A consumption VAT depends on identifying consumption _goods_ and differentiating them from investment goods. Rebates make it non-regressive
In contrast a consumption tax depends on identifying consumption expenditures and differentiating them from non-consumption. Progressive rates on consumption make it non-regressive
Neither is without problems. My sense is that some of both would be optimal. This would be the result of a VAT to finance a mildly progressive social insurance system (~ rebate) and progressive taxation of consumption. This latter has the advantage of steep rates on consumption of high consumption people, whihc I think are justified on marginal utility grounds.
I really like the framing on #4. I'll have to read more of your posts on this.
What would you count as public investment? R&D, infrastructure sure. Would you count baby bonus and welfare programs as well? What about defense spending?
I’ll cheat by answering that “investent” is any activity with an NPV>0 when both costs and benefit are valued at social marginal costs and revenues. So DOES a baby bonus raise real incomes in the future?
One more question, if a baby bonus raised aggregate real incomes by virtue of increasing population growth but left average incomes the same, would that count as investment?
Nice work here, Sam. Land Value Tax is a no-brainer to the extent that we can successfully disassociate land values from the improvements built upon them. This, I have found, is no easy feat, but can probably be done with enough accuracy to reach a 50% LVT on land rental values, maybe more.
I have long held that as a consumption tax, VAT should be our "second choice" for government revenue, but as you noted here, there are other revenue sources to consider before we get to VAT. So I have moved it to "third place." I have an essay coming on a Resource Rental Tax, which is essentially an LVT for minerals and resources found within the land.
You're correct to note that VAT's only real downside is that it's regressive. Well....kinda. It's not as regressive as most believe it to be because all savings and investments are future consumption and will be taxed at the same rate. If we don't look at the time value of money, VAT is a proportional tax.
So, I am less concerned with making VAT progressive than most, especially if LVT and RRT are already in place. A low VAT, levied as broadly as possible, will do what we need it to without being overly distortionary. That said, if the time value of money argument is persuasive, there is still a good case for a prebate.
You can probably push up the LVT rate in exchange for VAT. Under various models the theoretical optimum is 100%. Probably 20% is still realistic is my guess.
I guess I should define what I mean by the LVT rate more. Most LVT proposals aim to tax ~80% of the *rental* value of land. In a perfect world, we could tax 100%, but assessment is hard so it's safer to tax less.
But as I'm thinking about it in this post, the LVT would assess the land value and then ask for 10% of that annually.
Using this second definition of the LVT rate, you mostly want to stay close to the capitalization rate discussed here:
Okay, I see what you are saying, you want to base your tax on the market value of the land. I agree that market dynamics would certainly change, and if we go above the *rental* value of the land the market value would be negative.
This setup seems a little unstable to me because the *market* value of land drops as rent gets taxed away. So if you are not careful you might get weird feedback loops (?).
I think it's better to just always talk in terms of *rental* value, even if it is less intuitive.
A consumption VAT depends on identifying consumption _goods_ and differentiating them from investment goods. Rebates make it non-regressive
In contrast a consumption tax depends on identifying consumption expenditures and differentiating them from non-consumption. Progressive rates on consumption make it non-regressive
Neither is without problems. My sense is that some of both would be optimal. This would be the result of a VAT to finance a mildly progressive social insurance system (~ rebate) and progressive taxation of consumption. This latter has the advantage of steep rates on consumption of high consumption people, whihc I think are justified on marginal utility grounds.
Easy in the US at least:
1) Replace the wage tax with a VAT at rate to fully pay for whatever level of social insurance is chosen
2) Eliminate taxes on business income
3) Allow unlimited contributions to deductible savings vehicles.
4) Raise rates enough that deficits = public investent.
Seems good to me.
I really like the framing on #4. I'll have to read more of your posts on this.
What would you count as public investment? R&D, infrastructure sure. Would you count baby bonus and welfare programs as well? What about defense spending?
I’ll cheat by answering that “investent” is any activity with an NPV>0 when both costs and benefit are valued at social marginal costs and revenues. So DOES a baby bonus raise real incomes in the future?
I see, so it becomes an empirical problem.
One more question, if a baby bonus raised aggregate real incomes by virtue of increasing population growth but left average incomes the same, would that count as investment?
I'd say so. But surely we're expecting average incomes to be rising over time anyway.
Nice work here, Sam. Land Value Tax is a no-brainer to the extent that we can successfully disassociate land values from the improvements built upon them. This, I have found, is no easy feat, but can probably be done with enough accuracy to reach a 50% LVT on land rental values, maybe more.
I have long held that as a consumption tax, VAT should be our "second choice" for government revenue, but as you noted here, there are other revenue sources to consider before we get to VAT. So I have moved it to "third place." I have an essay coming on a Resource Rental Tax, which is essentially an LVT for minerals and resources found within the land.
You're correct to note that VAT's only real downside is that it's regressive. Well....kinda. It's not as regressive as most believe it to be because all savings and investments are future consumption and will be taxed at the same rate. If we don't look at the time value of money, VAT is a proportional tax.
So, I am less concerned with making VAT progressive than most, especially if LVT and RRT are already in place. A low VAT, levied as broadly as possible, will do what we need it to without being overly distortionary. That said, if the time value of money argument is persuasive, there is still a good case for a prebate.
I outline why VAT is not actually regressive here: https://www.lianeon.org/p/rethinking-taxation
You can probably push up the LVT rate in exchange for VAT. Under various models the theoretical optimum is 100%. Probably 20% is still realistic is my guess.
I guess I should define what I mean by the LVT rate more. Most LVT proposals aim to tax ~80% of the *rental* value of land. In a perfect world, we could tax 100%, but assessment is hard so it's safer to tax less.
But as I'm thinking about it in this post, the LVT would assess the land value and then ask for 10% of that annually.
Using this second definition of the LVT rate, you mostly want to stay close to the capitalization rate discussed here:
https://www.astralcodexten.com/p/does-georgism-work-is-land-really
I think they range from 5-10% so I'm worried that pushing up to 20% might create too many distortions.
Okay, I see what you are saying, you want to base your tax on the market value of the land. I agree that market dynamics would certainly change, and if we go above the *rental* value of the land the market value would be negative.
This setup seems a little unstable to me because the *market* value of land drops as rent gets taxed away. So if you are not careful you might get weird feedback loops (?).
I think it's better to just always talk in terms of *rental* value, even if it is less intuitive.
Point taken!
You may also be interested in this old post going through the distortions of auctions for land valuation:
https://splittinginfinity.substack.com/p/some-thoughts-on-using-auctions-for