The 1% US tax is not the biggest cost of sending money to Kenya
In my last piece I looked at the fall in remittances from the USA to Kenya, and at the 1% US tax on cash-funded transfers that began on 1 January 2026. I argued that the tax is too small and too narrow to explain the fall.
That raised a simpler question. If the tax is 1%, what does it cost to send money to Kenya in the first place?
The World Bank tracks this in its Remittance Prices Worldwide survey. It records what providers charge to send the equivalent of US$200: the fee plus the margin on the exchange rate. The latest round was collected in August 2025, a few months before the tax took effect.
Paying with cash already cost twice as much
From the USA to Kenya, the average cost was 4.26% across the 40 services surveyed. That matches the figure the World Bank publishes for this corridor.
The average hides a wide spread. When the sender paid with cash, the average was 7.70% (five services). When the sender paid from a bank account or with a card, it was 3.77% (35 services). Debit cards were the cheapest way to pay, at 2.93%. Transfers paid into a mobile wallet in Kenya averaged 2.27%.
On US$200, that is about US$15.40 for cash against about US$7.54 for a card or bank transfer. The tax adds US$2, and only on the cash route. The gap between channels was already more than three times the tax.
This fits the argument in my last piece. A sender who switches from cash to an app funded by a card avoids the tax and pays less in fees. That money still arrives through licensed channels, and CBK still counts it. A switch like that should not show up as a fall in recorded remittances.
The big corridors are getting cheaper
Sending money from the main diaspora countries has become cheaper. From the USA, the average fell from 7.54% in the second quarter of 2016 to 4.26%. From the UK, it fell from 6.62% to 3.43%. From Canada, from 10.16% to 6.00%.
For context, the global average in the same round was 6.36%. Sub-Saharan Africa, at 8.46%, was the most expensive region to send money to. The UN and the G20 want the global average down to 3% by 2030. The UK corridor to Kenya is already close.
The expensive corridors are next door
Within Africa the picture is very different. From Tanzania to Kenya, the corridor average was 59.99%. From Rwanda it was 22.56%, and from South Africa 10.62%. In 2016, Tanzania was at 13.07% and Rwanda at 12.54%.
The Tanzania figure is driven by banks. The survey priced nine bank transfers of TZS 324,500, the local equivalent of US$200. They cost between 57% and 88% of the amount sent: fixed fees of TZS 145,200 to 231,300, plus exchange-rate margins of 13% to 22%. M-Pesa did the same job for 7.86%. MoneyGram cost 3.86% and Western Union 0.72%.
This matters more than it used to. In August 2026, Tanzania sent Kenya US$11.7m through recorded channels, slightly more than Saudi Arabia (US$11.6m). Regional flows are growing, and some of the routes they use are priced as if nobody is watching.
Quoted prices, not paid prices
These are quoted prices for a standard transfer. They are not what people actually paid. The averages are simple means across services, not weighted by how much money each service carries. A few expensive bank products can pull a corridor average up even if few people use them. Most senders in Tanzania may well use mobile money or a money transfer operator instead. The survey has no data on volumes, so I cannot say.
The survey also leaves gaps. It has no price data for sending money from Saudi Arabia or the UAE to Kenya. And the latest round predates the US tax, so it cannot show whether providers changed their prices for cash senders after January 2026.
Who actually pays these prices?
These are open questions, not conclusions.
- Volumes. How much of the money from Tanzania and Rwanda goes through banks, and how much through mobile money or money transfer operators?
- Gulf prices. What does it cost a Kenyan worker in Saudi Arabia or the UAE to send money home?
- Pass-through. Did US providers absorb the 1% tax, or add it to the price for cash senders?
Post-tax prices and CBK's September file
The next World Bank round, the first to be priced after the US tax. And CBK's September release, to see whether Tanzania stays ahead of Saudi Arabia.
Rebuilding the World Bank's corridor averages
I downloaded the World Bank's complete Remittance Prices Worldwide dataset, which lists every service surveyed, and kept the corridors that end in Kenya: the USA, the UK, Canada, South Africa, Tanzania and Rwanda. For each sending country and quarter, I took the simple average of the total cost of sending US$200 across transparent services. This reproduces the World Bank's published corridor averages for the USA (4.26%) and Tanzania (59.99%). There was no survey round in the second quarter of 2025. The global and regional averages come from the World Bank's report for the third quarter of 2025. The CBK figures come from its Remittances by Source file for August 2026. I processed the data in Python. Data accessed 3 October 2026.
Sources
- World Bank, Remittance Prices Worldwide, complete dataset 2011 to Q3 2025 (Excel)
- World Bank, Remittance Prices Worldwide, Issue 54, September 2025 (Q3 2025 report)
- World Bank, Remittance Prices Worldwide, corridor page: United States to Kenya
- World Bank, Remittance Prices Worldwide, corridor page: Tanzania to Kenya
- CBK, Remittances by Source, August 2026 (Excel)
- Central Bank of Kenya, Diaspora Remittances
- 26 U.S.C. §4475, remittance transfer tax (Cornell LII)
- Elly Okinyo, "A record month, a shrinking US corridor: what Kenya's 2026 remittance data are telling us", 28 September 2026
Code to reproduce this analysis: github.com/eokinyo/kenya-remittance-costs