How Mobile Money Interoperability in Ghana affects fees Mobile Money Users Pay: Telco Price Analysis
I took a deep dive into Telco Mobile Money prices in Ghana to determine what interoperability means for fees mobile money users pay, specifically low-income people. Find the full analysis here.
For reproducibility, I outline what inspired my analysis, how the Telco mobile money fees data was collected, the code and tools used in the analysis and the insights below.
- Say a mobile money user wants to make a transaction, after factoring in cash in fees (if any), transfer fees, and the cash out fees, what is the final amount withdrawn? Or put another way, what 2 combination of networks allows the receiver to cash out at least cost in terms of fees?
- Given that mobile money users can switch to any network since the market is not fragmented anymore, how competitive are interoperable transaction options to same network transactions?
- What does interoperability mean for low-income people who typically send small amounts?
Mobile Money transaction fees data was sourced from AirtelTigo Ghana, MTN Ghana and Vodafone Ghana and can be found below.
A .csv file of the table below can be found here.
| Price Ranges | MTN | Vodafone | AirtelTigo | |||
|---|---|---|---|---|---|---|
| Across Network | Same Network | Across Network | Same Network | Across Network | Same Network | |
| GHS 1 - GHS 50 | 0.75 | 0.5 | 0.5 | 0.5 | 0.75 | 0.5 |
| GHS 51 - GHS 75 | 1.50% | 1% | 1.50% | 0.5 | 1.50% | 1 |
| GHS 76 - GHS 100* | 1.50% | 1% | 1.50% | 0.75 | 1.50% | 1 |
| GHS 101 - GHS 250 | 1.50% | 1% | 1.50% | 1 | 1.50% | 1.5 |
| GHS 251 - GHS 500 | 1.50% | 1% | 1.50% | 2 | 1.50% | 2 |
| GHS 501 - GHS 1000 | 1.50% | 1% | 1.50% | 3 | 1.50% | 2.5 |
| GHS 1 000 - GHS 5 000 | 15 | 15 | ||||
| * Range included to accommodate Vodafone's unique 50 -75/76 - 100 fee point split |
For each Network, I collected fees data on same network (on-net) and cross network (off-net) mobile money transfers. Data was collected from GHS 1 - GHS 1000, however for networks where price points beyond GHS 1 000 was available, it was also collected.
| Price Ranges | MTN | Vodafone | AirtelTigo |
|---|---|---|---|
| GHS 1 - GHS 50 | 0.5 | 0.5 | 0.8 |
| GHS 51 - GHS 75 | 1% | 1.5 | 1.5 |
| GHS 76 - GHS 100* | 1% | 1.5 | 1.5 |
| GHS 101 - GHS 250 | 1% | 2.5 | 2.8 |
| GHS 251 - GHS 500 | 1% | 4.0 | 4.0 |
| GHS 501 - GHS 1000 | 1% | 6.0 | 6.0 |
| GHS 1 000 - GHS 5 000 | |||
| * Range included to accommodate Vodafone's unique 50 -75/76 - 100 fee point split |
momo_value_for_money(network_from, network_to , initial_amount)- This function takes 3 arguments: network the transfer is from, network the transfer is going to and amount being sent. It then returns adata.frameof the network combination and the amount withdrawn.
Eg: momo_value_for_money(network_from = "AirtelTigo Money", network_to = "Vodafone Cash", 10)
| Mobile_Money_Service | withdrawn_amount |
|---|---|
| AirtelTigo Money to Vodafone Cash | 8.75 |
Within the momo_value_for_money() function, 2 functions operate:
momo_transfer(initial_amount, transaction_fee , effect = c("flat", "percentage"))- Computes the amount left after transafering.momo_withdrawal(remaining_amount, withdrawal_fee , effect = c("flat", "percentage")- Computes the amount left after withdrawing.
Each network combination amount is iteratively computed from GHc1 to GHc1000 using purrr::pmap_df().
I found that, users who send high amounts have relatively good options (further expanded by interoperability), but if it's less it doesn't matter what network they use.
If Government of Ghana has plans to tax mobile money transactions as being reported, Telco's could consider looking at higher amount ranges (500 - 1000, 1000+) since these amounts generally pay lower fees in percentage terms, although it feels like the platform is still in its infancy.
Find the full analysis here.