It all started with an impassioned presentation by the US ambassador to Kenya, Margaret ‘Meg’ Whitman. She made three claims about M-Pesa, the Kenyan mobile money transfer and payment service that has won accolades for its novelty, versatility and convenience.
The three are:
- “M-Pesa is involved in 70% of Kenya’s transactions.”
- “59% of Kenya’s annual GDP flows through M-Pesa.”
- “70% of Tanzania’s GDP flows through M-Pesa.”
Whitman made these claims to US business leaders visiting Kenya in March 2023, at a business event in New York City in April 2023, at Kenya's devolution conference in August 2023 and in San Francisco in September 2023 during a US-Kenya business roadshow.
Showing their reach, a news commentator repeated the claims in a widely read South African publication, citing the ambassador's presentations as the source when we asked.
Throughout September, we repeatedly contacted Safaricom, the Kenyan telecoms company that owns M-Pesa, by phone and email, seeking clarification on the figures behind the mobile money service's phenomenal story.
Apart from promises that the information we sought was being worked on, we have met a wall of silence. Should we ever hear anything, we'll update this report.
There are two key problems with the claims:
Problem 1: Ambiguity
Take, for example, the claim “M-Pesa is involved in 70% of Kenya’s transactions”. Which “transactions” are referred to?
Is it all financial transactions in Kenya, specifically money transfers, or all mobile money transactions, including payments? Are we talking about the number of transactions or the value of those transactions?
To try to figure this out, we asked the Central Bank of Kenya. A spokesperson referred us to the latest banking supervision report, which includes data on all mobile money transactions.
The bank used its own payment statistics and data from the Communications Authority, Kenya's telecommunications regulator, as sources.
The report showed an average monthly volume of 207 million mobile money transactions in 2022, valued at KSh708 billion (US$6 billion at the average 2022 exchange rate).
The latest report from the telecoms regulator doesn't give details of mobile money transactions, apart from data showing that Safaricom, accounted for 98% of revenue from “other services”, which includes mobile money.
A spokesman for the Communications Authority referred Africa Check back to the central bank, a circular loop we see too often.
Problem 2: Missing data and inconsistency
According to the Kenya National Bureau of Statistics (KNBS), the country’s data agency, mobile money provisionally recorded total transfers of KSh7.9 trillion ($67 billion at the time) in 2.28 billion transactions in 2022.
The “transactions” are for money to and from agents. A further KSh20.3 billion ($172 million) was transacted via “mobile commerce”, but the number of transactions wasn’t available.
We pulled up Safaricom’s most recent annual report, 2023. It showed that M-Pesa processed “2,600 transactions a second”, which worked out to “21.03 billion transactions per year, all valued at KSh35.86 trillion” (about $304 billion - the company’s financial year starts in April).
Now for the puzzling part: the statistics bureau reported 2.28 billion transactions per year. Safaricom reported 21.03 billion, which itself is more than nine times the total number of all mobile money transfer transactions recorded by the central bank over the year.
There's a clear inconsistency. The data is for 12 months - even if there were differences in the reporting periods, the variations would not be that large. The KNBS also did not have details of mobile commerce transactions. Could this missing data explain the difference, given the ubiquity of mobile payments?
We attempt a solution: working backwards
We tried another approach, using Safaricom's figures.
- There are 86,400 seconds in a day (24 hours * 60 minutes * 60 seconds).
- A year, therefore, has 31,557,600 seconds (86,400 seconds * 365.25 days).
- Multiplied by the 2,600 transactions per second, we get 82 billion (82,049,760,000) transactions, not 21.03 billion as stated in the annual report.
Furthermore, if we take the total number of transactions per year as stated in the company’s annual report (21.03 billion) and divide it by the number of seconds in a year (31,557,600), we arrive at 666.4 transactions per second. This is significantly lower than the 2,600 stated in the annual report.
We asked Safaricom to clarify if we were missing anything, such as technical nuances in the definitions of transactions and calculations. What exactly counts as a transaction? Since M-Pesa is available 24/7, how does Safaricom calculate transactions per second?
We also shared data from the statistics office and the central bank, and asked Safaricom to help us crunch the numbers and perhaps share their regulatory filings.
We emailed, phoned, emailed, phoned again, texted and phoned and phoned. Silence. They burrowed deeper.
‘Unfortunate wording’
Fact-checking the two claims that “59% of Kenya’s annual GDP flows through M-Pesa” or that “70% of Tanzania’s GDP flows through M-Pesa” is complicated.
The use of the word “flows” suggests that M-Pesa is a major player in handling a significant proportion of the financial activity that contributes to Kenya's economic output.
However, it also means that M-Pesa is not just facilitating transactions, but is having a significant impact on economic activities that contribute to the gross domestic product (GDP).
“The confusion is due to what I would term as unfortunate wording,” Prof Isaac Mbiti, co-author of a 2016 paper on the impact of M-Pesa in Kenya, told Africa Check.
He is an associate professor of public policy and economics at the University of Virginia's Frank Batten School of Leadership and Public Policy.
In his paper, Mbiti referred to the value of money transferred via M-Pesa as “equivalent” to GDP. “It is not trying to say that M-Pesa is driving GDP. It is just a way of benchmarking a big number,” he said.
‘Messy statistics’
This framing of the value of M-Pesa transactions relative to GDP has been used by journalists in foreign publications, in some country reports crediting M-Pesa with "processing" nearly half of GDP, in workshop reports and in Kenyan news outlets. In 2023, newspapers reported that M-Pesa had transacted “three times Kenya's GDP”.
Even Sitoyo Lopokoiyit, the managing director of M-Pesa Africa, juxtaposed the value of transactions with GDP using the word “flows”.
Dr Susan Johnson, a financial inclusion scholar who has worked in Kenya, while affiliated with the University of Bath in England, strongly criticised the framing as “messy” in 2014.
That’s because mobile money transaction values are about the total flows in and out of mobile wallets, while GDP is about the value of goods and services produced in a given time.
“Comparing the GDP figure with the transactions figure, therefore, compares figures that are not conceptually equivalent to each other,” Johnson wrote in a blog for Bath.
The Global Mobile Service Association, the international industry lobby for mobile service providers, corrected its data following Johnson’s brief on the misleading statistics on M-Pesa and Kenya’s GDP.
M-Pesa’s impact is not captured in GDP numbers
Given the headline-grabbing statistic and the prevalence of M-Pesa in Kenya, we checked the country’s GDP figures to see if M-Pesa was on the radar.
Agriculture is the biggest contributor (21%) to Kenya’s GDP, provisionally valued at KSh13.4 trillion (about $114 billion) in 2022. Telecommunications and financial activities, where M-Pesa falls, together contribute 7.2% of the GDP.
You would expect a more outsized share of the GDP if the claims about M-Pesa added up.
It is not every day that verification is elusive. Nor is this elusiveness bad news. In our search for the facts we have gained valuable insights into the shortcomings that allow misleading narratives to flourish.
Every unverified claim is an opportunity to uncover the nuances and strengthen the foundation of accurate information. We know we need better data, presented clearly and transparently, to enable honest public debate.
Indeed, fact-checking is empowering when we keep it simple, transparent and honest, like Safaricom's brand promise.
We will continue to knock, Safaricom.
Add new comment