Meta-Blog SEARCH QandO BLOGROLL QandO Recent Posts The Ayers Resurrection TourSpecial Friends Get Special BreaksOne HourThe Hope and Change Express - stalled in the slow laneMichael Steele New RNC ChairmanThings that make you go "hmmmm"...Oh yeah, that "rule of law" thing ...Putting Dollar Signs in Front Of The AGW HoaxMoving toward a 60 vote majority?Do As I Say .... QandO Newsroom Newsroom Home Page US News International News Blogging Regional Publications News Publications
In Praise of Regressivity
Posted by: Dale Franks on Monday, January 30, 2006

A QandO reader, Logan Boettcher, drops the following on me via email:
I consider the pure consumption tax to be a proportional tax on the tax side of the equation. Rich people save more and it would seem to be a regressive tax as the poor would spend the full 23% of their income on taxes and the rich would pay less. But the purpose of saving is for future consumption, so eventually every last dollar of the rich person's would be spent and would be taxed at the 23% rate. But even in that static, annualized analysis, I'm not entirely sure if the consumption tax is really that regressive on the annual basis.

I base this on the assumption that taxes are levied for the purpose of spending (revolutionary, I know). When they get spent, the taxpayers receive either a direct governmental transfer (food stamps, EITC, etc.) or a government service or benefit (national defense, clean air, etc.). This would mean that X=taxes paid, Y=direct government payments, Z=value of services provided by government, then the adjusted tax burden would be % Tax = (X - Y - Z)/(Income). We already see this calculation in discussions of Earned Income Tax Credits, when poor people are said to have negative tax burdens. But what of the Z part of the equation?

Let's say that police service was the only government service provided in Smallville and it was funded by a 10% consumption tax. Poor person A makes \$10,000 and spends it all, paying tax of \$1,000 and having a tax burden of 10%. Rich person B makes \$1,000,000 and spends \$500,000, paying tax of \$50,000 and having a tax burden of 5%. Aha! says a liberal, that is a regressive tax.

But if we adjust the tax burden to allow for the value of the government service, then things get more interesting. Let's say, for the sake of argument, that police service is worth \$2,000 to each person equally. The adjusted burden for person A is then (\$1,000-\$2,000)/\$10,000 = -10%, while it would be (\$50,000-\$2,000)/\$1,000,000 = 4.8% for person B. In fact, for the tax rates to equalize, police service for A would have to remain constant while the service for B would have to be valued at \$600,000! What the above calculation really says is that the rich person is subsidizing 48 poor people making \$10,000 to provide them with police service.

The point that I'm trying to make is that the value of government services and direct welfare payments influence poor people's effective tax rates to a far greater degree than a rich person's and does it in a progressive way.
It's an interesting idea that Logan proposes, but it doesn't really touch the question of regressivity, as it's commonly understood.

The big problem here is that we don't know, to use the example above, that police service is actually worth \$2,000 for each person. All we know, empirically, is what police service costs. The value that each person places on police service is pretty nebulous. As I explain in my book, SlackernomicsBook, Ebook)—(yes, it’s a shameless plug. It's called Capitalism. Look into it.):

Let’s say you own a 1985 Yugo that you want to sell. You decide you want to sell it for \$40,000 and not a penny less. When prospective buyers start coming around, none of them are going to meet your price because your price is not a true reflection of the car‘s value. Not even if you’ve maintained it really well.

The only way anyone is going to pay you forty grand for fifteen-year-old Yugo is if they are incredibly stupid or certifiably insane. The value of the Yugo is probably a lot closer to \$200, and sooner or later you’ll have to lower the price to that level, or decide to let it just rust away in your driveway.

Either way, you’re not getting the 40 grand.

The difficulty of being the seller or producer of any good is that you have to try and correlate its price with its value. Some people may be willing to pay \$500 for the Yugo. Some will only pay \$100. The value of the car is kind of vague. On the other hand, the price you put down as the seller has to be specific. You have to decide on a price that most potential customers will be willing to pay.

It’s the same for the producer of any good. First you have to add up what it costs to make the goods you produce. Then you have to try to determine the value that customers will set on your goods—in other words, how much they’re willing to pay. The goal is to set a price that is greater than the cost of making the goods, and is consistent with the value customers will put on them.

Let’s take a look at new cars and see how pricing and value interrelate. Let’s say General Motors makes the Chevrolet Caprice and the Pontiac Bonneville SSEi. The Caprice costs about \$18,000. The Bonneville costs about \$36,000. For all practical intents and purposes, they are the same car from GM’s viewpoint. They use the same frame and body. They use many of the same parts. The workers who make them and the plants in which they are made cost about the same to run. The Bonneville is slightly more expensive for GM to make because it comes with more expensive interior appointments and a racier engine. Other than that, the difference between the two cars is mainly cosmetic. They are the same car. But the price of the Bonneville is more than double that of the Caprice. Why?

Customers value the two cars entirely differently. Customers see the Caprice as the car their grandfather drives. It’s sedate. It’s boring. It’s respectable. It’s middle class. It comes in tan, gray, white or black. It is utterly unexciting.

Aaah, but the Bonneville! A 225-horsepower engine! Excitingly chunky leather seats! A heads-up display that projects driving information right into your field of vision! A dashboard covered with knobs and dials and pretty lights! Flashy colors like British Racing Green! This isn’t your grandfather’s car!

So, a customer who would pay \$16,000 dollars for the Caprice might do it as if he was doling out vials of his own blood. But that same customer would take one look at the Bonneville and toss out \$36,000 in rolls of C-notes like a drunken sailor on Singapore shore leave just to drive that baby home.

The image the Bonneville provides adds a value that translates to an extra couple of thousand dollars of profit for GM. They can jack up the price of the Bonneville simply because people are willing to pay more for it. Between the Bonneville and the Caprice, the Bonneville has greater value.

Value, at the end of the day, is pretty nebulous. And it's even more nebulous when you remember that different persons place different values on the same item. So, you can't just wish the value problem away by saying, "for the sake of argument, that police service is worth \$2,000 to each person equally," because it's just not.

All we can really do when we talk about regressivity is to look at things that can actually be measured, which in this case, is a person's actual income, and the amount they pay in taxes. In the proximate example, if the poor person pays 10% of his income in taxes and the rich person pays 5% of his income in taxes, then the tax is regressive. Suppositions about value are just that: suppositions. And unprovable suppositions at that.

But, despite the aggressively un-PC nature of my position, I don't care about regressivity. I just don't. Yes, I know that we are supposed to bleed for the poor, and exempt them from taxes to the extent possible, but I don't. Moreover, I didn't care about regressivity when I was a Security Policeman making \$700 a month, either.

Because what happens when you start programming progressivity into the tax system is that you inevitably begin to disconnect the cost of government services from the price the electorate pays to receive them. What you end up with is a system very similar to the one we have now, where the top 1% of income earners pay 37% of all income tax revenues, while the bottom 50% of income earners pay 4% of all income tax revenues.

As soon as you disconnect the price paid for government services from their actual cost, you've automatically created a system where the electorate can constantly vote for ever-increasing government spending, secure in the knowledge that they aren't actually paying for it. I can't think of a better system for incentivizing the electorate to increase the size and scope of government.

You know, the price mechanism is pretty darn important. It forces people to make decisions about what they want, because it makes the cost of getting what they want abundantly clear. That's why it works. Even Karl Marx and Friederich Engels wrote about how important the price mechanism was for making economically rational decisions. But progressive taxation inevitably removes the price mechanism from spending decisions. For the majority of the electorate, it pushes the costs onto third parties. So, as far as I'm concerned, regressive taxation is the single best available method for forcing the electorate to balance their desire for increased government programs with the cost of those programs.

As far as I'm concerned, the best way to incline me to support a tax program is to convince me of its regressivity. Regressivity is a good thing.

Let the flaming begin.