was utilised to test the assumption that a given sample $X_s$ originates from a normal distribution and the estimated residuals for each model are normally distributed. The JB test of normality does better when used on samples of more than 50 observations. The test follows a chi-square distribution with 3 degrees of freedom for sample size of 2000 and above but when the sample is less than 2000, the JB test follows a normal cumulative distribution (NCD).

3.4.2. Serial correlation

For high order test, the Breusch–Godfrey test was used in this study. The null hypothesis is not rejected if the calculated probability is greater than the observed probability value implying that the estimated residuals are not correlated to each other over time.

3.4.3. Heteroscedasticity

The rejection rule is that reject the null hypothesis in favour of the alternative if the observed probability value is greater than the calculated probability value of the test statistic.

3.4.4. Nonlinear test

The results in Table 1 show that linearity null hypothesis does not hold up at all levels of significance for FDI and entrepreneurship. At the conventional levels of significance of 1%, 5% and 10% both the Tsay and BDS test probability values come below. These findings suggest that the proposed models are correct treatments in dealing with asymmetrical and nonlinear data.

Table 1. Nonlinear test.

	FDI	Entrepreneurship
TSAY test
Coefficient	2.664	2.245
P-value	0.0048	0.052
BDS test
Coefficient	48.745	32.353
P-value	0.0000	0.0000

3.4.5. Nonlinear unit root test

In Table 2, the reported modified Wald test indicates that FDI and entrepreneurship series possess some properties associated with nonlinearity since the reported probability value is less than the level of significance. Hence, the null hypothesis of the linear unit root is rejected in favour of the nonlinear unit root. Preliminary analysis of data confirmed that this data is suitable for analysis with the Threshold models.

Table 2. Modified Wald test.

	chi-square	df	p-Value
FDI	753.6	2	0.0001
Entrepreneurship	780.7	2	0.0001

3.5. Empirical results

Prior to the estimation of $\mathrm{T}\mathrm{V}\mathrm{A}\mathrm{R}\left(p\right)$ and $\mathrm{T}\mathrm{V}\mathrm{E}\mathrm{C}\mathrm{M}\left(q\right)$ models, the $\mathrm{A}\mathrm{R}\left(p\right)$ model is estimated and tested for the linear specification. The optimum lag length for bivariate threshold models is selected using the Bayesian Information Criteria (BIC), Akaike Information Criteria (AIC) and Hannan and Quinn Information Criteria (HQIC). Moreover, Diks–Panchenko nonparametric causality test is estimated.

3.5.1. Maximum lag length selection and model estimation

The selected lag length is the one that is used to estimate the models. As reported in Table 3, the BIC and HQ select the optimal lag length as 1. The AIC has the smallest value at the lag length of 2 contradicting the suggestions by the SBC and the HQ criteria. The current study adopted the suggestion by (Tsay, 2005; Scott & Hatemi, 2008), hence the optimal lag length for the study is 1.

Table 3. Lag-length selection.

Number of lags	Information criterions
	AIC	BIC	HQ
1	0.696	0.791**	0.732**
2	0.689**	0.854	0.755
3	0.713	0.942	0.806
4	0.730	1.025	0.840
5	0.763	1.127	0.913
6	0.796	1.219	0.962
7	0.811	1.303	1.019
8	0.842	1.399	1.067

**Signifies a significant lag.

3.5.2. TVECM model and threshold cointegration test

The long-run relationship between the variables suggests that there is a positive nonlinear long-run significant relationship between FDI and Entrepreneurship in South Africa. This study also found a positive nonlinear short-run relationship between FDI and Entrepreneurship. Table 4 reports the results of the TVAR model. Although the econometric tests select a two-regime VAR model, the threshold vector autoregressive of order one, i.e. $\mathrm{T}\mathrm{V}\mathrm{A}\mathrm{R}\left(1\right)$, is estimated using the expectation-maximum (EM) method. Based on the estimation results the TVAR is significant at a 10% level of significance. The variables are stationary, and a TVAR model is estimated to examine the long-run nonlinear effects of FDI towards entrepreneurship. A threshold denoted by $\delta$ is significant in lower and upper regimes. The coefficient metrics (both the parameter 1.1 and 2.1) for lower regime indicates a higher relationship between FDI and entrepreneurship. For the upper regime, the parameter 1.2 is significant at 10% level of significance and the coefficient metrics (parameter 2.2) is highly significant. This clearly indicates a long-run nonlinear relationship between FDI and entrepreneurship.

Table 4. Estimation results of the TVAR model.

Parameter	Estimate	Std error	Prob
Lower regime
${\delta }_1$	511.6996	20.5849	0.0809
${\mathrm{\Phi }}_{11}$	1.0148	0.1051	0.0001
${\mathrm{\Phi }}_{21}$	0.9522	0.0663	0.0002
Upper regime
${\delta }_2$	483.047	23.149	0.0953
${\mathrm{\Phi }}_{12}$	148.465	87.669	0.0756
${\mathrm{\Phi }}_{22}$	0.8630	0.0833	0.0004

To correct the long-run disequilibrium, a short-run dynamic model was estimated with the shock term. The long-run co-movement will be at its equilibrium. To achieve this, a threshold vector error correction model (TVECM) was estimated. The misleading error results from people’s change in preferences, economic down turns, government or policy changes and institutional development. Table 5 reports the results from the TVECM.

Table 5. Parameter estimates of TVECM (1) model.

Parameter	Estimate	Std error	Prob
Lower regime
${\delta }_1$	0.017	0.650	0.076
${\mu }_1$	0.084	0.684	0.003
${\mathrm{\Phi }}_{11}$	0.325	0.0067	0.0043
${\mathrm{\Phi }}_{21}$	0.531	0.0001	0.0002
Upper regime
${\delta }_2$	0.547	0.0004	0.0001
${\mu }_2$	3.857	0.0065	0.0001
${\mathrm{\Phi }}_{12}$	0.234	0.0456	0.0564
${\mathrm{\Phi }}_{22}$	0.546	0.0833	0.0004

Applying equation 3.27–3.28, $\gamma =168$ is the parameter produced by the procedure. Based on this parameter;$\mathrm{T}\mathrm{V}\mathrm{E}\mathrm{C}\mathrm{M}\left(1\right)$ is alienated into two regimes. Regime 1 is defined by those quarterly FDIs where the absolute deviation from the short-term equilibrium is below 69.7% for every observation and in regime 2 the absolute deviation from the short-term equilibrium is above 30.3%. This indicates that the speed at which the SA economy corrects the long-run disequilibrium between FDI and entrepreneurship is at 30.3% quarterly.

The results of a TVAR model indicated a short-run and long-run nonlinear relationship between FDI and entrepreneurship. Aleem & Lahiani (2014) call this nonlinear co-movement the exchange rate pass-through. The effect of FDI on entrepreneurship is that for every 1% increase in FDI, entrepreneurship will increase by 16.8%. Therefore, the implication is that the outcome of an increase in FDI will be a nonlinear increase in domestic entrepreneurship in South Africa, which is in line with Onwuka & Chigozie (2014) results in Nigeria and also to the findings by Ayyagari & Kosova (2010) in the Czech Republic.

A significant positive long-run nonlinear relationship was found between FDI and entrepreneurship, using the cointegration analysis that confirmed the presence of one threshold cointegrating vector. While a convergence towards equilibrium was found in the short-run, even though the adjustment is strong at 69.7% per quarter in regime one which indicates a stable time in the economy. And it is weak above 30.3% for each observation in regime two. Indicating that the speed of adjustment in regime 2 is at 30.3% per quarter. To determine whether the model had a challenge of serial correlation, heteroscedasticity and normality a diagnostic and stability test was done which proved this to not be the case.

Furthermore, the Diks–Panchenko causality test was used to assess the causal link between FDI and Entrepreneurship in South Africa. The results confirmed that indeed there is a bidirectional causal relationship between FDI and Entrepreneurship. There is a positive relationship between FDI and entrepreneurship in South Africa in both the short-run and the long-run. Based on the findings of this study, it can be concluded that there are long-run and short-run nonlinear dynamics within the two series.

3.6. Key findings

The objectives of this research were to determine the short-run and long-run effect of FDI on domestic entrepreneurship in South Africa. The study found that there is a statistically significant relationship between FDI and entrepreneurship in South Africa, both in the short-run and the long-run. The estimated threshold is 168, meaning that an overall 1% increase in FDI will increase entrepreneurship by 16.8%. The low intentions of starting businesses and the low TEA rate suggest that this is not due to would-be entrepreneurs opting to be employed in MNEs, but can be linked to the intrinsic structural challenges of inequality in South Africa.

These findings are in line with Onwuka & Chigozie (2014) study in Nigeria, where they found that FDI had a positive effect on entrepreneurial development and encouraged the inward flow of technological expertise, managerial expertise and capital. This, in turn, stimulates entrepreneurial activities and increases the rate of development of entrepreneurship. Various studies have found a positive effect of FDI on entrepreneurship in the host country. Ayyagari & Kosova (2010) found in the Czech Republic that foreign participation encouraged the entry of domestic firms in the same industry, concluding the presence of horizontal spillovers from MNEs to domestic enterprises.

The findings of this study, as well as other studies that found a statistically significant relationship between FDI and entrepreneurship, are in line with free-market theorists. Free market theorists maintain that FDI has extensive benefits for the host country, especially in emerging markets (Danakol et al., 2014), where positive spillovers such as business management and ethics, human capital and the transfer of efficient production methods occur.

4. Conclusion and recommendations

Governments of developing countries have focused on increasing FDI to their respective countries in the hopes of supplementing the gap that exists because of meagre levels of domestic savings that are insufficient in meeting the investment needs of these countries. In the face of growing levels of inequality, poverty, and unemployment, compounded by lacklustre economic growth, South Africa is no different. The need to stimulate the economy has been a constant topic in recent times. New firms create more jobs than existing businesses and understanding some of the challenges behind the low levels of entrepreneurial intentions in the country is becoming increasingly pressing. It is clear that understanding this would serve as one of the key levers towards poverty alleviation. This understanding would ultimately increase domestic savings rates, which in turn would decrease the dependency on the need for capital inflows.

Short-run policy recommendations: To foster a conducive environment to entrepreneurial activity in South Africa, some of the policy options that government should consider are:

• Lower the regulatory burden on local start-up businesses, thereby increasing ease of doing business in South Africa;

• Tighten the localisation policy around the inclusion of small businesses in the value chain of big corporate players in key sectors.

Long-run policy recommendations: According to the literature, the effect of FDI on the local economy can only be ascertained over the longer term. Therefore, some policies associated with entrepreneurship need to have a long-term focus if desired returns are to be realised from the presence of FDI in South Africa, which include:

• Offering entrepreneurship as a discipline of study at primary and secondary level schooling. This will ensure that those that do not make it to tertiary level are also equipped with knowledge and socialised with the benefits of creating their own enterprises.

• Targeted sectoral incentives to encourage increased FDI to South Africa. For example, in primary industries: Offer tax breaks to corporate organisations that localise procurement and encourage value addition in country, versus exporting raw materials.

• Subsidise small businesses that are participants in these sectors from a technological capability perspective.

This study, motivated by the need to understand the economic issues facing the South African economy, such as high unemployment and poverty levels and the key drivers of these adversities. In addition, the need to comprehend the strategies to work towards solutions for these economic challenges necessitated a study that would look closely at what could have a positive effect on entrepreneurship. The successful implementation of the suggested recommendations could result in:

• Increased local participation in entrepreneurship;

• Reduction in inequality and poverty; and

• Encourage sustainable economic growth.

Empirical studies revealed that entrepreneurship and new venture creation are core to the creation of jobs/provision of employment in an economy. Furthermore, understanding such would serve as one of the key levers towards poverty alleviation and ultimately increased domestic savings rates, which would decrease the dependency on the need for capital inflows. The only way that South Africa can experience sustainable growth in the long term is to become more inclusive and increase participation in the economy of marginalised communities.

Disclosure statement

No potential conflict of interest was reported by the authors.