^{1}

^{1}

^{2}

^{1}

The purpose of this paper was to study the influence of R&D (which included capital, labour and human capital) on economic development in the West African sub-region. However, due to unavailability of data to cover these variables, proxies were used to represent the independent variables. The proxies used were gross fixed capital formation, GDP per person employed, government expenditure on tertiary education as a percentage of GDP and gross enrolment ratio of both sexes to represent capital, labour, R&D and human capital respectively. The study covered all the 15 member-states of the Economic Commission of West African States (ECOWAS). The study covered a period of 2001-2014. The Generalised Least Squares (GLS) method of estimation was adopted with a panel dataset. The findings revealed that the variables had positive impact on the economy while R&D did influence GDP per capital significantly though weakly positive.

The African continent is by far the second-largest and second-most-populous continent in the world, endowed with rich natural resources and yet the poorest, [

The economic situation in West Africa remains fragile and vulnerable to domestic and external shocks [

Hiking poverty rates, market assimilation difficulty into world markets, stagnant economic growth rates, higher unemployment etc. are some challenges the West African sub-region faces. In addition, the following are some destabilising factors posing threats to the sub-region: the fragility of states in the region, internal power struggles, growing militarisation, rapid population growth, a general climate of insecurity which fuels trafficking of all kinds, and lastly various foreign interventions aimed at gaining a toe-hold in this strategic corridor and taking control of its wealth [

Despite the above findings from Ghura, Basu and Calamitsis [

Could these accessions raise questions ranging from the accuracy of these views? Do countries with huge R&D expenditure enjoy faster economic growth rate? And do those countries with more R&D expenditure enjoy from higher productivity? These are questions worth finding answers. Consequently, it is incumbent on governments to support the growth and development of R&D through funding or any other means possible particularly if it is lacking in the private business environment.

Irrespective of the importance of R&D to economic growth that is theoretically having a positive effect on economic growth, empirically there are some challenges in its measurement. These measurement challenges may include differences in aggregated level (companies, industries or countries), sources of data (time periods, countries) and measurements of key variables (stocks, flows or differences) [

The aim of this paper was to investigate the role of research and development expenditures on economic growth of all member states of Economic Community of West African State (ECOWAS). The pushing factor for this paper is that literature [

To the best of our knowledge, there has been little or no information regarding the impact of R&D on economic growth for ECOWAS member states^{1}. This derives our passion for writing this paper. The paper is organized as follows: Section 1 presents the introduction; Section 2 reviews of the literature; Section 3 is an overview of Gross Domestic Product (GDP) and R&D in West African (WA) countries; Section 4 is an augmented Solow model that incorporates R&D and human capital; Section 5 is the methodology and the econometric results; while Section 6 outlines the main conclusions and policy implications of our research.

The engine of economic growth depends on the pace of technological advancement and level of innovation among entrepreneurs in a country. Entrepreneurs have a major role in the development agenda of a country. That is their performances should be based on innovative ways which could generate well designed and efficient production techniques in the current production process and the creation of new and improved products for existing and emerging markets [^{2} and rate of economic growth through the skilful innovations of research workers whom might have accumulated knowledge over the past years of doing research. This may imply that R&D maybe a necessary but not sufficient condition for economic development. A skilful and active labour force (human capital) maybe needed to augment the investments to develop the R&D sector.

Most governments over the past years have acknowledged the significance of R&D to economic growth. They therefore have invested more large amounts of money on regular basis to the development of research institutions. However, some scholars are calling for yet more increased investments in this sector. This has led to most governments facing various competing demands for public funding, especially on how much allocation to education, health, research and other sectors is required. In line with this, in the endogenous growth literature of Romer [

In regards to this, Salter and Martin [

Varying researches have revealed several findings with regards to the influence of R&D on economic growth using panel data analysis. These researches vary from different panels, time periods, variables and econometric methods. A study was conducted by Gumus and Celikay [

Sadraoui et al. (2014) analysed the causality between R&D collaboration and economic growth by using the data of 32 industrialized and developed country for the period between 1970-2012. Results obtained support the argument that there is a strong causality between economic growth and R&D collaboration. On the contrary, the non-causality between R&D collaboration and economic growth couldn’t be refused in several contexts. However, these results show that, if there is such a relationship, a Granger causality test with one or two variants can’t be defined easily.

From the African experience, Alani [

Due to the limitations of data accessibility, annual data for ECOWAS member states for the period of 2001-2014 was used to conduct an analysis of the macroeconomies of the member states. A gap in the literature for ECOWAS member states was filled by estimating an empirical model which is consistent with an augmented Solow growth model that includes in its specification the ratio of R&D to GDP and also human capital [

The main objective of our study was to find out whether R&D influenced economic growth in ECOWAS member states. An auxiliary objective was to find out whether stock of knowledge had an influence on R&D. Since the analysis dealt with various countries, attempts were made in order to ensure cointegration existed in the long run relationship of the variables. A unit root analysis was conducted to check the presence of panel unit root in the variables. Based on the outcome of the unit root analysis the cointegration test was be conducted if there existed unit root in the variables to evaluate the long-term relationship between the variables. Then the Generalized Least Squares (GLS) method with dynamic panel data estimators was used to find the relationship between the variables. The GLS was used because it more robust in curbing certain degree of correlation between the residuals in a regression model. Therefore the GLS is preferred over Ordinary Least Squares (OLS) because it can be statistically inefficient, or even give misleading inferences

It is agreed by African head of states and some scholars that for the development of the African continent, R&D must be taken seriously. This implies that the development of the African continent requires the metamorphosing from the reliance on natural resources to a value and knowledge-based economy. The over reliance on natural resources does not promote economic growth but rather a stagnant development. The developed nations have moved pass this stage only through innovation and advancement in technology [

In July 1979 and April 1980, the Monrovia Strategy and the Lagos Plan of Action (LPA) respectively were espoused by the heads of states of Africa (1980-2000). Their intentions for assumption of these action plans were to enhance the development of economic growth of African countries. The LPA was the first of its kind to be proposed on how the African continent learnt to independently cooperate on issues that bordered on the growth of the continent. This however, laid the grounds for successive African development agendas.

The LPA inspired the following gatherings and enactments. First, the CASTAFRICAII conference was organised by UNESCO/OAU/ECA, where economic recovery tactics for Africa were looked at. It was well attended by 26 Science, Technology and Innovation (STI) ministers from African countries and experts. In addition, the Abuja Treaty was also embraced in 1994. Its aim was to give birth to an African Economic Community (AEC) which ensured a common gain for the continent through economic assimilation. Similarly, in July 2001 the Organisation of African Unity (OAU) was transformed to African Unity (AU) in Lusaka, Zambia with the aim to “build an integrated, prosperous and peaceful Africa, an Africa driven and managed by its own citizens and representing a dynamic force in the international arena” [

^{4}Such calls for increased investment in R&D included the Monrovia Declaration of 1979, the Eighth Ordinary Session of the Executive Council of the AU that met in 2006 in Khartoum, Sudan and the Ninth Executive Council of the AU held in Addis Ababa, Ethiopia in 2007 [

Knowing the prospects emanating from STI, several calls^{3} have been made to increase the 1% investment of each African country’s GDP towards R&D enhancement. In light of this, a conference of Ministers in charge of Science and Technology (AMCOST) was instituted by the AUC. Its purpose was enable the ministers have a unanimous say on matters bordering on STI. In 2005, the (AMCOST) developed a Consolidated Plan of Action (CPA) for science and technology in Africa. Its focus was for the action plan to serve as a guide in determining the total expenditure towards R&D in each African country. In September 2007, the CPA gave birth to the African Science, Technology and Innovation Indicators Initiative (ASTII). During the creation of ASTII, the Intergovernmental Committee on ASTII meeting in Maputo resolved that African countries should apply internationally recognised mechanisms and guidelines to assess R&D and innovation programmes. The Organisation for Economic Development’s (OECD) Frascati and the OECD/Eurostat Oslo Manuals were recommended as key points of reference in conducting surveys and developing standard indicators of STI in Africa [

The model adopted for the purpose of this work was the standard growth equation model. It explained a country’s aggregate production function as a Cobb- Douglas type as in [

Y = A K a 1 L a 2 R D a 3 H C a 4 (1)

where Y is the real output, K denotes aggregate capital input, L is the total labour input, HC represents the human capital stock and R&D represents the research and development stock. Also, A denotes a positive constant. To add, α_{1}, α_{2}, α_{3} and α_{4} constitute exponents (positive fractions) of K, L, RD and HC respectively.

We start by taking the natural log of both sides of Equation (1) we find that:

ln Y = A + a 1 ln K + a 2 ln L + a 3 + ln R D + a 4 H C (2)

The availability of data in the West African sub-region on the independent variables was challenging. Therefore, proxy variables were used in replace of the independent variables.

These proxies are gross fixed capital formation (gfcf), GDP per person employed (gdpppe), government expenditure on tertiary education as a % of GDP (geote) and gross enrolment ratio, both sexes % (ger) for capital (K), labour (L), R&D (RD) and human capital (HC) respectively.

From the presentation of the theoretical framework in the preceding section, we can proceed by applying econometric tools to our Cobb-Douglas production function. That is by obtaining the estimates of the parameters A, a_{1}, a_{2}, a_{3} and a_{4}. The implication of the production function is that the output level Y depends upon K, L, RD and HC. A regression analysis is done on the transformed logarithm of the production function. The regression would provide us with estimates of a_{1}, a_{2}, a_{3} and a_{4}. The estimates of a_{1}, a_{2}, a_{3} and a_{4} could be used to determine phenomenon of returns to scale. If sum of these estimates is equal to one, then it implies constant returns to scale; if it is less than one, then it is decreasing returns to scale and if it is greater than one, it is increasing returns to scale. The production function has an implicit assumption that A (level of technology) is fixed over time. This however is not applicable in the real world therefore A needs to be varied. This we do by differentiating the transformed logarithm of the production function.

ln ( Y ) = 1.49013 − 0.0039 ln ( gfcf ) + 0.25095 ln ( gdpppe ) + 0.05031 ln ( ger ) + 0.11327 ln ( goete ) ( 0. 237 ) * * * ( 0.0 75 ) * * * ( 0.0 44 ) * * * (3)

Variables | Obs | Mean | Std. Dev. | Min | Max |
---|---|---|---|---|---|

goete | 203 | 0.7397634 | 0.4141411 | 0.0002 | 1.92761 |

Ger | 200 | 4.6505989 | 4.903917 | 0.00211 | 22.98218 |

Gdpppe | 125 | 1.32958 | 1.777474 | −4.66128 | 7.874914 |

Gfcf | 132 | 17.9934 | 9.758108 | 2.075241 | 46.44408 |

Gdpg | 210 | 4.605989 | 4.903917 | −30.14522 | 33.73577 |

_{2} indicates that a 1 percent increase in the employment level leads to a 0.251 percent increase in the real GDP level. Similarly, the 0.11327 estimate for α_{4} indicates that a 1 percent increase in the R&D level leads to a 0.113 percent increase in the GDP level.

In the case of Equation (1), it is possible to determine the level of change (either increase or decrease) on GDP to a corresponding change on all the input factors. Therefore from (1) we can get:

Variables | Description | Source |
---|---|---|

lngdpg_{i}_{,t} | Natural logarithm of GDP per capita growth (annual %) | Authors’ calculation based on World Bank―world development indicators |

lngfcf_{i}_{,t} | Natural logarithm of Gross fixed capital formation (% of GDP) | Authors’ calculation based on World Bank―world development indicators |

gdpppe_{i}_{,t} | Natural logarithm of GDP per person employed (constant 2011 PPP $) | World Bank―world development indicators |

lnger_{i}_{,t} | Natural logarithm of Gross enrolment ratio, tertiary, both sexes (%) | Authors’ calculation based on World Bank―education statistics |

lngoete_{i}_{,t} | Natural logarithm of Gross expenditure on tertiary education (% of GDP) | Authors’ calculation based on World Bank―world development indicators |

lngdpg | Coef. | P > t | [95% Conf. | Interval] |
---|---|---|---|---|

lngfcf | −0.003894 | 0.963 | −0.1687577 | 0.1609697 |

(0.082904) | ||||

lngdpppe | 0.2509485*** | 0.001 | 0.1016541 | 0.4002429 |

(0.0750748) | ||||

lnger | 0.05031 | 0.123 | −0.0139428 | 0.1145628 |

(0.0323104) | ||||

lngoete | 0.1132657*** | 0.011 | 0.0265207 | 0.2000106 |

(0.0436209) | ||||

_cons | 1.490128*** | 0 | 1.015632 | 1.964624 |

(0.2386068) | ||||

Number of obs = | 89 | |||

F( 4, 84) | 8.06 | |||

Prob> F | 0 | |||

R-squared | 0.2773 |

Notes: 1) The figures in the parenthesis are the standard errors. 2) *** denotes statistical significance at 5% level.

C ( y , r , w , p , q ) = h ⋅ y y 1 ϕ ⋅ r a 1 ϕ ⋅ w a 2 ϕ ⋅ p a 3 ϕ ⋅ q a 4 ϕ

where ϕ = a 1 + a 2 + a 3 + a 4 is the returns to scale parameter and

h = ϕ [ a 1 a 1 ⋅ a 2 a 2 ⋅ a 3 a 3 ⋅ a 4 a 4 ] − 1 ϕ

Taking the log of this equation and adding an error term (ԑ_{i}) yields the equation estimated by [

ln C i = β 0 + β y ⋅ ln ( y i ) + β K ⋅ ln ( K i ) + β L ⋅ ln ( L i ) + β R D ⋅ ln ( R D i ) + β H C ⋅ ln ( H C I ) + ε i (4)

In this model we know that if returns to scale are going to depend on the value of φ = a 1 + a 2 + a 3 + a 4 .

Then:

If φ = 1, the production function has constant returns to scale.

If φ > 1, the production function has increasing returns to scale.

If φ < 1, the production function has decreasing returns to scale.

To find out the level of GDP that is being influenced with a corresponding input factor, it can be seen from _{2} implies there is a decreasing return to labour with respect to GDP. That is the 0.25095 estimate for β indicates that a 1 percent increase in employment leads to a 0.251 percent increase in the GDP level. Similarly, the estimates for a_{4} imply there is decreasing returns to R&D with respect to GDP. That is the 0.11327 estimate for γ indicates that a 1 percent increase in R&D leads to a 0.113 percent increase in the GDP level.

Similarly, the sum of a 1 + a 2 + a 3 + a 4 = 0.41063 . The outcome of 0.41063 is however less than one, which denotes that GDP shows characteristics of “decreasing returns to scale.” A decreasing return to scale implies that a percentage increase in all inputs leads to a less than percentage decrease in the GDP. Therefore, doubling the inputs would mean that the outcome of the GDP will also be doubled but still not more than one. That is, a 1 + a 2 + a 3 + a 4 = 41.063 which shows that GDP had doubled as a result of a hundred percent increase in the inputs. The results still indicate decreasing returns to scale since the outcome of GDP is less than the amount of increase in inputs.

From the production function in Equation (1) it is observed that there is an implicit assumption that A is fixed over time (or has a positive constant) which is assumed to be the level of technology. On the contrary, technology is expected to be dynamic and therefore does not reflect happenings in the real world. For technology to be assumed to be dynamic, Equation (2) will be differentiated with respect to time and while assuming time is dependent. Doing so, we obtain

1 Y d Y d t = 1 A d A d t + α 1 K d K d t + β 1 L d L d t + γ 1 R D d R D d t + ρ 1 H C d H C d t (5)

where 1 Y d Y d t denotes the growth rate of Y, we can define a new variable g Y

for the growth rate of Y. Similarly, we can define the variables g A , g K , g L , g R D and g H C for the growth rates of technology, capital, labour, R&D and human capital and we can rewrite (5) as

g y = g A + α g K + β g L + γ g R D + ρ g H C (6)

Just as in the case of Equations (1) where the level of technology was assumed constant, so will the growth rate of technology be in model (5) while the level of technology changes. The growth rate of GDP on that of the growth rates of the input factors. Doing so, we obtain

g Y = − 0.0150057 − 0.2249671 g K + 0.0157849 g L + 0.1267035 g H C + 0.1453109 g R D ( 0. 1267 0 35 ) * * * ( 0.0 392197 ) * * * (7)

Equation (7) shows that the average growth rate for technology (average of the variable g A ) is −0.0150057 = −0.015% per year. The coefficients of a_{3} = 0.1267035 and a_{4} = 0.1453109 are the assumed elasticities for HC and R&D respectively. These coefficients imply that a 1 percent increase in HC and R&D will lead to a 0.127 and 0.145 percent increase in GDP respectively. This outcome also indicates a decreasing return to scale. Summing the coefficients of a 1 + a 2 + a 3 + a 4 = 0.0628322 . Comparing the outcomes of GDP when technology was constant and when it was varied, the initial outcome had a higher GDP than the former. This however suggests that the West African sub-region is reluctant to embrace technological change and therefore will prefer to invest in other sectors.

lngdpg | Coef. | P > t | 95% Conf. | Interval |
---|---|---|---|---|

lngdfcfD1 | −0.2249671 | 0.323 | −0.6767771 | 0.226843 |

(0.2257112) | ||||

lngdpppeD1 | 0.0157849 | 0.904 | −0.2438433 | 0.275413 |

(0.1297027) | ||||

lngerD1 | 0.1267035*** | 0.007 | 0.0356188 | 0.2177882 |

(0.0455033) | ||||

lngoeteD1 | 0.1453109*** | 0.000 | 0.0668042 | 0.2238176 |

(0.0392197) | ||||

_cons | −0.0150057 | 0.84 | −0.1631513 | 0.1331399 |

(0.0740092) | ||||

Number of obs = | 63 | |||

F( 4, 58) | 7.26 | |||

Prob> F | 0.0001 | |||

R-squared | 0.3337 |

Notes:1) The figures in the parenthesis are the standard errors. 2) *** denotes statistical significance at 5% level.

The purpose of this paper was to study the influence of R&D (which included capital, labour and human capital) on economic development in the West African sub-region. However, due to unavailability of data to cover these variables, proxies were used to represent the independent variables. These proxies are gross fixed capital formation, gross domestic product per person employed, government expenditure on tertiary education and gross enrolment ratio (both sexes) to represent capital, labour, R&D and human capital respectively. The West African sub-region is made of 15 countries^{3}. The study covered a period of 2001-2014 while the GLS method of estimation was adopted with a panel dataset. The findings first of all revealed that the independent variables (i.e. gross domestic product per person employed and government expenditure on tertiary education) had a significant influence on the independent variable (per capita) with the level of technology being constant. The level of significance was however weakly positive. The results also indicated that there was decreasing returns to scales on the GDP. That is, a proportionate increase in all inputs leads to less than proportionate decrease in the GDP. This therefore, meant that the amount of resources invested in the input did not yield an equal or more GDP but rather less. However, operating in a real world would mean that technology could not be constant. Therefore, it was varied in order to observe the growth rates of the inputs and GDP. The findings revealed that both R&D and HC had a positive influence on GDP (per capita). It was again noticed that they both had a decreasing returns to scale on GDP.

^{4}Benin, Burkina Faso, Cape Verde, Cote d’Ivoire, The Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Niger, Nigeria, Senegal, Sierra Leon and Togo.

It can however be said that the member states of ECOWAS continuous investment in the R&D and the development of the HC sectors especially in education would encourage innovation, novelty and creativity resulting in competition and efficiency thereby leading to higher production. The findings are in line with the endogenous growth literature of Romer [

Most governments over the past years particularly in the developed economies have acknowledged the significance of R&D to economic growth. They therefore have invested more large amounts of money on regular basis to the development of research institutions. However, based on this revelation about the importance of R&D coupled with a strong human resource base on economic development as could be seen in the outcome of the results, it is recommended that West African governments should adopt policies that are appropriate and create conducive operation conditions which promotes science, technology and innovation within each country and possibly within the sub-region. These policies should be geared towards encouraging interest in science and technology research in higher institutions of study and research bodies (laboratories) on one hand and the manufacturing sector on the other hand. It is also recommended that head of states of the West African sub-region should endeavour to be committed to meeting the 1% investment of their various GDP towards the enhancement of R&D as agreed by the African Union Commission (AUC), since most developed nations have moved pass this stage only through innovation and advancement in technology [

By these results, we conclude that government’s expenditure in education is the best proxy for R&D as compared to other variables used in the study.

This work was sponsored by the Qing Lan Project of Jiangsu Province and the Key Members of Outstanding Young Teacher Training Project of Jiangsu University.

Caesar, A.E., Chen, H.B., Udimal, T.B. and Osei-Agyemang,^{ }A. (2018) The Influence of R&D on Economic Development in the West African Sub-Region. Open Journal of Social Sciences, 6, 215-228. https://doi.org/10.4236/jss.2018.63015