Research shows that the introduction of taxes on mobile money nudged Ugandans away from using digital payments to cash. Photo / Edgar R Batte
It began with a simple promise: send money home. But what started as a workaround for the unbanked has quietly grown into one of Africa’s most profound financial revolutions.
Today, mobile money is more than just a tool - it is a lifeline. According to the GSMA 2023 State of the Industry Report on Mobile Money, more than 55 percent of adults across Africa now rely on it not just to send and receive funds, but save, borrow, insure, and manage everyday financial needs.
Globally, mobile money moves $3.5b every day - much of it flowing through Africa’s digital rails.
This is the real meaning of when technology meets necessity, it can redefine how economies work. But as mobile money expands, so has government picked interest.
In 2018, government introduced the digital money tax - originally at 1 percent - but later reduced to 0.5 percent on withdrawals. This tax sought to increase revenue from the telecommunications and financial sectors.
The question, however, is not just about revenue collection. It’s about consequences. Who ultimately bears the cost of taxing digital money? And how does this reshape financial behavior?
To answer this, we need to understand how mobile money interacts with banking, particularly through the lens of credit markets and financial inclusion.
This is where economists Lorenzo Spadavecchia (chief research officer at ION Group and a consultant with BoU), Jimmy Apaa Okello (BoU’s director of financial markets), and Samuel Musoke (BoU economist) enter the picture.
They built what they call a “toy model” of currency substitution, a focused framework that captures real-world tradeoffs consumers face: Ease and convenience of storing money digitally versus hidden transaction costs.
Through this lens, they unpack how shifts in digital transaction costs can ripple through the financial system, affecting how money is saved, borrowed, and moved. Their focus is on Uganda’s 2018 July digital tax that was introduced with almost no public warning.
The result?
A sudden and unexpected increase in the cost of using mobile money. Before the tax, mobile money was Uganda’s go-to for payments and deposits. But after the tax, card payments, which had been rare, picked up.
This pointed to mobile money’s rising cost, which made it less attractive, and in some cases, made people return to banks.
Entry of banking agents
If banks were always there, why weren’t people using them before? The answer lies in digital financial innovations such as agents in shops and kiosks, offering banking services on behalf of traditional institutions.
After the tax was introduced, there was a surge in bank deposits, particularly through agent banking, who were already spread across Uganda, offering a convenient way to deposit money, especially in areas where banks had a physical presence through automated teller machines (ATMs).
The pattern was clear: People deposited funds via agents and withdrew cash from ATMs to make payments. BoU data confirms the shift, with mobile money deposits dropping from $121m in mid-2018 to $70m in the quarter after, while bank deposits through agents rose from nearly zero to $9m.
And it didn’t stop there. By the second quarter of 2019, deposits through agents hit $50m, while mobile money balances remained low at around $100m.
This shift strongly suggests a knock-on effect, in which the tax nudged users toward traditional banking channels.
Digging deeper
The researchers didn’t stop at local data. They dived into global literature on digital currencies, financial intermediation, and mobile money, analysing studies from 2011 to 2023, including three specifically on mobile money's impact on banks.
They also partnered with a mobile money provider, gaining access to a trove of data: two billion transactions from 2018, 20 million users, and geolocation data from two million random users.
The result? A rich, data-driven exploration of how even a small policy tweak - a few percentage points of tax - ripples through an entire financial system, influencing where people store, move, and spend their money.
The researchers validated their findings using data from the Uganda National Panel Survey, focusing on 3,000 households, along with new banking technology adoption data across 136 districts.
They also tapped into monthly and quarterly bank balance sheet data from 26 banks, and leveraged data from the national credit registry, covering two million loans, tracking borrowers’ locations, demographics, and credit histories.
Their identification strategy hinged on time variation - the unexpected shock from the 2018 mobile money tax - and geographical variation, particularly the presence of ATMs, which acted as a proxy for how easily users could substitute mobile money with bank-related services.
Their core finding? The mobile money tax increased the cost of using mobile wallets, prompting a shift in user behaviour toward banking agents, a relatively underutilised financial innovation already present in Uganda.
This led to a rise in deposits and withdrawals through banks, and a broader rise in usage of cash.
They also observed that in areas with high ATM density, mobile money usage dropped by nearly 10 percent - a pattern supported by their difference-in-differences analysis and validated using household-level survey data.
One striking shift was the rapid adoption of banking agents in high-ATM-density districts.
Within three to six months of the tax’s introduction, the number of agents in those districts surged from two to 120. In contrast, low-ATM-density districts saw a more modest increase, from just 0.1 to eight agents.
The trend wasn’t just geographical. At the bank level, institutions with broader ATM networks saw sharper increases in banking agents and corresponding spikes in cash deposit inflows.
The deposits, however, were primarily short-term, aligned with a high turnover of liquidity.
BoU data confirmed the shift with an uptick in cash usage and ATM withdrawals in high-density districts.
In essence, the mobile money tax nudged users toward hybrid financial behaviour: banks for safekeeping, cash for spending.
For banks, this shift meant a new inflow of liquidity, but one characterised by rapid movement in and out, rather than long-term stability.
The credit market
So, what happens to the credit market amid this shift - and what are the implications?
To answer this, the researchers dug into Uganda’s credit registry, examining how the bank lending channel responds to liquidity shocks. They focused on credit terms, loan volumes, the number of loans issued, and interest rates.
The data revealed that loan maturity dropped across the board - banks started issuing shorter-term credit, consistent with the volatility introduced by rapid cash movements in and out of banks.
Interestingly, banks increased lending to low-risk borrowers with established credit histories, while scaling back credit to high-risk borrowers, particularly those without prior credit records.
The result? Higher borrowing costs for riskier borrowers.
How taxes on mobile money assisted banks
This behavioral shift underscores a key policy debate now echoing across Africa: How do digital taxes affect the broader financial system? Uganda’s case provides a rare data-rich window.
Here, the mobile money tax didn’t crowd out banks - in fact, it did the opposite. As the cost of digital payments rose, users shifted to bank deposits and cash.
“The data disproves the idea that digital currencies inherently replace bank deposits. In Uganda, higher digital costs simply nudged users back to banks,” Lorenzo says.
It also highlights a deeper friction: Fintech monopolies can stifle innovation.
In this case, mobile money’s dominance appeared to delay growth of complementary banking innovations like agent banking.
“As soon as mobile money became expensive, we saw a surge in banking agents,” he adds. Yet this transformation hasn’t come without consequences.
The volatile liquidity flow through agents - fast in, fast out -has strained banking models. Banks now face reduced loan maturities and a credit reallocation that favours low-risk end of the market.
