Here is a first stab at putting together an index that relates the cost of mobile services to income at the bottom of the pyramid in Africa. I found some ILO data on minimum wage that covers 24 African countries and a I found a couple more by googling. Here are the assumptions that I’ve made so far.

Mobile Costs

In order to relate mobile charges directly to income, I needed data that hadn’t already been converted into USD or similar.  This mean going to the operator’s websites for data.  Obviously there are hundreds of different rates and packages for mobile access.  I started by choosing the operator with the largest market share in the country.  From that operator, I chose one minute of air time during the day i.e. prime time and outside of the mobile network.  I was in some doubt as to whether to choose outside to the fixed line operator or outside to other mobile operators and went with the latter assuming it would represent the widest possible access.  I deliberately chose off-net calls to bring in the issue of interconnect fees.  This represents about the most expensive local call you can make in a country.

Minimum Wage

As I mentioned, the ILO were kind enough to point me at a spreadsheet they maintain which has a fair amount of data on minimum wage in African countries.  They have up-to-date data from 24 countries.  However, as we start to unpack things, there is some complexity here that I haven’t fully processed yet.  For instance, when I looked for data for Kenya (which was absent from the ILO data), I found this article which mentions two minimum wages, one for urban and one for rural agricultural workers.  In South Africa I know there is a specific minimum wage for domestic workers.  I would like to focus on the lowest wage earners but it is challenging to try to pick something that is consistent across countries.  I also thought of nurses wages or a day labourer’s wage.

Another assumption I made was that people work five days a week.  The ILO data gave a monthly wage but I  wanted a daily wage for the index so I divided by the average number of working days in a month (22).  I’m not sure whether this is a fair assumption at the bottom of the pyramid.

Needless to say, I welcome corrections and insights here.

Number of SMSes a days work will earn you at minimum wage

Number of call minutes a days work will earn you at minimum wage

Katrin Verclas and I and a few others have been kicking around the notion of Fair Mobile for some time now.  The essence of Fair Mobile is the idea of developing some metrics for equitable, competitive mobile markets that deliver optimal value for money to mobile users, particularly in developing countries.  It has taken me far too long to get going with this but I am finally finding some steam.  So why bother with this?

Magazines like The Economist have embraced the miracle of mobiles (see their recent special report “Mobile Marvels”) and among development agencies, adding the letter “m” for mobile to existing initiatives e.g. m-health, m-governance, m-learning, etc has become the latest in tech-savvy development.  Unfortunately not enough attention has been paid to two significant downsides to the current status of mobile infrastructure in Africa, namely uncompetitive telecommunications markets and walled garden practices by mobile operators.

1.  Uncompetitive Markets

Mobile markets are dominated by incumbent operators and are typically uncompetitive and overpriced.  Telecommunications regulatory expert Ewan Sutherland neatly summarises the issue:

A very small number of market players, protected by high politico-regulatory barriers to market entry, can easily result in price shadowing and even in collusion. Analyses of the markets for mobile call origination in France, Ireland and Spain have illustrated this problem, despite operators competing on the market for some years. In the case of France there has been shown to be collusion between the three operators, resulting in heavy fines (Conseil de la Concurrence 2005). To date, there has not been detailed analysis of markets in Africa, nor the regulatory action to remedy the lack of effective competition.i”

African countries still lack detailed market analyses that could lead to regulatory action but we know enough to suggest that such analyses are needed if prices are to be brought down.  By any standard, mobile markets in Africa are uncompetitive.  Vodafone and MTN, two of the largest operators in Africa, command on average more than 50% of the market in the nearly 20 African countries in which they operateii.  Safaricom, the largest operator in East Africa, holds more than 80% of the mobile market in Kenya.

Evidence from the pan-African research network, ResearchICTAfrica, points to a remarkably high percentage of income being spent by the poor on mobile services.  For low income earners across 17 countries studied, the average African is paying more than 50% of their disposable income on mobile services.

At the same time, mobile operators are posting impressive profits.  Kenyan operator Safaricom generated over 900 million USD in revenue last year of which a staggering 40% was Earnings before Interest, Taxes, Depreciation, and Amortization (EBITDA).  Other operators are also posting impressive profits with most operators on the continent announcing year on year increases in revenue.

The startling contrast between the remarkable benefits of mobile infrastructure and the high price being paid for mobile services in Africa while mobile operators post record profits leads to the conclusion that more competitive mobile markets in Africa would lead to even greater social and economic benefit for all but especially the poor.

2. Walled Gardens

Most of the discussion of the benefits of mobile infrastructure focuses on the increase in efficiencies that access to communication can provide.  What is not often discussed is the environment for innovation and entrepreneurship that mobile infrastructure can provide.

There have been some significant innovations with mobile services in Africa.  The most significant one is the development of mobile banking and money services such as mPesa in Kenya.  This and other similar innovations are largely top-down from the operator and have been slow to evolve in spite of the obvious demand.  What is missing on the continent is end-user innovation of the sort that leads to serendipitous discovery of social and commercial services.

In Africa, mobiles are too expensive too encourage the kind of experimentation that leads to innovation. Consider that in Rwanda, 4 minutes of local mobile communication or 10 SMSes constitutes an entire day’s wage for a labourer.  While this is expensive, it is obviously useful enough to spend a significant proportion of income on its use for essential calls.  What will never happen in this context is the kind of playful use that happens when one doesn’t have to consider the cost of each call and each SMS.

The cost of SMSes in particular are a barrier to innovation because they represent a gateway to the Internet for every mobile phone no matter how simple.  Because an SMS message carries data, it offers the opportunity to extend the reach of Internet applications whether for data collection, commerce, social networks or other innovative services.  Yet SMS services represents the most profitable aspect of mobile operator networks.  SMS charges are estimated to be between 80 and 90% profit.

However, it is not just cost that is a barrier to innovation.  Mobile operators jealously protect their mobile services and use them to help lock users on to their network.  It is impossible to launch a mobile service across mobile networks without negotiating access with every mobile network involved.

This is in stark contrast to the generative environment of the Internet which has spawned what are now some of the largest companies in the world (Google, Facebook, et al) from the minds of individuals who developed these services without huge corporate backing and in markets that did not exist until they created them.

If we were able to drive down the barriers to mobile voice and SMS use through reduced cost and more Open Access style networks, individual and small-business innovation in the delivery of novel voice and data services would very likely blossom on the continent.

So What Can We Do?

When I initially chatted with Katrin about Fair Mobile, what I had in my mind was an index of how fair mobile markets are in various countries.  Now the ITU already maintain an excellent ICT Development Index but I was thinking of something simpler and more focused, something like the Economist’s Big Mac index which uses the cost of a McDonald’s hamburger as a proxy for purchasing power parity (PPP) but perhaps even closer might be the Alternative Big Mac Index which measures how long one has to labour in a given country to earn the price of a Big Mac.

So what would that look like?  Well, what initially sent me down this path was Nathan Eagle’s description of nurses in Kenya and how they refused to send SMS updates to an online blood bank database simply because the cost of an SMS represented too significant a proportion of their daily wage.  Since then Katrin and I have discussed various indicators and metrics that might best make up a Fair Mobile Index.  Armed with these ideas, I went off a couple of weeks ago to the IDRC’s Acacia Research and Learning Forum 2009 in Dakar and had the opportunity to organise an OpenSpace session on Fair Mobile at the event.  About a dozen people attended the event, including among others, Alison Gillwald, Christoph Stork, and Godfred Frempong of the ResearchICTAfrica network as well as Willie Currie from the Association for Progressive Communications (APC).

We discussed the basic dilemma of the high cost and closed nature of mobile networks and brainstormed a number of factors that affect equity in mobile access, including:

• transparency and market complexity
• number portability
• tariff complexity and transparency
• issue of quality of service:  dropped calls, network coverage, voice quality, etc
• demand perspectives – cultural influences on usage
• taxation issues both on handsets and on usage
• the need to establish mobiles as a generative platform for innovation
• foreign ownership
• interconnection rates

A Simple Ratio

In the end, Alison suggested that it might be simpler to start with something very basic such as the ratio of basic mobile costs to the national minimum wage in a country.  Discussion followed on how to get the average cost of mobile and the suggestion was made to take a minute of peak time use on the largest operator in the country.  In the end, I propose to take 2 minutes of air time and 4 SMSes as the proxy which hopefully represents what might be a typical day’s phone use.

Obviously this simple ratio of mobile usage cost to minimum wage only begins to scratch the surface of Fair Mobile but it seems to me now that it is better to start with something simple and easily understood and refine it over time rather than try to come up with the perfect basket of indicators.

The next thing I think I would like to add is the percentage EBITDA of the largest operator in the country.  This would add a sense of proportion or disproportion between what people are paying and what sort of profit the operators are making.  More complex will be developing a metric around the “thickness” of the walled gardens.

So stand by for more as I start to compile some of this data.  I’m hoping that AfricanSignals will prove an ongoing useful resource on mobile pricing.

Finally, I should emphasise that these are just emerging thoughts on Fair Mobile.  I reserve the right to recant, adapt, and evolve rapidly. 

Blue Gum Tree - Knysna

I see in the press yesterday that Safaricom have won an innovation award for their MPesa service from a UN agency.

The Habitat Business Award for sustainable urbanization, which is organized by the United Nations Human Settlements Programme (UN-HABITAT), aims to recognize and publicize outstanding achievements contributing to sustainable urbanization through responsible corporate practices.

In reading this article, something pushed my buttons. If you happen to be a reader of this blog and have read Nathan and the Mobile Operators or We Need The Eggs or my rant about SMS charges, you may have already decided that I have African mobile operator “issues”.   Perhaps I do.   Certainly I have an issue with the apparent myopia, especially from development agencies, regarding any possible downside to the rapid growth of mobile infrastructure in Africa.  Everyone is happy to celebrate their achievements without paying much attention to the effective monopolies/oligopolies that they have become.  I am at a loss to explain why, with few exceptions, there is not more public comment on what mobile operators are doing to the general communications (and by extension, innovation) ecosystem in African countries.

Let me preface my rant by once again acknowledging the miracle that mobile phones are. There can’t be many people who still doubt the direct value that mobile phones provide to people.  I’ll go further and say that the best is yet to come.  There seems little doubt that there is a coming revolution in mobile-enabled banking and transactions. Yet all is not sweetness and light and that is because the wealth that is being generated by mobile operators in Africa is not being distributed evenly.  Here are a couple of examples:

• In South Africa, Sameer Dave, Chief Technology Officer for MTN, recently acknowledged that MTN were subsidising their 3G data traffic with revenue from their voice and SMS business.  This means, when it comes to communication, that the poor in South Africa are effectively subsidising the wealthy.

• Or consider the microeconomics of the edge cases of mobile access.  Let’s pick a remote village served by a single cell tower.  We know a couple of things about mobile access in Africa.  One, the majority of calls are local.  Two, people are spending substantial amounts of their disposable income on access.  So, in a remote village if 50% of the phone calls are local and people are spending 50% of their disposable income on mobile access, that means that 25% of their disposable income is being siphoned out of that village.

Can’t we acknowledge the achievements of mobile phones on the continent while at the same time look at the larger ecosystem?   Ecosystem…. the word set me thinking.  Then suddenly it came to me… the Blue Gum tree.  The Blue Gum tree is an excellent metaphor for understanding mobile operators on the continent.  If you have spent time in Africa, you will have likely have noticed the Blue Gum trees especially in Southern and Eastern Africa, and the Horn of Africa.  The Blue Gum tree (Eucalyptus) is not native to Africa but was imported to the continent in the late 19th century because it is fast growing and suitable for both timber and fuel.

However, in recent years the Blue Gum tree has gone from panacea to parasite.  Not only does the Blue Gum consume 80 to 200 litres of water per day but it also attempts to kill off surrounding plant life by releasing a chemical into the soil.  From South Africa to Kenya the effect of the Blue Gum tree on water scarcity is being recognised.  So while no one doubts the utility of Blue Gum trees, the impact on the water table as well as the killing off of indigenous competitors needs to be taken into account.

I don’t want to get rid of Blue Gum trees or mobile operators but I would like them to stop retarding the competition and for there to be a more balanced telecoms ecosystem.  The telecoms sector in Africa needs a thriving understorey that can develop new technologies and innovations that can ultimately grow up to challenge the “old growth”.  Without it we all remain thirsty for innovation on the continent.

A final note.  When giving out Mpesa awards, just once I would like to see Simon Batchelor of Gamos Consulting mentioned.  Simon has been a tireless champion of mobile money since I have known him and I know how hard he lobbied Safaricom to take up the idea.

This third  “What Google Should Do In Africa” post could be subtitled “Grow some balls”.  Why, oh why, is it that Google, so unafraid to tackle telco and broadcast market behemoths in the United States, behaves like a timid NGO in Africa?

Although this post has been in the queue for a while, the timing now could not be better as two days ago Google launched innovative new SMS-based services in Uganda in partnership with the Grameen Foundation and MTN.  This provoked a response from Katrin Verclas (@mobileactive) in which she queried the apparently high costs of the premium SMS charges being levied.  This was riposted by Erik Hersman (@whiteafrican) who rephrased Katrin’s post as the question “If you provide services to the poor, should you make a profit?”.

For me the problem is not whether the poor should pay for services.  I don’t think anyone engaged in this discussion believe the poor shouldn’t pay for services that are of value to them.  It is not even whether they should pay a premium rate for services.  The problem for me is the base rate itself of SMS charges.  Google and Grameen have correctly identified the tremendous potential power of SMS as a technology that can effectively provide data services to the poor.  However, this transformative technology, whose marginal cost of deployment is effectively zero, is being throttled by mobile operators charging a disproportionately high price for the service.  Mobile operators in Africa still embrace the economics of scarcity.

In his post, Erik makes a provocative statement.  He says,

If there’s a problem with collusion and price fixing in an industry (like there sometimes seems to be with SMS services in a country), that’s something beyond the scope of individuals and needs to be tackled separately by regulation.

If only it were that simple.  Telecoms regulation in Africa is in a parlous state.  With few exceptions (Nigeria being a notable one) communications regulators in Africa are under-resourced and often insufficiently independent from governments who maintain a substantial investment stake in the incumbent fixed-line and mobile operators.

So what happens in practice?  Regulators often start out well.  They issue a call for public input on issue X,  be it interconnection, local-loop unbundling, carrier pre-select, spectrum licensing or what have you.  Who responds to these calls for input?  Right now in most African countries the only organisations effectively lobbying the regulator are fixed-line incumbents and mobile operators…. with PREDICTABLE RESULTS.  The old joke about a telecom company being a law firm with an antenna stuck on top is actually not that funny in Africa. Incumbent operators are experts at both influencing policy and regulation development and at stalling any efforts to reform unfair practices.

For the rest of us, lawyers are notoriously expensive and there are few civil society organisations with the resources to actually draft the kind of input that regulators need in order for there to be a balanced debate.  There is a desperate need for organisations like Google who have a vested interest in seeing more data traffic to help lobby for more competition, for lower barriers to entrepreneurship in the telecom sector, and for cheaper access for all.

So when I see the company that wagered billions in the 700MHz spectrum auction in the U.S. to effectively arm-wrestle Verizon into OpenAccess conditions, the company that has made countless submissions to the FCC to lobby for unlicensed access to television white spaces spectrum, announce that they have “partnered” with a single mobile operator in Uganda to deliver SMS services, you will understand me if I seem a little let down.  The new SMS services for Uganda ARE innovative and I believe they have been well-conceived.  Kudos to the Grameen Foundation for developing them and to Google for supporting them.  Am I wrong to want more from one of the most innovative companies in the world?

Imagine the innovation in services that might be unleashed if SMSes were priced so that Africans didn’t have to think twice about sending them.  Imagine the economics of abundance being applied to the telecoms sector in Africa.  Sadly, voices calling for this on the continent are not nearly loud enough.  Google, which represents that principle so well, disappoints by failing to stand up for it.

It’s not sexy, it’s not whizz-bang, but it is true that Google could have a more profound effect in Africa by hiring a few lawyers, lobbyists, etc who can help level the playing field for the regulators on the continent than a dozen SMS services.

Ironically, Chris Anderson’s new book that explores the economics of abundance quotes Google CEO Eric Schmidt on the cover. Schmidt says:

“With the cost of distribution relentlessly driving towards zero, Chris Anderson, has once again identified the next big thing”

If he really believes that, why not Africa too?

This the second installment in a series of posts in which I have the hubris to reflect upon What Google Should Do In Africa (#WGSDIA). There is some context for this post in the preface to the series.

Imagine that you are a smart entrepreneurial African geek who has recognised that the huge opportunity of providing low-cost voice and Internet access in Africa.  You put together some new but remarkably inexpensive WiFi equipment from a company like Ubiquiti and set up a wireless mesh network in your community.  You backhaul it to the Internet with an ADSL line if your lucky enough to be near one or with a wireless backhaul to your most obliging ISP.  So far so good.  But what you discover with your community is that while there is a small and slowly growing demand for Internet access, the real demand is still for affordable voice.  So you offer free voice services over your wireless mesh and the community loves it because, after all, the majority of phone calls are local anyway.

The community is very excited about this and the love the service but they want more.  They want to be able to receive calls from mobile networks and the Publicly Switched Telephone Network (PSTN) and they want to be able to make calls out as well.  They want to have real phone numbers that people can call them on from anywhere.  Now things become more challenging.  In South Africa, in order to be allocated a block of phone numbers, you need an Electronic Communication Services (ECS) license from the regulator.  This is not too big a deal and requires only modest sums of money and slightly larger amounts of patience.  Having achieved that you now have to negotiate interconnects with the major operators.  This IS a big deal.  The operators each require a million Rand (approx 125,000 USD) as a bank guarantee to create the interconnect.  Some of them still require extremely expensive carrier-grade equipment in order to connect.  Worst of all, some of them either flat-out refuse to interconnect (which is against the law but some operators are a law unto themselves) or bureaucratively render the process so slow that they might as well have refused.

Here the entrepreneur runs into a brick wall that requires both deep pockets and deep patience to resolve.

Cut over to Mountain View.  On March 11th of this year, Google launched GoogleVoice, a service which offers you a single phone number which will ring all of your phones.  In addition, Google Voice offers free calling and texting in the continental U.S. as well as voicemail, conference calling, voicemail via email, the list goes on.  Google Voice is essentially a re-branding and upgrade of Grand Central which launched was founded in 2005 and acquired by Google in mid-2007.  Currently Google Voice is in private beta for existing Grand Central users.

For the present, this is a niche service in North America that will appeal more  to geeks and early-adopter technophiles because most people in North America already enjoy reasonably advanced and inexpensive phone services.  So why pitch Google Voice in Africa?

Consider the opportunity.  One could launch Google Voice but do it slightly differently.  Launch the existing service which might appeal to anyone with existing broadband but also launch it as an API for entrpreneurs.  Then all the small entrepreneur needs to do is find a way to get connected to the Internet and Google Voice can provide upstream services.  They could allocate phone numbers to the entrepreneur (or he/she could bring their own) but most importantly GoogleVoice could manage the interconnections with the mobile operators and the PSTN.   Obviously anyone selling voice services would still need an ECS license but that is a comparatively tiny hurdle to overcome.

Full disclosure:  this sort of service would tie in perfectly with the development Village Telco and the Mesh Potato currently being supported by the Shuttleworth Foundation.

I believe the launch of Google Voice in the U.S. signals a sea change in the telecoms sector around the world. Most significantly it points to the inevitable future of all voice services being Internet services. Creating a flat IP-based infrastructure for phone calls in Africa could begin to break down the mobile operator walled gardens and open up the possibility of operator-neutral value-added services in Africa.

So What Should Google Do About Google Voice in Africa?

I would pick a country in Africa where getting a communication network and/or services license is not too onerous and in which providing access on unlicensed WiFi spectrum is not illegal.   Thanks to the Altech decision last year, South Africa is currently a good place to do this but it could be Kenya, Nigeria or somewhere else.  Setting up something like GoogleVoice would require some work but probably much easier to pick a company like Dabba who are doing most of this already and work with them to evolve and adapt an African Google Voice.