Tag Archives: model-based cluster analysis

Africa’s Mobile Commerce Segments

M-Pesa Kiosk An article posted earlier this week acknowledges Africa as the global leader in the adoption of mobile commerce (m-commerce). Published on 9 November 2011 “Africa Leads World in Mobile Commerce” highlights the rapid diffusion and adoption of m-commerce, in general, and mobile payments systems, in particular, within the continent.

This article reinforces something that my colleagues and I have known for a while. Africa is using mobile technology to leapfrog the older, more capital intensive, communications infrastructures in use in developed countries. In a recent publication, Dr. Godwin Ariguzo (@ariguzo) and I develop segments for the diffusion of mobile commerce in Africa. With data obtained from the International Telecommunications Union (ITU)  and the 2011 CIA World Fact Book,  we use model-based cluster analysis to identify groupings of countries who are similar given three variables related to m-commerce: Internet Users per 100 People (IUP100 from ITU), Mobile Cellphones per 100 People (MCP100 from ITU) and Gross Domestic Product per Capita (GDPC from CIA World Fact Book).

The first two variables serve as surrogate indicators for the adoption of new(er) communications technologies within a country and the potential for accessing the Internet via mobile phones. The third variable, gross domestic product per capita (GDPC), is used as a surrogate indicator for the standard of living within the country. Not without its critics (and rightfully so, in our opinion), GDPC remains one of the most widely used indicators of the level of development an economy has achieved. Complete 2009 year-end data (data for all three variables investigated) is available for African 53 countries.

The results of the model-based cluster analysis suggest that three distinct m-commerce segments exist within Africa. The number of countries within each segments ranges from 15 to 21 (15, 17 and 21). Countries included in the three segments, along with the average segment scores for each variable investigated, are presented below.

African Mobile Commerce Segements and Segment Means

Based on the results, we propose that the 15 countries in the third segment offer the best opportunity for the roll-out of m-commerce within the African continent. Interestingly, the four North African countries (Algeria, Egypt, Libya and Tunisia) at the center of the political revolution in the Middle East and North Africa are contained within this cluster. The impact of mobile phones and mobile phone technologies in fostering and fueling these revolutions is well established.

The graph below presents the segments in three-dimensional space. Clearly, the three segments are distinct with little overlap. Just as interesting, the graphical presentation allows one to view the impact that each variable has in determining segment membership.

African Mobile Commerce Segements in Three Dimensional Space

In summary, those seeking to launch m-commerce endeavors in Africa should consider implementing them in the third segment (15 countries) first, followed by the first segment (17 countries) and then segment two (21 countries). The full article is available for download from the Social Science Research Network (SSRN).

Africa is a leader in the adoption and implementation of m-commerce. How long will it take for the rest of the world to catch up?

Citation:

Ariguzo, Godwin C. and D. Steven White (2011), “Africa’s Mobile Commerce Segments: A Model-Based Cluster Analysis”, Review of Business Research, Vol. 11, No. 4, pp. 38-44.

Share

Global Clusters Based on Energy Consumption, Carbon Dioxide Emissions and Paper Consumption Per Capita

Oil RigIn a recent journal article co-authored with Adam Sulkowski, we use model-based cluster analysis to examine 121 countries across three measures of environmental efficiency. The three measures – energy consumption (barrels of oil equivalent) per capita, emissions of CO2 (metric tons) per capita and per capita paper/paper products consumption (kilograms per person) – are all related to a country’s comparative resource usage efficiency and contribution to climate change, a primary global concern. Using the most recent complete data available from the World Resources Institute, and the R statistical software program, we find six distinct clusters (or groupings) of countries based on similarities in per capita resource use within the clusters and differences (known as distance) between the clusters.

We developed names for the resulting clusters based on the cluster centroids and the countries contained within each.

Developing countries:

The average per capita of energy consumption, CO2 emissions and paper product consumption in this cluster, as compared to Cluster 6, is 96, 98 and 95% less, respectively. More than half (14 out of 27) of the countries contained in this cluster are identified by the United Nations as Least Developed Countries. The countries contained in this cluster include Angola, Bangladesh, Benin, Cameroon, Congo, Democratic Republic of the Congo, Côte d’Ivoire, Eritrea, Ethiopia, Ghana, Haiti, Kenya, Morocco, Mozambique, Nepal, Nigeria, Pakistan, Paraguay, Peru, Senegal, Sudan, Tanzania, Togo, Vietnam, Yemen, Zambia and Zimbabwe.

Rapidly growing energy consuming countries:

The average per capita of energy consumption, CO2 emissions and paper product consumption in this cluster, as compared to the average consumption per capita of the members of Cluster 6, is: 85, 80 and 57% less. Of the 12 countries represented, the majority is from the Americas and Caribbean. The countries contained in this cluster are: Argentina, Azerbaijan, Brazil, Chile, Costa Rica, Iran, Jamaica, Mexico, Romania, Thailand, Uzbekistan and Venezuela. It is interesting to note that half (six of the twelve) are oil-producing states (Argentina, Brazil, Chile, Iran, Mexico and Venezuela).

Advanced developing countries:

In this cluster, the member countries consume on average 93% less energy, generate 99% less CO2 and use 79% less paper per capita than do the consumers of Cluster 6. Most of the countries in this cluster are considered to be emerging markets and/or contain moderate levels of income per capita. This cluster consists of the following countries: Albania, Algeria, Armenia, Bolivia, Botswana, China, Columbia, Dominican Republic, Ecuador, Egypt, El Salvador, Georgia, Guatemala, Honduras, India, Indonesia, Jordan, Kyrgyzstan, Macedonia, Moldova, Namibia, Nicaragua, Panama, Philippines, Sri Lanka, Syria, Tajikistan, Tunisia, Turkey and Uruguay.

Middle energy paper consuming countries:

The members of this cluster consume an average of 46% more paper per capita than do the members of Cluster 6, while consuming an average of 72% less energy per capita and expelling an average of almost 59% less CO2 per capita. It is interesting to note that 19 out of the 30 countries in this cluster are members of the European Union. Just as interesting is the membership of some developing countries within this cluster. The countries are Austria, Belarus, Bulgaria, Croatia, Cyprus, Denmark, Gabon, Germany, Greece, Hungary, Ireland, Israel, Italy, Japan, Korea, Latvia, Lebanon, Lithuania, Malaysia, Malta, Netherlands, New Zealand, Poland, Portugal, Slovakia, Slovenia, South Africa, Spain, Ukraine and the United Kingdom.

High energy tree destroying countries:

This cluster is characterized by the highest per capita consumption of paper products, with its members consuming an average of almost two times more paper products per capita than Cluster 6. While member countries consume an average of 45% less energy per capita and expel an average of 37% less CO2 per capita than do the countries contained in Cluster 6, they are the second least energy efficient cluster. Contained within this cluster are countries with the reputation as having some of the most green-oriented consumers and policies on the planet. The countries in this cluster include Australia, Belgium, Canada, Czech Republic, Estonia, Finland, France, Kazakhstan, Kuwait, Luxembourg, Norway, Oman, Russia, Saudi Arabia, Singapore, Sweden, Switzerland, Trinidad and Tobago and the USA.

Extreme energy usage and CO2 producing countries:

This cluster leads the world in average per capita energy consumption and average CO2 emissions per capita. This is the smallest cluster containing only Bahrain, Iceland and the United Arab Emirates. It seems fair to label these countries as the extreme in their energy use because of their relatively large average per capita consumption of energy: an average of almost 82% more than the second least energy efficient cluster, over 257% more than the third least energy efficient cluster and almost 567% more than the fourth least energy efficient cluster. These countries also produce extreme levels of CO2 per capita as compared to the other clusters (35% more than Cluster 5, 142% more than Cluster 4 and 393% more than Cluster 3). Furthermore, many of the countries in Cluster 4 and Cluster 5 have GDPs per capita and/or average standards of living that are higher than those found in the countries contained within this cluster.

White, D. Steven and Adam J. Sulkowski (2010)

By examining per capita consumption of resources, green marketers gain useful information on which to base their strategies. Likewise, the information may assist governments, public policy makers and private enterprises in their efforts to stimulate sustainable businesses and business practices. Thus, rather than being seen as a threat, measures of per capita consumption should be viewed as an opportunity – an opportunity to foster the development of new products or services designed to minimize resource use per capita while retaining or increasing standards of living globally.

Reference:

White, D. Steven and Adam J. Sulkowski (2010), “Relative Ecological Footprints Based on Resource Usage Efficiency per Capita: Macro-level Segmentation of 121 Countries”, International Journal of Sustainable Economy, Vol. 2, No. 2, pp. 224-240.

Share