Life Expectancy in Africa

?

Abstract

I do an econometric study of life expectancy in forty African countries with an aim to provide a comparative tool to governments and policy makers who strive to increase the average life expectancy of African countries. Only one "best" regression equation was picked among others as the most accurate linear relationship between life expectancy and its determinants.

Introduction

Life expectancy is a very hard variable to explain. Countries may have different life

expectancies no matter how identical they may be in terms of living conditions. Some

economists have tried to do regression analysis studying what influences life expectancy in

different countries. Most of the times though, they focus on Western and Asian countries. Few

researchers have done a study on African countries. It is in this perspective that, in this paper, we

will do an econometric study of life expectancies in 40 African countries. The independent

variables are: income per capita, HIV/AIDS infection percentage, literacy level, political stability

and the geographic location of the country referring to the Saharan dessert. Econometric as a

study of economic measurement can provide regression analysis as a tool to establish a

relationship between life expectancy and its respective determinants.

First and foremost it is important that we explain why the study focuses exclusively on

African countries. As observed by Gordon Anderson, in Life expectancy and Economic Welfare:

The Example of Africa in the 1990s, "there is no greater continental life expectancy variation

than in Africa where systematic health characteristics differences and major civil wars appear to

distinguish declining and non-declining life expectancy groups" (Anderson 456). Anderson's

study reveals the reason why life expectancy discrepancies among African countries are unique:

constant wars and worse health conditions compared to other continents. The scope of the

present study is knowing the intensity life expectancy factors have on it; it is not scientifically

enough to know the factors without knowing how important they are. Furthermore, it is also

important to make sure that the factors we think contribute to increasing or decreasing life

expectancy are significant.

On the other side, life expectancy of a population may serve as a measure of economic

forecasting. It can be assumed that the longer people live, the more productive to their country

they can be. Knowing the quantitative impact of factors that influence life expectancy is the first

step towards rising it and an important step in economic planning. In addition, knowing the

relationship between the life expectancy of a country and its income per capita, political stability,

HIV/AIDS infection percentage and literacy level is important. It can be an incentive for

governments to know which area to focus on in order for its people to live longer. For example, a

government may be thinking of which area to spend more money on and have difficulties

identifying which areas affect life expectancy the most. The equation that will be generated by

this study will be one of the best answers in a situation like this. Generally, the equation will

quantitatively show what happens to life expectancy of a country when one of the independent

variables that influences it increases by one unit, all other things equal.

This project intends to evaluate the relationship between life expectancy and variables

that influence it. Observations will be 40 countries in Africa. Only data that was released in 2011

will be considered. It is a cross section study. Income per capita and HIV/AIDS infection

percentages data were found on CIA World Factbook. Literacy level was taken from UNESCO

2011 data. Political stability, which is a dummy variable, is be established based on a list of 20

most dangerous countries (or failed states) by Fund for Peace. Before moving on, it is important

to look at their studies for both reference criticism purposes.

Literature Review

In 2001, at the World Bank, Desmond McCarthy and Holger Wolf published a study on life

expectancy in African countries, they entitled it: Comparative Life Expectancy in Africa. At the

beginning of their study, both authors hypothesizes that poverty is not an important factor

in determining life expectancy. They base this understanding on the fact that some countries

with low income per capita have higher life expectancies. "Countries with very low per capita

income, such as Tanzania, boast life expectancy comparable with those in much richer countries"

(McCarthy 7). In our data, Tanzania has a higher life expectancy than Congo (also known as

Congo Brazzaville) which is six times richer. Such an example and others of the same kind

pushed McCarthy and his colleague Wolf to do a regression analysis in their study. They doubted

that income was significantly correlated to life expectancy. However, the results of their

regression analysis rejected their null hypothesis. There was a strong evidence that life

expectancy and income are positively correlated; the initial assumption that life expectancy and

income per capita are not related was wrong. They concluded that "the primary determinant of

life expectancy is income per capita" (McCarthy 13). But they added that the relationship is not

linear (McCarthy 5). As it can be observed on Graph I in the appendix, this assumption is true.

Desmond McCarthy and Holger Wolf suggest that when income per capita goes up, life

expectancy goes up as well. However, this suggestion was doubted by some economists,

including Acemoglu and Johnson.

Daron Acemoglu and Simon Johnson, two scholars from Massachusetts Institute of

Technology, in a study they entitled: Disease and Development: The Effect of Life Expectancy

on Economic Growth, question whether it is income that increases life expectancy or the other

way around. The argument is that the longer people can live, the more money they can make,

thus increasing their income per capita. At the beginning of their study, both scholars explicitly

express the doubt: "while a range of micro studies demonstrate the importance of health for

individual productivity, as discussed below, these studies do not resolve the question of whether

health differences are at the root of the large income differences we observe today and whether

improvements in health will increase economic growth substantially" (Acemoglu 1).

From this argument emerges other problems about linking income to life expectancy.

Obviously, a person does not spend all the money he makes on just himself. Some personal

income is spent on other people and on things that have nothing to do with one's biological and

psychological well-being (healthcare