As promised: long and boring thread about income vs consumption tax. TL;DR they are not as different as you might think. Income taxation distorts saving and consumption taxes are harder to make progressive in the short term, but they are close cousins 1/29
Let’s start with (national) income identity:
Labor income (L)+Capital Income (K)+Transfers (G) -Taxes (T)=Consumption (C)+Saving (S)
where T should be thought of here as taxes other than consumption/income tax that we are about to introduce.
I’ll mostly ignore foreigners 2/29
Labor income (L)+Capital Income (K)+Transfers (G) -Taxes (T)=Consumption (C)+Saving (S)
where T should be thought of here as taxes other than consumption/income tax that we are about to introduce.
I’ll mostly ignore foreigners 2/29
What would be an income tax? It’d be a tax on the left hand side or it could also be a tax on L+K (not the same from distributional and political economy standpoint, but I’ll put it aside). I’ll mostly ignore T and G and think about L+K. 3/29
What would be a consumption tax? It would be a tax on C.
Right there, from the accounting identity, you have the key insight. A tax on consumption is the same as a tax on all sources of income with a deduction for saving. 4/29
Right there, from the accounting identity, you have the key insight. A tax on consumption is the same as a tax on all sources of income with a deduction for saving. 4/29
What do we have in the U.S.? Well, personal income tax taxes labor and capital income, it includes some but not all transfers, and (used to) allow to deduct some but not all taxes. A lot of saving is done through retirement accounts & is effectively out of the base 5/29
It is not exactly a consumption tax – none of the components is comprehensive, rates vary depending on source – but it is somewhere in between income and consumption. Where exactly it is in between changed over time with reforms. 6/29
What about distributional implications of the two forms of taxation? It depends on savings. Savings may be positive (during working years or for bequests) or negative (retirees). Income vs consumption tax has implication who will be taxed at a given point in time. 7/29
If you take a broad notion of consumption – everything will be spend eventually (even if by grand-grand-grandkids), then no saving will escape consumption tax. In practice, of course, you can worry that not every use of funds will be taxed as consumption 8/29
Savings also matter for what happens when the tax is introduced. A new consumption tax does not excuse old saving (only future saving) so it will tax whatever has been accumulated in the past. 9/29
Here is another way to look at it: intertemporal budget constraint. Consumption=Labor income+Old Assets*(1+r)-New Assets
where r is the rate of return. 10/29
where r is the rate of return. 10/29
An income tax is a tax on Labor income and r*Old Assets. Right there, you see the second insight: assets are a way to transfer consumption between periods. A tax on income from assets (or on assets) distorts that. 11/29
Putting it together – income tax distorts saving. Consumption tax does not tax saving and its distributional implications relative to income taxation depend on who the savers are. In practice, the U.S. income tax system is a mix of the two. Let’s turn to implementation. 12/29
Here is a very neat paper by Joel Slemrod jstor.org/stable/2950903… that walks you through steps to go from the U.S. income tax to VAT, with various interesting taxes in between 13/29 
It is fun to start at the end: VAT. Value Added Tax, consumption tax. VA of a firm: difference between revenue and costs. VA of all firms in the production chain: value of the final output. VA of all firms in the economy: consumption+net exports 14/29
Comprehensive consumption base: Value added of all firms in the economy plus imports minus exports (tax imports, make sure total value of exports is tax free.
Straightforward implementation: measure VA for each firm, tax that (subtraction method VAT; only? Japan uses it). 15/29
Straightforward implementation: measure VA for each firm, tax that (subtraction method VAT; only? Japan uses it). 15/29
Clever implementation: tax all revenue, allow for credit if there are receipts to prove that tax was paid on inputs (credit invoice method; pretty much everyone else uses it). 16/29
It is clever, because firms now want receipts and act as tax enforcers for their suppliers. It works. Hi @DinaPomeranz! aeaweb.org/articles?id=10… 17/29
Now this is where the fun starts. Part of value added is labor, payroll. What if, bear with me, we subtract payroll and tax it separately at the same rate? Economically, it is the same – still consumption tax. But it now looks like a business tax and a labor income tax. 18/29
If you keep the business tax and use subtraction method, you have #DBCFT. If you keep both parts and introduce an untaxed allowance for workers, you have Hall-Rabushka flat tax. If you start taxing wages progressively, you have a progressive consumption tax called X-tax 19/29
Part of the cost is investment. That’s key for what’s coming next. Treating full value of investment as cost for tax purposes is called expensing. VAT effectively has expensing. 20/29
There is an alternative: depreciation. Depreciation reflects recognizing cost of investment as it wears out rather than upfront. There is a problem: it’s hard to get it right, in practice you use depreciation schedules for each different type of investment 21/29
That by itself creates issues, but there is a bigger problem: if you spread out 100% of wear and tear over X years, then the present value of it will be less than 100%. That means that the cost of investment is not fully deductible. 22/29
That effectively amounts to taxing normal rate of return (assuming that depreciation schedule is not messed up), So, by switching from expensing to depreciation, we are now in the world of income taxation rather than consumption tax! 23/29
If you take subtraction method VAT, separate payroll, and do depreciation rather than expensing, then you basically have a great discovery of “National Income Tax” (not such a bad idea, but it is really just next door from consumption tax) 24/29
This thread is already way too long, but let’s just finish with perils of tax incidence. Let’s compare VAT – a tax on consumption, to an income tax implemented as business part of VAT with depreciation plus a payroll tax. In zero interest world, they are the same 25/29
Let’s pick our favorite incidence approach: expert-augmented statutory. Consumption on consumers, payroll on workers, business tax on shareholders. Taxes are effectively the same, but that’s what you get: 26/29
To be fairer: the previous tweet shows a noncomprehensive (light on taxing services) VAT that makes it even more regressive. Still, general pattern is what you’d get from such ad hoc incidence assumptions even with comprehensive one. 27/29
I hope that this was sufficiently long and boring. I still have glossed over many details that I’m sure some people will complain about and add. 28/29
I do not have a patreon account, but if the thread goes viral, I will create one. Good night! 29/29
