> Widespread gaming or no, the system pretty much works according to the internal aims. Cities get a list of their internal diaspora, and do make considerably more effort to stay in touch with them than they did previously. (This includes lovely holiday cards and sometimes even I-can’t-believe-they’re-not-alumni-magazines.) You really do get plums from childhood in your mail from your hometown (if you don’t optimize for cash equivalents). Cities with declining local tax bases really do get enough money to do material projects with. Tokyo takes a hit to revenue but can afford it.
The claim isn't that it doesn't work, but that the overhead is substantial, and those same goals could be reached with far less. I think it's self-evident from the description of the scheme - you could come up with a much more straightforward tax-and-redistribute scheme with less overhead.
But the overhead, as noted in the OP, is about 3% of the money transfer. That's low.
You could say that hometown marketing and alumni magazines are "overhead", but you can't say that at the same time you're listing them as a beneficial effect of the program.
That's the overhead of cashing the gift cards only.
And those other things are not overhead if and only if the person actually derives value from "hometown services". If all they do is send money there to get a gift card back and cash it, then all the bureaucracy in that town that handles gift cards is also overhead.
> Widespread gaming or no, the system pretty much works according to the internal aims. Cities get a list of their internal diaspora, and do make considerably more effort to stay in touch with them than they did previously. (This includes lovely holiday cards and sometimes even I-can’t-believe-they’re-not-alumni-magazines.) You really do get plums from childhood in your mail from your hometown (if you don’t optimize for cash equivalents). Cities with declining local tax bases really do get enough money to do material projects with. Tokyo takes a hit to revenue but can afford it.