Join the Conversation log-in

Thursday, August 13th, 2009 - 4 comments

Poverty is one part of HIV risk, but not the most important part

It has been widely assumed for many years that poverty and underdevelopment were driving factors behind the transmission of HIV, but empirical evidence does not appear to bear this out.  For example:

Country-level evidence

Viewed globally, the highest prevalence of HIV is found in the poorer countries of the world, with more than two thirds of people living with HIV residing in sub Saharan Africa.

However, within sub Saharan Africa, the highest rates of HIV prevalence are found in the relatively more wealthy countries, such as Swaziland, Botswana and South Africa, and in general, the countries with higher GDP per capita have higher levels of HIV prevalence – although the country level correlation is not statistically significant.

Direct measures of poverty, such as the percentage below $1 per day are not correlated between countries.  However, most of the other commonly understood poverty measures such as literacy rates, educational enrolment, under-five mortality or nutritional outcomes seem to show an inverse relationship at country level between poverty and HIV – the poorer countries (or countries with worse development indicators) have lower rates of HIV prevalence.

Household level evidence

At household level, strong evidence from population-based surveys (DHS+) has shown that HIV prevalence is lower among the poorer members of African society, and higher among those who are better off, and those living in urban areas.  Note that the associations are stronger and clearer for women – i.e. women living in better-off households have clearly higher HIV prevalence in most African countries.

Therefore, it is not correct to say that HIV risk is primarily associated with severe poverty (which is also concentrated in rural areas), although the poor are also at risk.  What is undoubtedly true is that the poor are the most severely affected once they do become infected with HIV, since they are unable to afford treatment, and can least afford to lose their incomes.

Evidence linking income inequality and HIV risk

There is a striking association between income inequality (measured by the GINI coefficient) and HIV – countries with higher income inequality appear to have higher HIV prevalence rates.  Similar patterns (slightly less clear) can be seen with other inequality measures such as the decile or quintile ratio.

There may be many confounding factors in this apparent correlation, such as differences between countries in cultural practice (e.g. male circumcision and religion).  Nevertheless, it is tempting to conclude that income inequality may give rise to social conditions that fuel the HIV epidemic – for example by providing incentives for transactional sex.  However, the evidence is not yet strong enough for us to be able to say this, since the association that we observe may be a result of other underlying factors that affect both income inequality and HIV risk.  The association is interesting, but we cannot yet say that it is causal, and it has not been borne out by household-level evidence.


Multivariate analysis of DHS+ data has shown that the association between wealth and HIV could be attributed to three main factors that are correlated with wealth, i.e.:

  • More lifetime concurrent partners
  • Urban-residence
  • Age

It has also been shown that education is generally associated with lower HIV risk.  This, and the above analysis suggests that behavioural factors not directly related to poverty are more important drivers of HIV infection in Africa, and has clearly shown that economic status in itself is not a strong predictor of HIV status.

The primary implication is that prevention must therefore cut across all socioeconomic strata of society – especially those living in urban areas.  A purely poverty-based strategy is unlikely to succeed.

There is therefore no simple explanation for the distribution of HIV risk.  Poverty may be a part of the story, but it is clearly not the key.  The pathways and interactions linking socio-economic status and HIV risk are complex, and predisposing factors are different for different groups – i.e. the sources of risk are not necessarily the same for poor people as they are for wealthier people.  It is therefore vital to tailor interventions to the specific drivers of transmission within different groups.

The views expressed in this blog-post are solely those of the author.

Comments (4)

Friday 14th August, 2009, 4:07am


This is a very interesting article and it is quite contradictory to what we hear about poverty patterns and the related incidence of HIV/AIDS in South Africa. Under the Mbeki government the impact of poverty on the spread of HIV/AIDS was highly publicised so this sheds more light on the issue.

I would be interested in reading the research that you use as the basis for your argument - perhaps you can point me to where I can find it.


Robert Greener
Friday 14th August, 2009, 11:16am

Try these two references:

Stuart Gillespie, Robert Greener, Jimmy Whitworth and Alan Whiteside, "Poverty, HIV and AIDS: Vulnerability and Impact in Southern Africa", AIDS, Vol 21 Supplement 7, November 2007

Peter Piot, Robert Greener and Sarah Russell, "Squaring the Circle: AIDS, Poverty and Human Development", PLoS Medicine, Volume 4, Issue 10, October 2007

The first of these contains studies of HIV incidence and socio-economic status in South Africa (based on cohort studies) which find mixed patterns in mostly rural samples. I have not looked closely (i.e. not at household level) at the South African data from HSRC, but the results I have seen do not control for race - this may change the interpretation somewhat in a multivariate analysis. I believe that there is a great deal of interesting analysis still to be done on the South African data, which may indeed yield a different pattern to other African countries.

Tuesday 18th August, 2009, 7:38am

Thanks this looks very interesting and helpful

Twitter Trackbacks for Conversations for a Better World – Poverty is one part of HIV risk, but not the most important part [] on
Saturday 29th August, 2009, 8:34pm

[...] Conversations for a Better World – Poverty is one part of HIV risk, but not the most important par... – view page – cached a shared blog on population, gender and health — From the page [...]

Leave a reply

Name - required


Email - required, never published




Guest Editor

Robert Greener

Senior Economics Advisor, UNAIDS


Robert Greener is the senior economics advisor at UNAIDS, whose responsibilities include identifying priorities for research, intervention, and policy development on the economic aspects of HIV infection, AIDS impact and the effectiveness and consequences of the response to the epidemic. He has been involved in refining the global resource need estimates for a scaled up response to the epidemic, and in recent work on the associations between poverty, income inequality and the risk of HIV infection, and the development of sustainable financing for a long-term response in low and middle-income countries. Prior to joining UNAIDS in September 2003, he spent 7 years as a Senior Research Fellow at the Botswana Institute for Development Policy Analysis (BIDPA), where he conducted work related to rural development, poverty reduction policy, and the economic impacts of AIDS.

Register for Newsletter
Conversation Starters
Tag Cloud
Host a Conversation