Innovation capabilities in West African firms

Ananeh-Frempong, Betty-Ann

Innovation capabilities in West African firms

Thesis / Dissertation

2026

Publisher

University of Cape Town

Department

School of Economics

Faculty

Faculty of Commerce

Abstract

This thesis examines the factors that influence innovation by West African firms. It also seeks to determine the relationship between the different types of innovation. In addition, a case study is used to complement the quantitative chapters. It sets out to provide an in-depth understanding of how backward linkage firms in the Ghanaian gold mining sector develop innovation capability through knowledge transfer and how this is influenced by government policy on procurement and local content. There was little data on innovation outcomes and innovative capability activity of firms in the West African region until the World Bank Enterprise Survey conducted investment climate assessment surveys that included questions about innovation in firms in the region beginning in 2014. This has provided the opportunity to conduct research on the determinants of innovation outcomes in the region. The case study provides insights into issues not covered in the innovation surveys that influence innovation outcomes such as transfer of knowledge within an industry and government policy aimed at an industry. Chapter two uses innovation data for ten West African countries to ascertain how innovative activities influence the probability of innovating. The innovation activities used in this chapter are R&D investment, investment in training programmes for employees, the purchase of assets and equipment, and making an export sale. Product and process innovation are the dependent variables. The chapter uses two recursive bivariate probit models to ascertain correlation between product and process innovation. The estimation results show that for both recursive bivariate probit models, firms that engage in one innovation outcome have an increased probability by over 90% of engaging in the other innovation outcome. Thus, the recursive bivariate probit results show that product innovation influences process innovation and vice versa. The marginal effects show that of the innovation activities, R&D investment had the largest marginal effect on the probability that a firm would engage in product and process innovation for both models. The results also show that investment in training programmes for employees increased the probability of innovating in product and process innovation in the first model and in product innovation in the second model. Investment in assets and equipment increased the probability that a firm will engage in process innovation in the first model and in both product and process innovation in the second model. Making an export sale increased the probability that a firm will engage in process innovation in the first model and in product innovation in the second model. Chapter three uses innovation data for Ghana and Nigeria because it includes data on marketing and organisational innovation. The objective of chapter three is to investigate how product, process, marketing, and organisational innovation are related to each other. This chapter also assesses the influence of innovation activities on all the innovation outcomes. A multivariate probit model with the four types of innovation as the outcome variables and innovation activities such as R&D, worker training, export, and purchase of assets and equipment included as regressors is used for this chapter. The regression results show that three innovation activities namely R&D, worker training and purchase of assets and equipment increase the probability that firms will engage in all the four types of innovation. Export however is not statistically significant for product and process innovation although it increases the probability that firms will engage in marketing and organisational innovation. The results of the off-diagonal elements of the correlation matrix show that there is correlation between the error terms of the innovation outcomes. This means that the innovation outcomes are positively associated and dependent on each other. The values of the covariance matrix show that there is a stronger correlation between the error terms for product and process innovation. Similarly, the error term for marketing innovation is most strongly correlated to that for organisational innovation. Thus, the regression results show that there is a relationship between all four types of innovation. The case study is to assess how the local content policy has encouraged local participation in the backward linkage firms to the gold mining industry in Ghana. It assesses the prevalence of innovation in these firms. It also assesses the development of innovation capability through the transfer of knowledge from multinational corporations to local supply firms. This was done by interviewing the management of six firms that supply goods and services to the mining companies. These firms are all affiliate members of the Ghana Chamber of Mines. The chapter shows that government policy to promote the local participation of backward linkage firms by increasing Ghanaian content has been able to substantially increase Ghanaian participation in the supply of mining inputs to the mining firms. The management of the firms believed that the policy has also encouraged knowledge transfer between local firms and multinational corporations. Yet, it can be deduced from the supply firms interviewed that there is not much innovation in these firms. The present policy direction encourages the importation of mining inputs over local manufacturing and value-addition. The firms interviewed that import finished goods expressed an interest in adding value through the establishment of assembly and or manufacturing plants if supported by the government.