Discover more from The Pursuit of Happiness
Before opining on taxes, I’d like to say how gratifying it is to see names I haven’t heard from in years show up in my email as new subscribers. I can’t say I blame people for abandoning the comment section over at TheMoneyIllusion, and with your help I hope to make this comment section both friendlier and more thoughtful.
The issue du jour seems to be whether to tax at least some unrealized capital gains. Cowen and Tabarrok have done a better job than I could have in pointing to all of the real world complications that make that approach problematic. Here I’d like to take a broader view; what is the problem with our tax system that makes people look for such unwieldy solutions?
In my view, the original sin of tax policy was the decision to focus on income, not consumption. Once we started down that road, we created a system where closing one loophole would inevitably create a couple more. Yes, if income really is the thing that should be taxed, then it makes logical sense to tax unrealized gains. But income is not the right base for our tax system; consumption is what matters.
You can think of a consumption tax as a system that taxes current and future consumption at the same rate. As always in economics, it’s possible to dream up theories where this is not optimal. But ask yourself the following question. If a neutral attitude toward saving is the baseline assumption, how likely is it that the optimal tax regime in a more realistic model would tax saved income at a higher rate than currently consumed income?
In his blog and in Stubborn Attachments, Tyler Cowen has made a pretty convincing case that growth is underrated. Consider a saver that invests in a biotech company that goes on to cure a major disease. How likely is it that the investors will capture all of the benefits from that medical breakthrough? In reality, savers produce enormous external benefits by spurring more rapid economic growth. Those gains are also enjoyed by the less thrifty share of the population. If the baseline assumption of a neutral attitude toward saving is wrong, it’s likely to be due to the fact that we should tax future consumption at a lower rate than current consumption—-the exact opposite of an income tax.
The simplest way to tax consumption is with a VAT. But if the federal government needs at least 20% of GDP to fund our welfare state, then a VAT will almost certainly be inadequate. Later, I’ll argue that it might make sense to use a VAT as one part of a well-designed tax system, but I’d be opposed to adding it to our current poorly designed system.
The other approach is to tax income minus saving. Under this system, you would start with money inflows from wages, interest, dividends, rent, asset sales and borrowing, and then subtract outflows that went to savings vehicles such as stocks, bonds, commercial property and bank accounts.
[Owner-occupied housing is a tricky issue, which can be handled in several different ways. The key is to be consistent, so that the consumption value of housing is taxed once, not twice or zero times. Note that state governments already tax owner-occupied housing.]
The basic idea is that your consumption is equal to your money inflows minus your various forms of saving. A 401k basically operates along these lines, until the principle is violated when you are forced to start withdrawing money at age 73, regardless of whether it is immediately spent on consumption.
This basic approach simplifies the tax system in some respects, but not all. You don’t need to worry about short and long run capital gains, or realized vs. unrealized capital gains. Indeed capital gains don’t matter at all for tax purposes, what matters is consumption. Ditto for “depreciation schedules”. It’s all based on inflows of income and borrowing and outflows into various saving vehicles.
One advantage of this approach is that it allows the government to make the tax system as progressive as desired. Thus “it favors the rich” is not an argument against consumption taxes, although it may be an argument against a VAT-only consumption tax.
Would there be problems with this sort of system? Yes, but they’d mostly be the same problems that we face with our current income tax systems. (I use the plural for systems, as there are two personal income tax regimes and a corporate income tax.) People will try to get consumption relabeled as investment. Tax authorities will need to take a close look at what entrepreneurs are calling “business expenses”. But for most of us it would have the simplicity of a system with unlimited 401k privileges. The proverbial “income tax on a postcard.”
So that’s the theory, but how about in the real world? How do real world complications change things? I’d emphasize two broad principles:
It’s far easier to develop a sensible tax system when tax rates are relatively low. This is because the (deadweight) welfare cost of taxes does not go up linearly with the tax rate; it increases with something closer to the square of the tax rate. One way to keep rates low is to replace a big part of your welfare state with a Singapore-style forced saving model of social insurance. But if you cannot do that, and need more than 20% of GDP at the federal level, then it’s probably best to have two tax regimes, each raising between 8% and 15% of GDP.
When you discover that there are real world problems with a consumption tax, the right response is not to throw up your hands and say that we need to go back to an income tax. That approach will never work satisfactorily. Instead, the right approach is to keep the basic consumption tax framework, and then tweak the tax system to address particular problems, such as how to tax owner-occupied housing, or how to distinguish between consumption and investment. Real world solutions will always be a bit messy under any tax regime. But if you begin with the correct (consumption tax) framework, it will be easier to develop sensible tweaks.
Let’s consider the recent debate over unrealized gains. People complain that a Silicon Valley tech entrepreneur has zillions in unrealized gains that will never be taxed during his lifetime. (Even worse, perhaps never at all due to the stepped up basis at death.) So how does a consumption tax help?
Suppose the billionaire wishes to live a lavish lifestyle, but doesn’t wish to sell his stock and use the money for consumption. One common technique is to borrow money and use the stock as collateral. Under our current regime, that borrowing is not taxed. Under a consumption tax regime, that borrowing is taxed as consumption unless the money is put into another investment vehicle.
Again, there would be tax evasion. But that’s also true of our current tax regime, which faces thorny issues of what to do when businesses try to disguise consumption as a business expense. My view is that the tax authorities should err on the side of viewing things as consumption when there is any doubt as to which category it belongs to. I say that because under the current regime they err too much in the other direction, most notably when they deny that consuming food at a business lunch is consumption. There is literally no purer example of consumption than consuming food.
We’d all benefit from lower tax rates if we move to a more sensible tax regime, especially one with an expansive definition of consumption. Compare marginal tax rates in the US and Singapore if you wish to see the advantages of an efficient tax regime.
To summarize:
Collect roughly 10% of GDP with a VAT, possibly with a rebate so that the poor don’t have to pay any taxes on the first $25,000 in consumption. Collect 10% to 12% of GDP with a steeply progressive tax on money flows into consumption.
Then call it a day.
PS. States like Texas and Florida seem to get by fine without a state income tax. But I’ve always wondered why states that favor income taxes feel the need to have their own separate (and complicated) income tax system. Why don’t they tell taxpayers “send us a copy of your federal forms and then pay us X% of what you paid the IRS”? Are there any states that do that?
Subscribe to The Pursuit of Happiness
Nostalgia for the Neoliberal Era
"Why don't" states have a simplified tax scheme? Because tax policy is another source of governmental power, used to provide incentives to change public behavior. It's just another tool, along with various regulations and laws and fees. It's much the reason that the federal tax system is so complex, and it's such a political challenge to do tax reform. Everyone wants the whole system to be simpler ... but then each particular exception is of great value to some narrow special interest.
And there are more benefits to this approach. A progressive consumption tax could (and should) go negative on the low end, eliminating the need for all of the distortive and complex subsidies that litter our welfare state (like Section 8 housing), and also could (and should) go above 100% on the high end, so billionaires burning millions of dollars of resources on vanity space trips could do a little more to help others. A progressive consumption tax is a simple and efficient way to bend the consumption curve into a more socially desirable shape without creating perverse incentives and with a minimum reduction in productivity. Our productive billionaires can continue their work unimpeded while the consumptive ones can start paying their fair share.