Africa, Technology and Lindy Hopping

Over the last few weeks since I announced Functional Africa, I’ve had the pleasure of talking to several people who have been kind enough to give me some of their time to talk about the project, and their experiences.

I always saw this as a learing experience for me, and I entered into it well aware of my lack of knowledge about the vast continent of Africa—I’ve spent just a couple of months in one African country—so I’ve been extremely grateful for the opportunity to listen and grow my knowledge.

One detail I mentioned in my presentation at Functional Scala was an anecdote I had known for many years: that despite having less widespread access to technology, mobile payments have seen greater adoption in Africa than in many countries in the rich world.

As I discovered during my conversations, the adoption of technologies like M-Pesa was not so much a rare success of advanced technology in Africa, but more a pattern that has repeated itself many times over.

The Mobile Revolution

In 2004, Africa became the first continent to have more mobile phones than fixed-line phones. In the USA, that threshold was only crossed in 2016. And only in 2019 did the number of calls in Germany made with a mobile phone overtake the number originating from a fixed line.

It seems astounding that it was nearly two decades ago that Africa, as a continent, reached this particular milestone, while that same milestone has only been passed in recent years elsewhere in the world. Nevertheless, the statistic does flatter Africa for one clear reason: there was very little penetration of fixed-line phones in Africa to begin with.

Fixed-line phones were ubiquitous in Europe by the 1990s. The ability to communicate with other people by phone—socially and commercially—combined with the Network Effect made communication by phone extremely convenient, and businesses changed to make the telephone central to their operations, to the extent that it was impractical to get through life at the turn of the millennium without a phone.

But on a continent where even electricity was never universal, that fixed-line phone revolution just didn’t happen. And the same network effect which helped adoption in Europe hindered it in Africa; put simply, when there is nobody else to phone, there is not much point having a phone!

Mobile phones, however, required less expensive infrastructure—no endless metres of cable to lay and telephone exchanges to build—and less commitment from the early adopters, while SMS messaging offered everyone a cheap and novel way to communicate that was not previously possible. So mobile phone technology did not take long to overtake fixed-line phones in countries where the latter had never taken hold.

And so it appeared that most of Africa leapfrogged an entire generation of technology, to lead the world (by over a decade) in being having a mobile-oriented society.

Electronic Payments in Africa

And so, when M-Pesa was launched in Kenya in 2007 by mobile operators, Vodafone and Safaricom, its mobile-ready society provided fertile ground for the service.

With this launch, it became possible for anyone with a mobile phone to deposit and withdraw money, transfer money to friends and family, pay bills, and even borrow.

M-Pesa filled a void for many members of society. Traditional banking services were expensive and not widely used in Kenya, particularly in more rural areas.

And since its launch fourteen years ago, M-Pesa has spread to other African nations: Tanzania, South Africa, Lesotho, Democratic Republic of Congo, Ghana, Mozambique and Egypt, as well as Afghanistan.

Much as weak penetration of fixed-line phones in Africa led to more rapid adoption of mobiles, less widespread engagement with the banking system—without any lack of inherent need for banking services—meant that electronic, mobile payments were fully embraced in Africa when they became available.

And that couldn’t have happened elsewhere in the world.

In many European countries, even in the 2020s, shops which only accept cash are still common, and the UK government has been trying for years to shut down payment by cheque—only to face resistance from the population.

The Lindy Effect

The Lindy Effect is a phenomenon which can be used to predict the future life-expectancy of a technology based on its age. It states that the longer a particular technology has been in use, the longer it is expected to continue to be used.

For example, consider the text editors, Atom and Visual Studio Code are both about six years old, and by almost any measure are more modern editors than Emacs and vi, which are 45 years old.

But despite the recency of the technology, the Lindy Effect is generally a better predictor of longevity, and it suggests that vi and Emacs will still be around long after VS Code and Atom are no longer maintained.

This offers no commentary or analysis on the quality or design choices of any of those four editors. Nor does it offer any guarantee that one technology will outlive another. It just observes that, statistically, past survival is one of the best indicators of future survival.

We can explain this phenomenon in two different ways.

Firstly, the lifespans of technologies obey a power law distribution: at any given time, the technologies that are considered “current” will include many that are recent and new, and a long tail of fewer and fewer current technologies from the more distant past.

If we take the derivative of the relationship, it implies a decreasing mortality rate, meaning that as a technology gets older, its chance of disappearing goes down (within a unit of time).

Alternatively, and more tangibly, the longer a particular technology has existed, the more people, organizations, services, commerce, investment and mindshare will have come to depend upon it, and the harder it becomes for those interactions to be undone in order for the technology to disappear. And so it doesn’t; not so easily, at least.

Lindy Hopping

In technology-saturated Europe and America, the Lindy Effect can hold us back.

A long-established incumbent technology—be it cheques, fixed-line telephones, or a particular text editor—is much harder to displace than a recent invention. That means that a new technology challenging an incumbent technology must work harder to beat it.

But new inventions are arriving and competing for adoption in their space all the time. And for the most part, the rich world is the target market for that competition. While more products are brought to market first in Europe and America. And despite the innovation, most of those products are short-lived, not so much because of contemporary competition, but because of established competition.

Many technologies never became widespread in Africa because theirs proponents always thought first about the rich world as their target market. But many technologies never become widespread in the rich world because they can’t compete against incumbent technologies in the same space.

The opportunity is there for Africa to recognize innovative ideas in the areas where their precursors never took hold, and to adopt them while the rest of the world makes the long slow migration away from legacy.

Africa has the chance to leapfrog yesterday’s technologies and go straight to tomorrow’s. And I see Functional Programming as precisely such a technology.

Functional Programming’s Opportunity

As a software developer, I primarily use Scala, a functional language, and I work with many other developers who use Haskell, another FP language. And we recognise the benefits these languages bring over more established imperative languages like Java and Python.

But we are nevertheless in the minority within the software industry, while the majority continues to write code in the imperative paradigm.

Functional programming offers many advantages—code that is easier to reason about, and software that is smaller, simpler and more reliable—but to many businesses, these advantages are less important.

Most companies prefer to invest their development time in long-established programming techniques because they are perceived as proven by their longevity; safe, even if unremarkable, and supported by huge pools of developers in a dynamic jobs market. That helps to creates jobs, too, as young developers learn these languages in order to get work. Consequently, the large body of programmers working in those languages further perpetuates their adoption.

Africa’s software development industry is active, but still small. Compared to more economically-developed countries, there are fewer software projects, fewer developers and fewer jobs in software development.

But that liberates technology decisions from the legacy of the incumbents: the technical advantages of functional programming languages remain, while the benefits of more established languages don’t exist. And technologies can be chosen on their technical merits, without bias for their place in the market.

I see Functional Programming as a huge opportunity for software developers in Africa. While the continent fell behind the rest of the world in software development in the last few decades, that gap now presents a chance to get ahead, and to “Lindy hop” yesterday’s technology.