3 ms·
You don't need the code even - though the code, as being already paid for by the taxpayers, should be available to them anyway - the rules are public, the data
by smsm42 3y ago
You don't need the code even - though the code, as being already paid for by the taxpayers, should be available to them anyway - the rules are public, the data should be available to the public, so you can just reimplement the rules and see if you get the same answer. They key here to make the data a) available b) portable c) every decision the system takes be documented and having the audit trail.
- AnthonyMouse 3y agoYou do need the code because if the public system gets it right for (or is otherwise used by) 90% of the taxpayers, a private system for the remaining taxpayers would have to cover its costs with only 10% of the revenue. Releasing the code reduces their costs because they don't have to duplicate the parts the government already got right, they only have to fix what it got wrong. "The rules" in this case is the tax code, which is thousands of pages and written in lawyer. You can imagine how converting that into computer code is a non-trivial undertaking notwithstanding that it's public information.
- smsm42 3y ago> a private system for the remaining taxpayers would have to cover its costs with only 10% of the revenue. Is it that hard? Americans filed about 260M returns last year. 10% of it is 26M. Let's say you get only 10% of those - since you're not the only person on the market - to pay you $100 (if you save a significant amount on taxes, paying $100 for it would make sense, won't it?) - that's $260M, or about as much as the government says their whole system, for all 100% of taxpayers, would cost. If you can do better job than the government on the efficiency (not a very high bar), and you only have to support 1/10 of the volume the government does, I think these numbers look ok for you. > You can imagine how converting that into computer code is a non-trivial undertaking notwithstanding that it's public information. It is. But there are multiple businesses who already have been doing it for years, so it looks like a solvable problem.
- AnthonyMouse 3y ago> Let's say you get only 10% of those - since you're not the only person on the market - to pay you $100 (if you save a significant amount on taxes, paying $100 for it would make sense, won't it?) - that's $260M, or about as much as the government says their whole system, for all 100% of taxpayers, would cost. TurboTax is ~$39. Also, the cost of developing the system doesn't depend on how many people use it, and those are only the development costs, not anything else a business has to pay, like marketing, billing or interest on capital. Moreover, suppose a business could actually get away with charging $100 and that would be enough to cover their costs. Why would you want this as opposed to releasing the source code, which reduces their costs and allows them to get by charging $25, while driving the market price down to that level by reducing the barrier to entry? > It is. But there are multiple businesses who already have been doing it for years, so it looks like a solvable problem. That was when you don't have a taxpayer-subsidized competitor providing an almost-as-good service for free.
- smsm42 3y ago$39 is the cheapest offer (well, beyond free ones). If you want to cover your mortgage, your charity giving, savings, maybe a little stock investments, maybe some side gigs, then you very quickly get to $100. And of course state is extra (my tax return, if I choose to file with Intuit, would be at around $120 with current prices). It's not all people, for sure, but in those 10% that need non-trivial filing, those would be much more frequent that "one salary, nothing else" fold. For the latter - the pre-baked offer would be the way. BTW, that's the cheapest offer TurboTax now does, where you do all the work. They have full-service offerings now, there it could be 3-4x from that easily. So I don't think their business is much threatened. > That was when you don't have a taxpayer-subsidized competitor providing an almost-as-good service for free. They already have the platform and 99% of stuff implemented. The only thing they need to do is year-to-year updates. True, it's some work, but as I just shown, their potential market is in hundreds of millions of dollars. Every year, recurring. I think many VCs would look favorably at such potential - and here we're talking about already established companies with all sunk costs already processed and paid off, needing to spend just on maintenance and recurring updates. I don't think it's as bad as you think it is.