Analysis of Factors Affecting Firm Innovation: An Empirical Investigation for Ethiopian Firms

ABSTRACT

The present study concentrates on answering following questions: (i) Do firm’s internal factors such as size, age, ownership, R&D spending, and exporting significantly predict the propensity of firms to engage in innovative activities in Ethiopia? (ii) Do firm’s external factors, namely, access to finance and competition, significantly predict the propensity of firms to engage in innovative activities in Ethiopia? The data source of the paper is a survey conducted by World Bank in 2015 on Ethiopian firms. Then, the factors influencing the innovativeness of Ethiopian firms are analyzed by using a logistic regression model. Results obtained through logistic regression suggest that internal factors such as the size of the firms, education level of the employees, technology adoption capability of managers, on-job training, and R&D expenditures have a significant positive impact on both types of innovations, while private and foreign ownership were only found to be significant for process innovation. Estimations also showed that access to finance has a significant (positive) impact on product innovation and the impact of the competition level is significant (negative) only in terms of process innovation. Most importantly, we found that, disregarding their statistical significance, all explanatory variables influence both process and product innovation in the same direction, probably because of the positive correlation between the two.

KEYWORDS:

• Innovation
• Ethiopia
• Logistic regression

Notes

¹ For a comprehensive survey examining all aspects of the topic, see Hall and Rosenberg (2010).

² See OECD (2018, p. 28) for the evolution of the Oslo Manual.

³ See http://www.worldbank.org/en/country/ethiopia/publication/ethiopia-great-run-growth-acceleration-how-to-pace-it.

⁴ See Table 3.

⁵ Exceptions are Spielman, Davis, Negash, and Ayele (2011) for small farmers and Gebreyesus (2011) for micro enterprises in Ethiopia.

⁶ See for instance, Acemoglu et al. (2006).

⁷ The raw data of the survey are available online at www.enterprisesurveys.org.

⁸ The current constitution of Ethiopia, ratified in 1995, establishes the federal structure based on nine semi-autonomous administrative regions.

⁹ For a discussion about the Oslo Manual’s innovation definition, see (Beyhan, Dayar, Fındık, & Tandoğan, 2009).

¹⁰ See for instance, (Ayyagari et al., 2011; Chadee & Roxas, 2013; Mahendra et al., 2015).

¹¹ The data were collected from 6 regions, as stated earlier, and 7 major sectors that include 1) Food and Beverages, 2) Textile and Garments including leather, 3) Non-metallic mineral products, 4) other manufacturing, 5) Transportation, 6) Retail, and 7) Other Services. Therefore, 5 dummy variables for accounting regional differences of innovation and 6 dummies for accounting sectorial fixed effect are used.

¹² Positive correlation between foreign ownership and process innovation only is also reported by Alvarez and Robertson (2004) and Arza and López (2010).

¹³ As mentioned above, our study results are based on a self-report survey data like many others in this realm (Mairesse & Mohnen, 2010). It is a well-known fact that in such surveys evaluations suffer from numerous problems such as understanding questions, processing information, capacity of memory, and reporting biases. In order to eliminate these problems, at least partly, we rerun all regressions by using the data by taking into account the data the interviewer considered “Truthful” only. These new estimates with “restricted data” gave nearly similar results with the ones using “unrestricted data”. Therefore, and to save space, we did not report the new estimates. Similarly, we re-estimated the models by using non-log forms of the continuous variables. These estimations also gave similar results.

¹⁴ For instance, introducing a new product or increasing the quality of an existing product is probably more costly than the implantation of a new delivery method which reduces transportation costs.