There's a calculation most sponsor banks have never run, and it takes about an afternoon.
Take one quarter. Total what the card networks (schemes) billed you. Total what you billed out to your programs for network fees over the same quarter. The difference is what you absorbed.
In our work with BIN sponsors globally, that number is usually larger than the people setting the pricing expect.
The reason isn't unwillingness to pass fees through. It's attribution. To recharge a fee accurately you have to know which program generated it, and for a large share of network fees the billing record doesn't tell you. Roughly half of network costs aren't directly linked to a transaction at all. They attach to BINs, licenses, registrations, product enrollments, etc. and many are assessed above the level of any single program.
So sponsors do one of three things. Pass through only what allocates obviously and absorb the rest. Apply a blunt allocation that overcharges some programs and undercharges others, which surfaces later as a commercial argument they usually lose, or worse, a lost client. Or build it into a blended rate set once and never revisited, which decays quietly as the fee schedule moves underneath it.
Where a sponsor can't attribute a fee, the price signal disappears for the party whose behavior generated it. The program manager never sees the consequence of its own choices, so it has no reason to change them, and the sponsor ends up holding both the cost and most of the incentive to fix it.
Run the number before anyone argues about method. Not because it tells you how to allocate, but because it tells you whether the structure is recovering cost at all.