I love the tax career agent idea, because it seems so simple, seems to have no losers, and doesn’t seem to step on any taboos.
The basic idea is this: If the government auctions off the right to get the income tax revenue paid by taxpayer X diverted to the auction winner, then that winner becomes a career agent. Such an agent has an incentive to advise and promote X, just as career agents do today in acting, music, sports, etc. Such auctions are a direct substitute for government selling debt, which governments already do to convert future taxes into present revenue. Furthermore, competitive auctions will make sure the government gets paid its full value, and would also tend to pick auction winners who are actually better able to advise and promote clients X.
At the moment the government is your tax career agent, and they are doing a terrible job; they do absolutely nothing. So selling the role to someone else is a net gain if agents on average find it their interest to do some advising or promoting, and if their audiences aren’t on averaged fooled by that advice and promotion. At least if the value agents get by doing so is a lot bigger than the simple transaction costs of setting up the auction and diverting the payments, and such costs that can be made very low. Agents need not help most clients, but even if they only help 5% of clients that’s still an improvement over the status quo.
Now economic theorists will worry about transaction costs due to asymmetric information; what if some auction bidders know more than others about a taxpayer’s chances of future income? In that case the auction winner might get “information rents”, and so not have to pay full price, cheating the government out of a bit of their revenue. A simple solution to this is to do the auction very early in the life of taxpayer X, such as at birth. Then few know much about X’s future income, and so info rents are almost zero.
However, a much bigger problem is that many people find it creepy to think of someone they didn’t pick trying to influence their life. Even though such agents want to help them make more money, and have no official powers whatsoever to influence their behaviors; they can only advise and promote to audiences who will listen to them. Even so, many worry tax career agents will do illegal things, or exert influence malicious via gossip or bribes. So it seems best to modify the proposal to placate these worriers, even if that less placates economic theorists. So we need to find ways to help people feel more in control of their tax career agents.
Now if the tax authority that sets this up whole thing has a limited geographic scope, like a state or nation, then taxpayer X can always threaten to leave that region, and completely destroy the value of the asset that the agent has purchased. Furthermore, if X simply makes it clear that they won’t listen at all to their current career agent, but would listen to other possible agents who might fill that role, that should create an incentive to sell for their existing unloved agent to “divorce” them by selling their asset to one of those other agents.
These do seem to be substantial powers of X over their agent. But let’s give taxpayers even more powers, For example, maybe the tax career agent isn’t even created until taxpayer X explicitly asks for it. And maybe X can choose a limited time duration to be auctioned off. Allowing them to decide whether to authorize another auction latter when that first time period runs out.
Furthermore, we could more strongly encourage an unloved tax career agent to sell their role to someone else via setting a self-set (“Harberger”) tax on the asset that the tax career agent holds. Just as self-set taxes on physical property encourage transfer of property to those better above to gain value from it.
Finally, we might authorize taxpayer X to give a 1% discount on this self-set property tax to any owner of their choosing. This would give a further push for a unloved agent to sell their asset to a better loved agent. And yet not much threaten government revenue.
All of this seems to me to give taxpayer X plenty of powers to discipline any disliked tax career agent. Thus allowing people to feel less creeped out by this idea. All while having governments suffer only modest info rents that cut their auction revenue.
This poll says the ideal self-set tax cuts the asset value by 7%. And this poll suggests that 70% support tax career agents; I might do a more nationally representative poll to see how well that generalizes.
A different concern, more on the overall population level, might be that tax career agents will increase differences in outcomes, maybe significantly so. Individual that for whatever genetic or environmental reasons are predicted (correctly or not) to be low tax value will necessarily receive proportionally lower investments. This will drive social structures toward being less egalitarian.
Well, there is an alignment problem of that kind, as there is in current private industry as well, but I wouldn't say that governments who administer income tax programs are really aligned with maximizing tax revenue either. The politicians are aligned with getting people to vote for them, which means they often campaign based on playing to the cognitive biases of their voters, and appeasing their corporate donors, rather than to any real goal like "maximizing total revenue." As long a politician can claim that their policy is going to improve everybody's income - and be believed by the uninformed average voter - then the politician can get re-elected, regardless of whether their policy actually does that. If the policy fails, that just creates an opportunity to blame the other side for the failure.