Posts Tagged ‘statistics’

Africa’s Middle Class: Development economics and marketing demographics conflating the holy grail

The most developed nation on the African continent, south of the Sahara desert, is considered to be South Africa with its financial and transportation infrastructure and systems, a legacy from history. In the first decade of the 21st century, the black middle class – known as Black Diamonds in marketer jargon – came into prominence on the back of numerous economic initiatives after the fall of apartheid.

img-south-africa-consumer-goods-02The rise of the Black Diamonds was meant to be the signal of a changing rainbow nation, one whose peoples would finally be included in the social and economic advancements long enjoyed by a privileged minority. This emerging middle class was also among the first to be noticed as African consumers in their own right, and their discovery pioneered the subsequent search for the now mythical African middle class. Even then, their total number was under scrutiny for its aspirational inclusivity versus actual households fitting the conventional definition of a middle class. From The Economist writing in 2007:

The University of Cape Town’s Unilever Institute of Strategic Marketing says there are now 2.6m “black diamonds”, as it calls the black middle class, a 30% increase in less than two years. Included in the definition are working professionals; those who own things such as cars, homes or microwave ovens; university students; and those who merely have the potential to enter these categories. The survey estimates that these black diamonds represent 12% of South Africa’s black adults, and make 180 billion rand a year ($26.2 billion), or 28% of the country’s (and more than half of all black South African) buying power.

For some, such as Lawrence Schlemmer, a sociologist in Cape Town, this definition is far too broad to be meaningful. He agrees that numbers are rising fast but argues that they are still tiny. Last year, he says, only 322,000 black South Africans (less than 1% of the black population of 38m) could be deemed “core” middle class, a far cry from 2.6m black diamonds.

Still, whatever their size, the buppies are affecting the economy and the political landscape.

This week, a comprehensive new survey by the South African government shows the on the ground reality in 2016. The National Income Dynamics Study (NIDS)‚ launched by the Department of Planning‚ Monitoring and Evaluation (DPME) in Pretoria surveyed 28‚000 people who were tracked every two years from 2008 to 2015. Very similar in fact to the recent household panel survey completed in India. Even their conclusions resemble each other:

According to the study‚ those in the middle class have a tendency to drop in and out of poverty.

And the size has not actually changed much since 1993 – the year before the fall of apartheid and the election of Nelson Mandela.

The study also shows that the South African middle class is much smaller than estimated‚ sitting at around 14.5% of the total population in 2014. Women are more affected by poverty, and even those who manage to climb the ladder may slip down again.

“…It has not grown much since 1993 — growing its share by only two percentage points in the past 23 years…”

20151024_mac237And, perhaps, the real challenge we face with the ongoing search for Africa’s middle classes is the conflation that took place back then between a consumer marketing segmentation and a socio-political demographic.  By allowing the aspirational reach of the consumer marketing driven research to inflate the size of the segment classified as middle class, it has given rise to an ongoing and complex muddle across teh entire continent. As the AfDB’s former president Donald Kaberuka said last year:

“I think we are wasting too much time on the definition of the middle class and the cut off point, it is a sterile debate.

“A dynamic middle class that rises with the sea increases domestic demand, the diversity of the economy, [its] resilience, and they also stabilise the politics of a country as well, since they have a stake in the system.”

He has a point. But perhaps not the one he intended to make. Instead, if we consider disentangling consumption and demand for consumer products from the increase in political voice and “stake in the system”, we may in fact discover that there is indeed a sizeable bourgeoisie emerging even though they may not possess all the qualifying criteria traditionally attributed to a middle class per se. (Previous posts on this topic have been tagged informal bourgeoisie)

There’s the demographic segment which is the middle, and then, there’s the conceptual body of solid citizens invested in the democratic stability and economic growth and development of their countries. As Jacques Enaudeau wrote in 2013:

But fixated on wealth, the discussion on middle classes in Africa misses out on the other two pillars of social stratification: social status and political power.

As soon as those two are factored in, discussing the “African middle class” as a homogenous entity seems absurd, and so it should. Thinking that what separates the senior civil servant from the street hawker or the country head of a multinational from the shop owner is a matter of daily expenditure amounts to looking at their reality through the wrong end of the telescope: the bigger picture is that they live in different worlds.

In the developing world, the formal sector with its white collar jobs populated by university graduates may jostle cheek by jowl with the informal economy’s life lived on the street but that proximity might be on the only thing they have in common.

For here lies the rub: the material culture that the notion of “middle class” posits as shared consciousness is articulated to a strong sense of individualism, which is borderline contradictory with the idea of class. All the more reasons for the analysis to consider the representations which members have of themselves as a group and the historical context in which such groups are being shaped.

This, however, is not the post to unpack those complexities of self image and collective consciousness. It’s one which pauses to ponder the newest set of findings on the dynamic nature of poverty and wealth in the more uncertain and volatile operating environments of the still developing world. And considers the South African example introduced today:

There has, however, been considerable demographic transformation within that band of the middle class, with Africans now outnumbering whites by about two to one, the report said.  Factors driving the surge include greater access to credit, improved education levels, BEE and improved economic growth until recently.

Transformation of societies is underway, just as the Indian researchers concluded in their analysis. This might be a much larger global trend underway, whose weak signals we’re just beginning to pick up now. I’ll be following up with these musings on the blog. The people with the real problem on their hands are the consumer companies looking to justify entering the African markets, and perhaps that’s a topic to take up in the next article.

Internet penetration by population in African countries: mapping opportunity

Since we’re currently working on a market entry exploration study for Village Telco in Kenya, I’ve been taking a deeper look at the spread and adoption of internet use.  It struck me that the landscape is actually far more fragmented than it used to be – things have been changing so fast that gone are the days where you could look at the situation in one Sub Saharan country and extrapolate it reasonably accurately for many others.  This is particularly true for ICT as cheaper rates and smarter devices impact some locations before diffusing to others. While playing around with the numbers, some interesting observations emerged:

Data Source: www.internetworldstats.com (click for large) Chart: Semacraft Consulting Partners

I sorted the internet usage numbers by size of country – the chart above shows the top 10 countries in Africa by percentage of total population i.e. almost 15% of the continent’s people live in Nigeria, and then added on what percentage of that country’s population was online.

Data source: www.internetworldstats Chart: Semacraft Consulting Partners

The findings were surprising when you compare to this chart where I’ve sorted the countries by percentage of the population accessing the internet. (I’ve removed the French island Reunion  which showed up in 3rd place nudging Nigeria out of the top 5). Their proportion of the continent’s population is seen next to them.

The only countries that fall in the top ten – both by total population and percentage of population on the internet – are Nigeria, Egypt, South Africa and Algeria.  I had started out thinking that if I looked at internet penetration rates by population it would give me some clues about where the internet was being most rapidly adopted (and then perhaps, why). But instead, I found myself surprised by the gaps instead – Tanzania being the unexpected. The reality may be entirely different in Ethiopia and the Democratic Republic of Congo.  Maybe if the data is looked at again separating North African countries from Sub Saharan, a different set of clusters will emerge.  I’d also like to remove all the little island nations to see what happens.

Update:  I decided to take a look at the GDP based on PPP per capita for 2011 (IMF data) for selected countries (based on the earlier two sets) just to see if there were any correlations between that and the internet.

Now this starts to get even more interesting:- Morocco, Nigeria and South Africa show internet adoption figures very different from their relative position in the comparative economy chart. You’d think that greater economic strength would demonstrate a higher internet adoption and vice versa. But South Africa’s internet adoption is  too low compared to its economic standing while Morocco’s is outstanding compared to its economy.  In the East African region, Tanzania is still the internet laggard compared to its neighbors Uganda and Kenya.

Lies, damned lies and statistics: how mythinformation can go viral for mobile data

Mark Kaigwa‘s comment on my previous post exploring the future of cyber cafes in Kenya made me take a deeper look at the data around mobile phones, internet access and of course, extremely affordable Android based smartphones. Perhaps it was not just in the urban areas that cyber cafes were facing a challenge to their business, maybe the rural ones were as well.  Even the photograph I had selected of a small town cybercafe touted “Phones New Arrivals” with the IDEOS first on the list. Given that the Android smartphone can double up as a wifi hotspot, and its sales figures (as far as we knew) were astounding, who knows just how much of a difference they were making?

I’d begun my analysis with the data available on the CCK’s website from their 2010 ICT Survey reports released this month and the Access Gap report released in August 2011.  Then I came across a sentence in Erik’s recent post Android Invasion which struck me forcibly:

I believe we’ll look back at the landing of the IDEOS phone earlier this year in Kenya as an inflection point, where in 2 years we’ll define the times up until then as, “before Android”.

What if all the findings in the CCK’s reports were already obsolete? Hadn’t the survey itself been conducted back in the May and June of 2010, half a year before the IDEOS was launched. After all, didn’t the Access Gap report say:

In summary, although the main place to access internet in Kenya is at cybercafés, cybercafés are mainly used in more developed counties (Nairobi, Mombasa and in eastern counties). In less developed counties (northern, eastern and southern counties) mobile phones and telecommunication centres are the main option to access internet. Between these two alternatives, mobile phones appear to be a more affordable alternative to access internet due to the lower costs of the equipments and lower operational costs (lower electricity consumption and do not require administrative personnel or maintenance). – page 24/25

And mapped usage patterns thus:

So I went looking to see if there was any information out there about the distribution of Android phones in Kenya – after all, a very significant number had been sold in the first half of 2011.  And here’s what I have not found: there are no mentions of sales figures quoting either Safaricom or Huawei officially except for these released on 17 May 2011 (sourced from the Huawei website):

Mr. Herman He, CEO Huawei announced that, “Since the IDEOS launch five months ago, so far over 60,000 pieces have been sold and we are moving towards the 100,000 piece mark with its share of the local smartphone market at 45% in the first quarter of the year, making it the top selling device with February alone reaching 73%.”

Suddenly, one month later on 23rd June 2011, MIT’s Technology review claimed that more 3 times that estimated number were sold:

Smart phones surged in popularity in February after Safaricom, Kenya’s dominant telecom, began offering the cheapest smart phone yet on the market—an Android model called Ideos from the Chinese maker Huawei, which has been making inroads in the developing world. In Kenya, the price, approximately $80, was low enough to win more than 350,000 buyers to date.

This was absolutely amazing to me so I tried to find something more from Huawei or Safaricom – after all, this was an unprecedented jump in numbers, completely beyond Mr Herman He’s wildest imaginings. But no, there was nothing out there  – at least as far as I could find.  All mentions of this sales figure are after June 23rd and those that refer to the source come back to the MIT article as the origin.  The closest alternate I can find is this “confirmed” report that Safaricom’s selling 2000 phones per day.  Given that there were 14 days left in May when Mr He made his estimate, it is possible to reach ‘towards the 100,000 mark’ by adding 28,000 phones more to the 60,000 odd sold = 88,000 and few more.

Then we have 23 days in June for MIT’s magical number of 350,000 – that is 250,000 phones sold in 23 days. Do you think that the global supply chain industry, who face shortages for popular iPhone and Nokia releases would not have made a squeak if suddenly daily sales shot up to 10,000 IDEOS smartphones a day in June?

Anyway, I hope I’m wrong and someone can confirm this number is real but I suspect that the truth may lie closer to it being a typographical error gone viral or the original wording being closer “in the hands of” where the average number of users who share a phone can be around 3 or so in most parts of Africa – that is, around a hundred thousand phones reaching the hands of three times as many Kenyans, a far more realistic assessment of the situation than the mythical sales figure gone viral.

In the meantime, I think I’ll put my secondary research efforts aside for a while and see what happens when we get out in the field and talk to people. Who knows, they might tell us a genuine miracle took place.