Separating innovation short-run and long-run technical efficiencies: Evidence from the Economic Community of West African States (ECOWAS).

• Author: Kouakou, Dorgyles C.M.

1. Introduction

   The quest for economic performance is of interest in nations and for-profit organizations. Innovation is one of the important sources of this performance (Steil et al., 2002). Indeed, for instance, it is known to be a source of economic growth (Akcigit and Kerr, 2018), productivity and competitiveness (Carayannis and Grigoroudis, 2016). In view of its importance, factors amenable to increase innovation production have been investigated extensively in many papers. In such a context, many studies have been carried out on the determinants of innovation (see for instance, Gebreeyesus and Mohnen, 2013; Paunov, 2016; Rooks et al., 2012), and especially on those of R&D (see for instance, Balsmeier, 2017; Maskus et al, 2019) which is generally shown to be an important innovation input (Mairesse and Mohnen, 2002).

   A relatively recent stream of literature has been developed concerning the analysis of innovation technical efficiency (henceforth, ITE), that is, the efficient production of innovation. Technical efficiency, indeed, refers to productive efficiency (Farrell, 1957). It is of particular interest as it measures a decision-making unit's ability to achieve the maximum output given a certain level of inputs. In other words, technical efficiency assesses a decision-making unit's ability to produce better. We are interested in ITE in this paper.

   In the current literature, to the best of our knowledge, papers do not measure innovation short-run and long-run technical efficiencies. This is interesting to do as short-run and long-run efficiencies capture different kinds of efficiency and are not associated with the same policy implications (Kumbhakar and Heshmati, 1995). In fact, innovation short-run technical efficiency assesses a decision-making unit's ability to achieve the maximum level of innovation in the short run given its inputs, while innovation long-run technical efficiency refers to this ability in the long run. Also, ceteris paribus, short-run efficiency should be more affected by short-run policies while long-run efficiency should be more targeted through long-run policies. The short-run part of efficiency can be adjusted over time for each individual, while the long-run part varies across individuals but is constant over time (Kumbhakar and Lien, 2017). Measuring short-run and long-run technical efficiencies permits to properly measure the overall technical efficiency which is technical efficiency both in the short and long run (Colombi et al., 2014; Kumbhakar et al, 2014, 2015).

   Furthermore, the current literature on the measurement of ITE has neglected developing countries from Africa, presumably due to lack of data. Indeed, to the best of our knowledge, Kao's (2017) study is a rare example of a research that considers an African country, that is, South Africa which is an emerging country. A study of ITE in African countries is of particular interest in view of the fact that a significant part of these countries is usually found to be part of the countries that exhibit the lowest levels of innovation in the world (1). Technical inefficiency could be an explanation for this fact as it leads to achieve a level of innovation that is lower than what is possible to be achieved given the inputs.

   In this paper, we therefore measure innovation short-run technical efficiency, innovation long-run technical efficiency, and innovation overall technical efficiency, using data from Africa. In particular, we exploit data from the Economic Community of West African States (ECOWAS). We also investigate the potential determinants of innovation efficiency. This is operationalized through the panel data stochastic frontier analysis framework proposed recently by Lai and Kumbhakar (2018), in which the variances of short-run and long-run inefficiencies are modeled as functions of the determinants, and possible endogeneity issue in the estimation of the frontier is handled.

   Our study follows the macroeconomic literature on ITE, where the investigation of African developing countries is a real omission. Among other factors, it is important to fill this gap since for the purposes of designing policies to improve the use and allocation of resources, it is necessary to evaluate ITE levels at the country level (Wang and Huang, 2007). Carrying out such a study in the ECOWAS area would permit to identify both the best innovation practitioners (for benchmarking) and the lagging ones, and then to investigate ways to improve ITE (Guan and Chen, 2012).

   The remainder of the paper is organized as follows. Section 2 highlights ITE as a development issue in West Africa. Section 3 is a literature review on the analysis of ITE at the macroeconomic level. Section 4 presents the econometric modeling. The data, variables and descriptive statistics are presented in section 5. The empirical results are presented and discussed in section 6. Section 7 concludes.

2. The efficient production of innovation in West Africa: A development issue

   West Africa is one of the most important economic areas in Africa. Indeed, it includes Nigeria, one of Africa's leading economic powers, and Cote d'Ivoire, one of the countries exhibiting the highest economic growth rates in Africa in recent years (2). West African countries are grouped within the ECOWAS which includes 15 countries, namely: Benin, Burkina Faso, Cabo Verde, Cote d'Ivoire, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Niger, Nigeria, Senegal, Sierra Leone and Togo.

   The real gross domestic product per capita (GDPPC) averaged 1,285 United States Dollars (USD) in 2019 in West Africa (3). There are significant differences in terms of GDPPC between countries like Niger (524 USD), Togo (631 USD), Guinea-Bissau (650 USD), Liberia (650 USD) or Sierra Leone (650 USD), and other countries like Cabo Verde (3,482 USD), Nigeria (2,503 USD), Cote d'Ivoire (2,328 USD) and Ghana (2,054 USD) (4). Beyond these differences, the figures presented show the need for West African countries to increase their levels of wealth per capita. Indeed, compared to more advanced countries, their levels of GDPPC seem quite low. For instance, in 2019, the GDPPC of France was 38,912 USD. That of the United States of America (USA) was 60,687 USD.

   More generally, at the level of human development, the average human development index (HDI) in West Africa in 2018 was equal to 0.49 (5). This implies a low level of human development, the value of the HDI being lower than 0.5. In 2019, we observed an improvement as the HDI was equal to 0.51. This suggests a kind of medium level of human development. We also note that in recent years, especially since 2009, the rate of monetary poverty (6) in West African countries has very often been between 10% and 69% (7), which is not negligible. From these figures, it appears a need to analyze the factors that can improve the level of development in West Africa and reduce monetary poverty.

   In the economic literature, innovation is acknowledged as a factor promoting economic growth and development (see for instance, Akcigit and Kerr, 2018; Schumpeter, 1912), among other factors, through an increase in competitiveness, productivity and social welfare (see for instance, Amable et al., 2016; Carayannis and Grigoroudis, 2016; Gambardella et al., 2016). A good redistribution of the fruits of the economic growth can then help to reduce the rate of monetary poverty. The production of innovation in West Africa therefore constitutes a development issue, all the more so since the average innovation output index in this area has been decreasing since 2009 (8) (see Figure 1).

   The innovation output index measures a country's level of production of innovation. It is published yearly by the World Intellectual Property Organization (WIPO), INSEAD and Cornell University since 2007, and ranges from 0 to 100. The greater the index, the better the level of innovation output. Figure 1 shows the evolution of the average innovation output index in the ECOWAS area over the recent years. The index fell from 29.94 in 2009 to 10.68 in 2020, which represents a decrease of 19.26 in less than 15 years. It can be seen that the index is stagnating at levels lower than what is achievable, i.e., 30, the value of the 2010 index. These figures show that the level of production of innovation in the ECOWAS area is becoming weaker. Therefore, there is an urgent need to study innovation in this area, in order to make policy recommendations to increase the level of innovation.

   In this context, some studies have investigated the determinants of innovation in countries from the ECOWAS area (see for instance, Fu et al, 2018; Kouakou, 2020; Robson et al, 2009). However, the issue of whether the decision-making units in ECOWAS countries are efficient in producing innovation or not has been completely ignored. This gap is important to fill as inefficiency could also explain the observed reduction of the production of innovation in the ECOWAS area in recent years. Stylized facts seem to give intuitions that inefficiency might be an issue in the ECOWAS area (see Figure 2).

   Figure 2 shows the evolution of both the average innovation output and input indexes of the ECOWAS area. It can be seen that the input index curve is located above that of the output index. This suggests that the countries could be being producing less innovations than what could be expected in view of their endowments in innovation inputs. It also emerges from Figure 2 that over the 2013-2014 and 2016-2017 periods, the innovation production has fallen while innovation inputs have increased. In other words, over these periods, in ECOWAS countries, the production units have increased the factors allowing innovation production, but this did not lead to more innovations; innovation production has rather decreased.

   A plausible explanation for this phenomenon is that the ITE levels of ECOWAS countries have decreased. Therefore, it seems important to...