Recent Thoughts On Taxes: In Praise of “Regressive” Taxes

I’ve been reviewing some of what has been said about taxes over the years. This includes:

The Flat Tax (1985), by Robert Hall and Alvin Rabushka
The Flat Tax Revolution (2007), by Steve Forbes
The End of Prosperity (2008), by Arthur Laffer, Steven Moore and Peter Tanous
The Return to Prosperity (2010), by Arthur Laffer and Steven Moore (these two books have long sections describing Arthur Laffer’s proposed tax plans).
A Fine Mess: A Global Quest for a Simpler, Fairer, More Efficient Tax System (2017), by T.R. Reid
The FairTax Book (2006), by Neil Boortz and John Linder
FairTax: Answering the Critics (2008), by Neil Boortz and John Linder
Toward Fundamental Tax Reform (2005), by Alan Auerbach and Kevin Hassett, eds.
For Good and Evil: the Impact of Taxes on the Course of Civilization (1993), by Charles Adams

These things have been discussed for a long time, and I think many people have got the jist of what is in these books. Instead, I would like to focus especially on a few things that are not in them.

These proposals have a distinctly U.S.-centric air about them. They were developed according to U.S. traditions, precedent, intellectual trends and existing institutions. One of these principles is the idea of progressivity — that is, that lower incomes should pay a lower rate of tax. I too have advocated this idea, suggesting that one of the best things that Republicans could do is have lower taxes on lower incomes — “don’t take their money.” And why not? The lower 50% of tax filers, ranked by adjusted gross income, only produce about 4% of income tax revenues as it is today. We could exempt them from income taxes altogether, and only lose 4% of revenue. “Flat tax” proposals have this progressivity in the form of a large basic deduction. In effect, there are two rates: 0%, and the “flat tax” rate of perhaps 18% or so. “FairTax” proposals have it in the form of a “prebate,” which is suspiciously similar to a “universal basic income.”

In practice, lower incomes do pay taxes, in the form of payroll taxes, sales taxes and excise taxes, and they are significant — especially when combined with an additional 10% marginal tax rate on low incomes. Today, these have been moderated somewhat by the Earned Income Tax Credit, which basically offsets some of the taxes paid in payroll taxes, sales taxes and excise taxes. It makes the overall tax system more “progressive.”

But, after reviewing all of these proposals, two ideas come to mind:

1) The potential benefits of “regressive” taxes — higher rates on lower incomes; and

2) The potential benefits of “proportional” taxes — the same rate on all income, from the first dollar earned.

If we look around the world at what governments are actually doing, we find that seemingly “regressive” taxes are quite common. Much of developed Europe has VAT of 15%-20% plus a payroll tax of perhaps 20%-30% plus some income taxes on moderate incomes — while corporations are taxed at perhaps 22%. The East European flat taxers commonly have marginal tax rates of 10%-20% on high incomes, while lower incomes face VATs of about 20% plus payroll taxes of about 30%.

And … maybe this is a good thing. Why?

Mostly, it is a matter of simple expediency. To generate the very high levels of revenue/GDP common today, in excess of 25% and sometimes in excess of 40%, you are going to have to collect a lot of tax. It is simply not possible for “the rich” — the top 1% of income — to pay this. Even setting aside the fact that the top 1% of income is not really “rich,” but more like upper-middle class, it should be easy to see that 1% of the people can’t pay 40% of GDP in revenue; or, in any case, they won’t, since to even attempt this implies very high tax rates, which all of history says are not paid.

High tax rates create economic disincentives — there is a lot of economic damage. There has been a lot of talk about the disincentives of marginal tax rates. But, what if the overall level of tax (revenue/GDP or tax payments/income) is high but the marginal rate is tolerably low? For example, let’s say that you paid a 40% rate on the first $40,000 of income, and then a 20% rate above that. For someone making over $40,000, the total tax liability (payment) is high, but the marginal rate is not so high, and thus not so much of a disincentive to payment.

The second question is: where does the money go? For the most part, it goes toward government services — services that would have to be purchased in the private economy, if the government didn’t provide it. These are some services used by nearly everyone — mostly education, health care and public pensions (Social Security) — combined with a “safety net” of needs-based welfare programs which, even though many may never use them, they would perhaps be willing to pay for them, as a form of insurance.

These programs are often conceived of as being a sort of government business, with “contributions” that are separate from tax payments. The Japanese system, for example, bundles all social programs (health, pensions, welfare) in a separate budget segregated from the regular national budget. They are “off budget.” The U.S. Social Security program used to be the same way, until it was moved “on-budget” 1968. The payroll tax which supposedly funds Social Security is inherently “regressive,” applied to the first dollar of income with an upper limit, producing a “marginal rate of zero” for higher incomes. Similar structures exist in most developed countries. This does not seem to bother any of the leftists.

These programs still have an element of “progressivity” simply because everyone gets roughly the same benefits, but the amount that they pay is proportional to income. In the private market — private health insurance for example — the same services have the same cost. The tax equivalent would be a poll tax.

I won’t go too much farther with these ideas today, but it has some implications for our previous proposals. For example, what if you had a “fair tax” (a universal sales tax), but NO “prebate”? For one thing, that would allow you to set a lower rate. In today’s system, what if you bundled all Federal healthcare and welfare-related into the payroll tax, with an upper limit to income much like today’s $127,000? Obviously, the tax rate would have to rise. But, this is more-or-less the system in use in many countries today. The associated idea is that we are all paying for certain government services, just like we might pay for car insurance in the private market. People with higher incomes pay more; but, there is an upper limit to how much we would ask them to pay. A person with an income of $10 million doesn’t have to pay effectively $3 million for government healthcare.

In general, all discussions of this sort tend to flounder on the need to raise very large amounts of money. You are going to have to tax a lot of people at uncomfortably high rates. There has been a tendency to invent wonderfully perfect theoretical tax systems, that are applied to today’s laughably imperfect spending programs. This doesn’t make much sense. Everything tends to work out much better if we assume revenue/GDP levels under 20%. Then, it is easy to design systems with tolerably low rates. These tend to imply the use of “provident fund”-type systems, for example, the private account pension programs used by over thirty countries today, including Hong Kong and Singapore. This, in turn, brings up the question as to the difference between a mandatory contribution to a privately-owned account that holds real-world financial assets, as opposed to a tax. I think they are indeed different, and are treated differently by taxpayers. Singapore’s mandatory contribution rate is 37%, into a privately-owned account that can be used for housing, healthcare, education and retirement savings. The overall effect is very different, I gather, than if Singapore applied a 37% payroll tax, and then provided these government services in return.