Who are Africa's middle class? And will they help to reduce poverty?

A report says a third of Africans are now middle class. Their interests coincide with the interests of the poor and should help to bring about change

African businessmen talk on their mobile phones. Photograph: Per-Anders Pettersson/Getty Images

One in three Africans are middle class. Or are they? A new report from the African Development Bank says: 34%, or 313 million Africans are now middle class (living on $2-$20 a day), after several decades without any change, a jump from 27% in 2000.

A similar report by the Asia Development Bank (pdf) last year found something similar in the data and noted that 56% of people in developing Asia lived at the $2-$20 a day level (all figures purchasing power parity ).

This is all good news. However, is someone living on just over $2 a day "middle class"? Two dollars is the average poverty line for developing countries, meaning you're poor or middle class – but what about in between?

There are various definitions of middle class in what is a burgeoning area of interest for poverty reduction research – stimulated perhaps by ideas about the middle classes as change-makers or catalysts. Think about the middle classes who voted with the poor for Luiz Inácio Lula da Silva as president in Brazil, or the vibrant civil society of India that brings together the poor and middle-class activists (or possibly even the middle classes who are the protesters in the Middle East).

There are at least five reasons why the middle classes (variously defined) might matter for poverty reduction. First, as small business owners their potential to hire employees. Second, their higher disposable income, a share of which could be saved and invested domestically. Third, their higher demand has potential to drive economic growth and attracts private investment. Fourth, their investment in human capital for children potentially leads to higher participation in education and a larger pool of skilled labour. And fifth, they're more likely to hold the government accountable for decisions.

In terms of evidence, development economist William Easterly (pdf) finds empirical evidence linking the middle classes and economic growth, and E Sridharan, from the University of Pennsylvania, outlines the role of the Indian middle class in promoting reform. The World Bank's Martin Ravallion (pdf) also provides evidence that empirically links a middle class to growth and poverty reduction. He notes: "Countries starting out with a small middle class judged by developing country rather than western standards [$2-$13 a day per capita] face a handicap in promoting growth and poverty reduction, though this too is largely accountable to differences in the incidence of poverty."

This means – taking this definition of poor and middle class (there is no in-between) – at any given mean consumption, the countries with lower poverty will have a large middle class and see higher subsequent rates of both growth and poverty reduction.

In the middle-class work there are a range of definitions: some relative – the middle class as the middle 60% of consumption expenditure; and some absolute – the middle classes as those living on a range of scales between $2-$100 a day, (a global middle class).

More fundamentally, any definition of the middle class needs to be clear about the extent to which the middle class "starts" at the point where poverty "ends", or whether there is an implicit distance between the characteristics of the poor and the middle class that would suggest a rather clearer demarcation. But at what point does saving start? At what point does consumption of non-essentials, such as mobile phones, start?

MIT economists Abhijit Banerjee and Esther Duflo point out, the middle classes are: likely to be less connected to agriculture; more likely to be engaged in small business activities; and benefit from formal sector employment, with a weekly or monthly salary, which enables them to adopt a longer-term perspective towards their finances.

However, in contrast to the rather idealised image of the middle class as "risk-taking entrepreneurs", Banerjee and Duflo highlight empirical evidence showing that many businesses operate at low profits or fail to experience significant growth.

What about the idea of a catalytic class? This is the idea: the catalytic class is a group, whose expansion triggers internally driven, self-sustaining, political and economic change, a group whose exertion of pressure for better governance and economic reform leads to change when the class hits a certain size (in population or economic or tax revenue size). The interests of this class coincide with the interests of the poor.

As noted in the Globe and Mail, the catalyst class (a) has an interest in accountability because they pay more taxes; (b) probably don't work for the state and thus don't see their loyalty and interests tied to the status quo; (c) have parents who led quite different consumption lifestyles to them; (d) probably have internet (cafe) access and cell phones; and (e) want "open business conditions, fair and honourable contracts, and a route to employment unclotted by corruption".

That sounds more like the middle class to me.