Cluster analysis of Nigerian real GDP

Abstract

Background: The Central Bank of Nigeria has itemized Nigerian real Gross Domestic Product (GDP) into the following: crop production, livestock, forestry, fishing, crude petroleum and natural gas, solid minerals, manufacturing, construction, trade, transport, information and communication, utilities, accommodation and food services, finance and insurance, real estate, professional, scientific and technical services, administrative, support and business services, public administration, education, human health and  social services, arts, entertainment and recreation and other services.

Objective:  This paper is an attempt to break Nigerian real GDP into clusters on the basis of their contributions in the year 2018 and the first three quarters of 2019.

Methodology: The researcher made use of  a combination of two distances: (i) the Pearson’s (correlation coefficient) distance and (ii) the absolute correlation coefficient distance to assist into breaking Nigeria’s GDP into clusters. Additionally,  two linkage methods were applied. They are: (i) The complete linkage method and (ii) the centroid linkage method.  The combination of the correlation coefficient distance and the centroid linkage method was chosen because it isolated the mainstay of the Nigerian economy namely, crude oil into a cluster of its own.

Results: The three clusters which emerged were: Cluster 1: crop production, livestock, forestry, solid minerals, manufacturing, trade, transportation, utilities, real estate, professional, scientific and technical services, administrative, support and business services, public administration, education, human health and social services. Cluster 2: fishing, construction, information and communication, finance, arts, entertainment and recreation, other services. Cluster 3: crude oil

Unique contribution:  The researcher suggested the most appropriate method of classifying Nigerian GDP by clustering of the mainstay of the Nigerian economy, crude oil, into a cluster of its own. 

Conclusion: A   combination of Pearson’s (correlation coefficient) distance and the centroid linkage method are the most appropriate method for examining Nigerian GDP because it results in the clustering of crude oil into a cluster of its own. 

Key recommendation: We recommend that the Central Bank of Nigeria should categorize Nigeria’s GDP into clusters on the basis of importance to the Nigerian economy.