CONTRIBUTION OF INTERNATIONALISATION TO SME GROWTH: EVIDENCE FROM THE KENYAN MANUFACTURING SECTOR

Sub-Saharan African countries are among the poorest countries in the world and there is a need to develop their economies. Researchers suggest the promotion of small and medium-sized enterprises (SMEs) to foster economic development in countries. Internationalisation has been proved to be a key strategy for SME growth. This study investigates the effect of internationalisation on manufacturing SME growth in Kenya. Kenya is a developing country and the leading economy in the East African community. Using data from the World Bank enterprise survey, a sample of 94 SMEs operating in Kenya between 2013 and 2018 was selected. Multiple linear regression analysis using ordinary least square (OLS) was applied and the results revealed that internationalisation through direct exports contributed positively to the growth of manufacturing SMEs in Kenya. The findings of the study suggest that policy makers should promote internationalisation of SMEs in Kenya to improve the doing business environment in general and remove external barriers to internationalisation of SMEs at the national and international level in particular.


INTRODUCTION
SMEs constitute an important driver of economic development in countries (Memili, Fang, Chrisman, & Massis, 2015). SMEs are critical to increase competitiveness and the creation of important innovation systems for developing countries. For the promotion of SME growth, it is important to improve the doing business environment and strengthen the capacity of SMEs to take advantage of trade and investment opportunities. This can lead to an increase in growth and a reduction of poverty in developing countries (Kesk, et al., 2010).
Internationalisation enhances significantly the performance of firms. The extent to which companies venture in foreign trade is positively associated with their performance (Lu & Beamish, 2001). Going international has become critical to the competitiveness of companies of all sizes. With globalisation, companies adopting a global strategy can easily take advantage of cross-border activities, which are important for revenue growth as well as the exchange of skills and capabilities that contribute to SME growth in an African economy and to formulate recommendations that can be used by policy makers in Africa.
This article is organised as follows. Section 1 presents the literature review; Section 2 provides the data and research methodology; Section 3 introduces the results; and the final section draws the conclusion.

LITERATURE REVIEW
This section presents the underpinning theory of SME growth (resource-based theory), empirical studies on internationalisation and SME growth as well as the drivers of enterprise growth.
The resource-based theory attempts to explain the reasons behind enterprise competitiveness. Growth is a component of the competitiveness of enterprises. The resource-based theory states that enterprises that use strategic factor markets can achieve sustained competitive advantages and economic rents (Dagnino, 2012). This theory can be used to explain the growth of enterprises when different strategies based on resources are involved, including involvement in internationalisation.
Internationalisation is considered the process of venturing in international trade. Many researchers have conducted studies on internationalisation of SMEs. Many reasons have been pointed out to be behind internationalisation of firms. Korsakien and Tvaronavi (2012), Korsakiene (2014), in a study on barriers and motives of internationalisation of Lithuanian SMEs, showed that profit goals and networks were the main factors affecting internationalisation. Increasing competition was proved to be the most important driver behind internationalisation of firms. A study conducted by Jekanyika (2012) in Kenya showed that internationalisation of companies was motivated by internal factors such as the reputation of the business, managerial orientation, increasing market share and revenue, innovations in technology and more flexibility in operations.
There are many ways for SMEs to internationalise. Mupemhi (2013) showed that internationalisation could be done through exporting, manufacturing in foreigh countries, partnerships or resource sharing with foreigh companies. The study of Korsakien and Tvaronavi (2012) on internationalisation of SMEs revealed that Lithuanian SMEs internationalised through exporting while Norwegian SMEs chose subcontracting, licensing or the creation of representative abroad. Norwegian SMEs were found to be at the more advanced stage of internationalisation. Mtigwe (2010) investigated entrepreneurial firm internationalisation process in the Southern African context, the results revealed that depending on the international experience small enterprises in the South-African region had a common preference for exporting.
Many studies proved the influence of internationalisation on SME growth. Chelliah, et al. (2010) showed that SMEs could go international to gain rapid growth. Internationalisation can promote performance and motivate firms to continuously grasp foreign markets. Lu and Beamish (2001) demonstrated that the extent of a company's involvement in foreign direct investment defined the impact of internationalisation on the company's performance. When companies are

Economics and Business
__________________________________________________________________________ 2020 / 34 264 involved in foreign direct investment activities, their profitability declines but a higher level of involvement in foreign direct investment results in higher performance. Love and Roper (2015) showed that increasing exports had a positive impact on the performance of companies; this performance was measured by productivity and growth. Golovko and Valentini (2011), using unbalanced panel data on Spanish manufacturing companies for the period between 1990 and 1999, revealed that engaging in international trade trough exports had a positive impact on SME growth.
SMEs can choose between direct exports and indirect exports. Direct exports are concerned with completing trade with a consumer in a foreign market without using an intermediary, while in indirect exports SMEs involve an intermediary firm to reach customers in the foreign market. Intermediary exports are smaller in size, ship more products but they reach fewer foreign countries compared to direct exporters (Bernard, 2011).
Apart from internationalisation, many other factors influence the growth of enterprises. Jamali, et al. (2012) found out that for the manufacturing sector, the type of ownerships influenced enterprise growth. Using a sample of Spanish SMEs, Fernández (2006) showed that the ownership type was linked with the decision to internationalise. Family ownership was found to be negatively linked to internationalisation, while internationalisation and corporate ownership were found to be positively linked. George, Wiklund & Zhara (2005), in a study on the relationship between internationalisation of SMEs and ownership using a sample of 899 Swedish SMEs, revealed that external owners were more likely to take the risk of internationalisation than internal owners. Cerrato and Piva (2012) carried a study on the drivers of internationalisation using a sample of 1324 Italian manufacturing SMEs; he revealed that the level of foreign ownership in SMEs had a positive impact on internationalisation. Jamali, et al. (2012) analysed the growth of manufacturing companies in Iran. He used panel data on 137 sub-sectors between 1995 and 2002. Using a generalised method of moment (GMM), his findings showed that apart from internationalisation (export, import), the growth of SMEs and large firms was driven by capital labour ratio, technology usage, sunk cost, age, minimum efficient scale, and total factor productivity.
Technology innovation is a critical strategy for promoting the growth of SMEs (Golovko & Valentini, 2011). Love and Roper (2015) showed that there was a strong positive association between innovation and performance in terms of productivity and/or growth. Innovation is important to improve performance. Nunes (2013) investigated 1845 Portuguese SMEs, using a two-step estimation method. The study revealed that age and size influenced positively growth of young SMEs, while age and size had no impact on the growth of old SMEs. Levratto et al. (2010) investigated the determinants of growth for SMEs. They carried out a longitudinal study on French manufacturing firms and found out that the size of enterprises was one of the important drivers of enterprise growth. Morone and Testa (2008) studied the determinants of growth for Italian SMEs, using a sample of 2600

Economics and Business
__________________________________________________________________________ 2020 / 34 265 SMEs. They found out that the age of SMEs was negatively related to growth. Young SMEs experienced more growth than older ones.
While the previous studies have been based on the reasons behind SME internationalisation and how internationalisation influences SME growth in several countries, the present study focuses on the impact of internationalisation on SME growth in the Kenyan manufacturing sector.

RESEARCH METHODOLOGY
This section presents the data used in the present study and the data analysis techniques.

Data
The dataset used in the present study is from the World Bank Enterprise Survey completed in 2018. The period concerned is between 2013 and 2018. After treatment, cross-section data on 94 manufacturing SMEs were selected for our analysis. The selection of SMEs was based on the availability of data related to the value of sales between 2013 and 2018.

Empirical Model Formulation and Measurement
In the present study, a multiple linear regression analysis using ordinary least squares (OLS) is used to investigate the interrelationship between internationalisation and growth. OLS approach is used because it is suitable for a quantitative continuous dependent variable and independent quantitative continuous or dummy variables (Baltagi, 2011).
Thus, we specify the empirical model denoted by the following regression equation:

Growth
Referring to Fisman and Svensson (2007), Cieslik and Kaciak (2009), Beck and Maksimovic (2002), Hessels and Parker (2013), we proxied growth by the change in the volume of sales. Thus, the percentage of SME growth is measured by the

Economics and Business
__________________________________________________________________________ 2020 / 34 266 difference between sales for two years. In the present study, Growth in 2018 = Sales 2018 − Sales 2013; the long-time horizon is explained by the availability of data.

Internationalisation
Internationalisation can take different forms: export, import, indirect export, and indirect import (Wolszczak-Derlacz, 2018). Internationalisation is measured by export intensity (Cieslik and Kaciak, 2009). In the present study, three variables were selected to measure internationalisation: direct exports (measured by the growth in direct exports between 2013 and 2018), indirect exports (measured by the growth in indirect exports between 2013 and 2018.

Control Variables
This study includes several control variables, such as foreign ownership (measured by the percentage of foreign ownership in the SMEs), the size of SMEs measured by the number of full-time employees, innovation which is a dummy variable measured by the introduction of new or significantly improved products or services, and the duration of the SME activity measured by the number of years since the establishment started operating.

RESULTS
This section presents three points: the descriptive statistics, the correlation matrix and the relationships between the independent and dependent variables in the estimated model. Table 1 below presents the descriptive statistics of the quantitative variables. The values of the minimum, the maximum, the mean and the standard deviation for each variable are given in Table 1. Table 2 presents the frequencies for the variable innovation since it is a dummy variable. The results in Table 1 provide the summary of the descriptive statistics for the variables. The average growth rate of SMEs in the manufacturing sector of Kenya between 2013 and 2018 was 642.9 %. This is an important performance for a developing country where the growth of SMEs contributes to new jobs and wealth (Kesk, 2010). The minimum number of employees is two, while the maximum is 800, and the average is 40. The average growth of direct export was −8.99 %, which implied an average decreasing trend of direct exportations; however, the average increase in indirect exports was 19.20 %. The average duration of the activity of SMEs is 33 years of operation and 3 years is the minimum age. The average foreign ownership in Kenyan manufacturing SMEs is 9.48 %. The results in Table 2 show that 47.9 % of SMEs interviewed have introduced new or significantly improved products or services. This is a good move for Kenyan SMEs in the manufacturing sector as innovation is proved to impact significantly the growth of SMEs (Golovko & Valentini, 2011;Love & Roper, 2015). Table 3 below presents the results of the correlation analysis between the study variables. In Table 3, SME growth is significantly positively correlated with direct export, the size of SMEs and the duration of the SME activity. Variables that are not correlated with SME growth are excluded from the estimated model. The analysis of the correlation between the independent variables shows that there is no correlation between the three retained independent variables. Referring to Baltagi (2011), there is not a problem of multicolinearity given that all correlation coefficients between independent variables are not significant. The results of the regression analysis complete the correlation analysis, showing the direction of the relationship and also how the change in independent variables can affect the dependent variable.

Regression Results
Table 4 below presents the results of the estimated multiple regression model. The results in Table 4 indicate that three variables are significant with P-values below the significant level of 0.05 and 0.1. These are direct export, duration of the SME activity and the size of SMEs. The coefficients of direct export and size are positive; this implies that an increase in direct exports and the size of SMEs contribute significantly to growth of SMEs. The coefficient of duration of the SME activity is negative; this shows that younger SMEs have high chances of growing than older SMEs.
The Fisher test is significant with a P-value of 0.040, which is below the signification level of 0.05. This means that the model estimated is, in general, significant. An adjusted R-squared of 0.1638 (16.38 %) implies a low explanatory power of the estimated model that can be improved by adding more independent variables (Wooldridge, 2016).
Internationalisation through exports (direct) contributes significantly to increased growth of Kenyan manufacturing SMEs. Increasing direct exports results in a high growth rate of Kenyan SMEs. It is therefore a critical strategy to foster the growth of Kenyan SMEs. This result is in line with the findings of Love and Roper (2015), Golovko andValentini (2011), Lu andBeamish (2001). The findings of their research showed that SMEs involved in direct exportations had a high growth rate. In the context of African countries that are working to expand their economies and specifically Kenya, it is important to identify strategies that can be used to attain that objective. The results of this study support the key role played by exportations in expanding the activities of Kenyan SMEs. This can result in job creation, an increase of taxes paid to the government as well as the income of SME shareholders. In Table 3, the results showed that the relationship between indirect exports and SME growth was not significant; this could be explained by the fact that indirect exports took place through an intermediary firm that could increase the cost of products and consequently the final price of products as supported by Bernard et al. (2011).
The results in Table 4 depict that the size of SMEs has a positive impact on SME growth. This result implies that an increase in the number of employees results in an increase of the growth rate of SMEs. Generally, SMEs that use a large number of employees are more likely to produce more and to grow faster than those with a limited number of employees. This result confirms the findings of the study by Levratto et al. (2010) who demonstrated that in France the size of manufacturing companies was a driver of growth. The findings of the present study support that in Kenya the size of SMEs is a driver of growth.
The results in Table 4 support that in Kenya the duration of SME activities has a positive impact on SME growth. This result is in line with the study of Jamali (2012) that was carried out in Iran; the study showed that the longer the operation period of SMEs, the higher growth rate they had; this could be explained by the experience acquired by SMEs in their operations. In the African context, Akinboade (2015) obtained a similar result; he showed that in Cameroon the likelihood of negative growth or zero growth for SMEs decreased with the age of SMEs. However, the finding contradicts with the study of Morone and Testa (2008), who demonstrated that in Italy young firms are more likely to experience positive growth.

CONCLUSION AND IMPLICATIONS
Many authors, e.g., Paul, Parthasarathy and Gupta (2017), Ibeh et al. (2011) have recommended research on the performance of companies and internationalisation.
The present study sought to find out the impact of internationalisation on the growth of manufacturing SMEs in the leading East African economy, Kenya. Based on the study results, it is clear that Kenya made an important progress for the development of SMEs in the last decade with an average growth rate above 100 %. However, given the need for research on the effects of internationalisation of SMEs in Africa, it was important to investigate the extent to which internationalisation could contribute to the growth of Kenyan SMEs.
The results of the estimated model revealed that internationalisation through direct exports contributed to an increase in the growth of manufacturing SMEs in Kenya. Among the control variables, age and size of SMEs were found to have a positive influence on SME growth.
Kenyan SMEs should therefore invest significantly to develop their exporting capacity through the development of new products and the introduction of innovations in the production process. This will contribute significantly to their growth as well as the development of the Kenyan economy.
Policy makers should enhance internationalisation of SMEs in Kenya, by improving the doing business of SMEs through removing external barriers to internationalisation of SMEs at the national and international level (this can be done through trade agreements in the East African community and the common market for Eastern and Southern Africa where the country is a member), as well as reducing administrative burdens as well as formal and informal trade barriers as emphasised by Wilson (2006). This will lead to the growth of SMEs with all the advantages related to it, such as the creation of jobs, and economic growth as supported by the World Bank (2020) and Kesk (2010).
Future researchers should investigate internal barriers to internationalisation of African SMEs as well as the motivations and barriers to internationalisation of SMEs in Africa.
The present study was based on one country and a limited number of SMEs. Given the nature of the available data, it was not possible to conduct a panel analysis, i.e., a study on African economic regions. Future studies should investigate the barriers to internationalisation of SMEs in African countries.

ACKNOWLEDGEMENT
Thanks to the World Bank for providing the access to the World Bank enterprise survey data.