Another WSJ lesson in how to deceive with statistics

Today's Wall Street Journal editorial on the tax code, which concludes that "[t]he U.S. tax system is already 'progressive' enough," uses a common trick: talking about the rising tax share paid by upper income Americans without once mentioning the rising pre-tax income share that they receive. Here's how the Journal frames it:

The authors found that over the course of 20 years the richest 0.1% of all taxpayers saw their overall tax share double -- to 11.05%, from 5.06%. The top 20% of all earners also saw their tax share increase sharply to more than two-thirds of all taxes paid. Meanwhile, the bottom 20% of earners paid only a tiny share in 1979 but saw even that share cut in half 20 years later -- including payroll taxes.

When you receive a disproportionate increase in pre-tax income, you're going to pay more taxes -- this isn't rocket science. (There are other changes that affect relative tax burdens, of course, but to not even mention the massive gains made by the rich over the last two decades is intellectually dishonest at best.) The paper (PDF) that the Journal cites in the editorial makes this clear:

The share of income accounted for by the top 1 percent of the income distribution has climbed steadily from a low of 9.58 percent (3.28 for the top 0.1 percent) for 1979 to a high of 21.55 (10.49 for the top 0.1 percent) for 2000... This pattern of an increasing share of total income is mirrored in the 1-to-5 percent class but to a considerably lesser degree. For this group, the income share increased from 12.60 percent to 15.14 percent in this period. The 5-to-10 percent class’s share of income held fairly steady over this period, going from 10.89 percent for 1979 to 11.28 percent for 2002. The shares of the lower percentile-size classes, from the 10-to-20 percent classes to the four lowest quintiles, show declines in shares of total income over the 24-year period.

Thus, it's not surprising that shares of income taxes shifted accordingly:

The share of income taxes accounted for by the top 1 percent also climbed steadily during this period, from 19.75 percent (7.38 for the top 0.1 percent) for 1979, then declined to a low of 17.42 percent (6.28 for the top 0.1 percent) for 1981, before rising to 36.30 percent (18.70 for the top 0.1 percent) for 2000. The corresponding percentages for 2000 for the 1-percent and 0.1-percent groups are 37.68 percent and 19.44 percent, respectively, accounting for the 2000 tax rebate, which is discussed below. The 1-to-5 percent size class exhibited relatively modest change in its share of taxes, increasing from 17.53 percent to 20.29 percent (20.52 including the rebate for the child tax credit) in the period. The 5-to10 percent class, and all lower income-size classes, had declining shares of total tax.

But you won't find any of this in the Journal - just their cherry-picked statistics.

Update 4/26: Glenn Reynolds links. See also Kevin Drum, who posted similar thoughts on the editorial. But most importantly, see the first comment below, which alleges that I'm a soak-the-rich leftist. How he reached that conclusion based on this post is beyond me -- do only leftists demand honesty in editorials on tax policy? As I've said before, I'm a DLC moderate, but this is not about my politics. Whether you care about growing inequality or not, the rising share of income going to upper income earners is a major factor in the distribution of the tax burden. Welcome to reality.

Update 4/27: Sadly, some people are having a really hard time understanding this, so here's a simple model to fix ideas. There are ten people in society. Three of them -- the "rich" -- make $100/year and pay a tax rate of 30%. Seven of them -- the "poor" -- make $40/year and pay a tax rate of 15%. Thus, the rich pay $90/year in taxes while the poor pay $42/year, which means that the rich pay 68% of the taxes measured as a share of the total tax burden.

Now, fast forward a couple of decades. Let's assume that over that period, the rich have done really well, increasing their incomes to $150/year, while the poor are doing only marginally better with incomes of $50/year. We see that the rich pay $135/year in taxes while the poor pay $52.50. Even though the tax structure did not change, the share of total taxes paid by the rich increased to 72%.

The trick the Journal uses is to ignore the changing distribution of pre-tax income and say, "The tax system has become more progressive because the rich are paying more of the taxes." But this happened in our model even though the tax code stayed exactly the same and the poor's income improved marginally. Now, in real life, the tax code obviously changed in various ways, but the fact remains that the rich now command a larger share of pre-tax national income than they did in the past. As a result, it's not surprising that they pay more of the taxes -- and it's not surprising that the Journal went to such lengths to conceal this fact from its readers.

Update 4/27: For more, see this Spinsanity post on how the progressivity of the tax system is spun (an issue I discuss below in comments).

Comments

[John]

No matter how much the rich pay, it's never enough for lefties like Brendan. I'm just glad the "objective" charade dubbed Spinsanity is over.

[Clement]

I'm not sure what you're complaining about. Does it make sense to rob Peter to pay Paul? I mean isn't that in essence what you're complaining about - that we're not robbing Peter enough?

Sure income may be rising faster for the the highest income brackets, but what does that have to do with whether or not the tax system is progressive since it is the rich who pay for government? Isn't the complaint that taxes aren't adequately progressive that the rich don't pay enough - and yet we can plainly see that they do.

That the wealthier are making much more money now than before isn't an issue if one recognizes as the Wall Street Journal does, that it's not a fixed "pie". Wealth is created - and isn't a function of redistribution. Simple economics really. So again, the WSJ, didn't ignore that, it just wasn't relevant to the issue at hand - particularly considering that if you cite statistics of overall income, you also have to cite the fact that even amongst the poor, purchasing power has been increasing - so if anything this is more a reflection if your bias.

[Murphy]

Exactly how much LESS is the poorest quintile supposed to pay?? And what are the societal consequences of having vast swaths of voters who essentially are paying no taxes? When politicians vote they vote to spend other people's money, the result being out of control discretionary spending and entitlement programs (SS, Medicare, Medicare Drug) that have us heading towards national bankruptcy. We don't need 40-50% of the population ALSO voting to spend other people's money.

[glenzo]

Lies, damn lies and statistics

As you know, the WSJ has to defend the current tax structure due to continuous attacks by Democrats trying to make it even more progressive than it is. 'Statistics' used in articles I read written by proponents of higher taxes on the 'wealthy' are just as biases as any published by the WSJ.

In any case, I am not overly impressed with using 1979 as the base case anyway. Pre-Reagan tax cuts and heading into a difficult recession so lots of the IRS statistics compare apples and oranges, my humble opinion.

Also, when anyone uses the tax year 2000 for comparison, when income taxes paid as a percent of GDP jumped to 10.1%, the highest in history, then there is something else going on here.

Although I cannot find the data [I will look harder]it appears to me that individual income taxes may have had a large component of capital gains during the bubble period. This period was inevitably followed by a downturn where some of these gains were wiped out.

My sister got stock in a high tech company that she worked for that year in which she had to pay AMT taxes on. She would have fallen into the 'rich' catagory since her income fell into the $1 million+ area. However, that was the only year she would have fallen into the rich catagory by a very large margin. In fact, the stock plummeted before she was allowed to sell. Some of the company's other 'rich' taxpayers ended up declaring bankruptcy and there were even suicides due to the financial stress.

I have been overly impressed with the the ability to use government statistics to make a claim on what the right tax policy is for either side of the aisle. But I am a firm believer that if government tax policy increases growth and in the process makes some people wealthy, then we are better off as a nation. conversely, punative tax policy on the wealthy that causes growth to moderate or decline doesn't do anyone any good, although the relative statistics may look better for those that like to see greater income equality.

[Noumenon]

I'm not sure what you're complaining about. Does it make sense to rob Peter to pay Paul? I mean isn't that in essence what you're complaining about - that we're not robbing Peter enough?

No, I think he is making an entirely neutral point. It makes no sense to talk about what share of income taxes any group pays without talking about their share of total income. For example you can't tell whether the bottom 50% is paying their fair share if you only know that they are paying 8% of federal taxes. You need to know whether they earn 4% or 12% of total income. (It's 12%, so they're coming out a bit ahead.)

Basically, the Journal is complaining that obese people have to pay more money to McDonald's than they used to, without noting that they've been eating more at McDonald's too. It's hard to trust someone who doesn't think that's pertinent information or doesn't want to reveal it to you.

[david]

Amazing how almost everyone managed to miss your point, Brendan. How's this: If you and I both make the same amount of money one year and pay a 10 per cent tax, and next year you make twice as much as I do, you will pay twice as much tax as I will. The mere fact that you end up paying more tax IS NOT punitive and it DOES NOT necessarily mean that you are paying more than you should be. So, if the rich are earning more than ever before, then it should not be surprising that are also paying more than before, regardless of whether the graduated tax is just or not. Even if the tax were not graduated but were a simple percentage of income, they would still be paying more. Does that help?

Now, to confuse things again, here's a little message from the father of free market thought: "When the toll upon heavy carriages of luxury...is made somewhat higher in proportion to their weight, than upon carriages of necessary use, such as carts, waggons, etc., the indolence and vanity of the rich is made to contribute in a very easy manner to the relief of the poor..." - Adam Smith, Wealth of Nations, V.i.5

[Clement]

Amazing how so many people just don't get simple economics - forget about free market thought.

Why does it matter how much income everyone makes? Let's break this down to a simpler example: a bunch of friends go out to dinner, each of these friends make a variable amount - does it matter how much everyone pays of the bill? With government, we want to say that rich pay more, and they do - but why should they pay a higher percentage of their income when they already pay more of the bill? Why does it matter if they make more money if it is wealth that is created (not redistributed from the poor)? Ultimately - why do we insist on penalizing success?

Further with respect to this point, if one makes a note of the rich's share of overall income - it is disingenuous not to also make a note of how all income is rising (and this isn't just inflation as this income is reflected in rising purchasing power for the poor).

A variable distribution of income should only be important if it's a zero sum gain (i.e. James wins only of Jack loses) but this is just not the case (one need only note how the west has generally speaking eradicated absolute poverty - compare this to the developing world where I've worked).

The point is not that the rich should not pay more taxes - by simple math they do - the point the Wall Street Journal was making was that tax policy need not explicitly tax the rich more with a graduated taxation system since the rich will pay more anyway - in fact, it is the rich that are taking the risks creating the innovations that create the bulk of the wealth (and this is not to say that the rich today are the same as the rich tomorrow). Thus lower taxes as a percentage of income of the rich actually has the effect of increasing the tax base - which is what the Wall Street Journal shows (there's a reason the US economy has done so much better than the Europeans and continues to do so).

[david]

All Brendan is trying to say is, even if there were no graduated tax, and taxes were merely percentages, the rich's share of the tax burden would have increased dramatically (look at the numbers), and since the WSJ does not point that out, its article is misleading. It could easily lead you to think that it is an increase in tax rates rather than a proportional increase in tax shares that reflects changing income distributions. This is true regardless of your position on the graduated tax.

[Clement]

David - I just don't get it. You state "It could easily lead you to think that it is an increase in tax rates rather than a proportional increase in tax shares that reflects changing income distributions."

Quite clearly the editorial states that the large proportion of taxes paid by the top 20% is DESPITE drops in income tax for the wealthy under Reagan and Bush. I don't see how any reasonable reader of the article could mistake that. (Read the editorial). But I completely fail to see how this is relevant.

If you include the fact that income has risen for the rich, then you must in good faith also note that it has for the poor, so as not to mislead people into thinking that wealth is created on the backs of the poor or that the wealthy are not paying their "fair share" - and yet Brendan does not do so.

For this reason I just can't figure out how it is "spin" or misleading by the WSJ. It's a problem if you are against inequality as Brendan acknowledges that he is - but it is also partisan and "spin" not to also acknowledge the betterment of even the poor over the same time frame. Is it better that we are all poor but equally poor, or broadly rich with some wealthier than others? These are the two choices history and economics shows that we have - Brendan clearly prefers the former - I wish he could just be honest about it.

[Jon Henke]

Now, in real life, the tax code obviously changed in various ways, but the fact remains that the rich now command a larger share of pre-tax national income than they did in the past.

You're absolutely right, Brendan. And yet, as the WSJ alleged, the tax code did become "more progressive". The separation between the bottom and top quintiles in 1979/1999/2001 was:

Spread 19.5% 21.9% 21.4%

That's directly from the Congessional Budget Office. You make a good point about the share of tax revenue being irrelevant, but that doesn't obscure the fact that the system did become more progressive.

[Brendan Nyhan]

Jon, I'm not sure you understand the statistics you're citing. The study he's referring to is here and the figures he presents are the differences between the "effective tax rates" (total taxes as a percentage of total income) paid by the lowest and highest quintiles in 1979, 1999 and 2001. But if you go back to the original CBO report in the series here, this is what you'll find:

Over the same period, however, the income of households with the highest income (which therefore face the highest tax rates) grew substantially faster than the income of other households. As a result, the effective federal tax rate for all households as a group increased by one-half of a percentage point, or from 22.3 percent to 22.8 percent.

The problem is that these aren't statutory tax rates; "effective" tax rates means simply the percentage of income that's being paid in taxes. Thus, when one group does disproportionately better than another, it pays higher rates on that additional income and ends up with a higher "effective" tax rate even if the tax structure does not change. We could illustrate this in the model I present above by implementing a two-tier progressive tax system.

A system becomes more progressive when the statutory tax rates paid on income increase more sharply as you move up the scale. That's what the term means. It does not become so when the rich do better than anyone else and end up paying more taxes as a result.

For more on this, see this Spinsanity post on how the progressivity of the tax system is spun, which I'm going to link above.

[Jon Henke]

I understand the point you're making, Brendan, but look at what Duncan Black wrote at Media Matters on the definition of progressivity:

Economists do not consider a tax system "more progressive" simply because high-income earners pay a larger share of total taxes. Rather, a tax system is "more progressive" if taxpayers pay a progressively larger share of their incomes in taxes as these incomes go up.

As the incomes went up, the spread between the highest and lowest quintile did, as well.

The point you are making is a structural one, dealing with the marginal tax rates. However, it's a complex tax system, and higher marginal rates won't necessarily result in more progressivity, except in theory. In actual practice (effective tax rates), the highest quintile is paying a progressively larger portion of their income than is the lowest quintile.

Again, I'm using the definition provided by Black and Media Matters.

[Noumenon]

Here's a link to that Media Matters article... http://mediamatters.org/items/200504260003. It seems like you are using Black's definition to support the WSJ editorial when he is presenting it to disagree with the editorial -- I am not understanding one of you.

[Brendan Nyhan]

Black's definition says that a tax system is progressive if the percentage of income paid in taxes increases as one's income increases -- that doesn't directly relate to the relative spread in effective tax rates between the rich and the poor, which is the statistic you're citing.

In addition, by Black's definition (which you endorse), the system has become less progressive. Consider these data from the Census department. The mean household income of the top quintile has increased from $97,133 to $145,970 in 2001 dollars. All else equal, if the system is more progressive according to Black's definition, they should therefore be paying a higher proportion of their income in taxes. But despite this 50% increase in real income the table you cite above shows that they went from paying 27.5% of their income in taxes in 1979 to 26.8% in 2001.

[Jon Henke]

Well, I did point out in my post that the increased progressivity was the result of the decrease in the effective rate paid by the lowest quintile. That is, taxes didn't become more progressive because the wealthy started paying higher percentages of their income, but because the lowest quintile's percentage decreased by more than that of the highest quintile. The spread increased, in the same sense as I could say that a 10/1 split is more progressive than a 15/5 split, even though the top rate would have been cut.

For what it's worth, you make a good point by noting that the rise in income has not come with an exactly corresponding rise in percentage paid.

[JWG]

For what it's worth, you make a good point by noting that the rise in income has not come with an exactly corresponding rise in percentage paid.

That's because it can't (at least I haven't found a way to do it without drastically increasing the tax rate spread). I pointed this out in the QandO thread, and no one has shown me how to make increasing income correspond to an equally increasing tax burden. As long as the higher quintiles increase their income more than the lower quintiles and there is a progressive tax rate, the income share will rise faster than the tax share.

[JWG]

Black's definition says that a tax system is progressive if the percentage of income paid in taxes increases as one's income increases -- that doesn't directly relate to the relative spread in effective tax rates between the rich and the poor

Of course it does. You're trying to argue that the only way for the system to become more progressive is to raise taxes. But if we were starting from a flat tax and then decided to lower the tax rate for lower income groups, wouldn't the system become more progressive? Or would you only consider it more progressive if we raised the tax rate for the higher income groups and left the others alone?

[Lance]

Brendan,

I don't want to get in a spitting match over exactly how to define progressivity along with everyone else. Though I must admit Jon Henke has at least shown your original point to be overstated at minimum.

What is more interesting is the way this fight has evolved over the years.

In the early eighties the argument was made that if you lowered tax rates on the upper brackets (at the time as high as 70%) that it would raise revenue coming to our treasury from those brackets. The argument made was that the taxes acted as a disincentive to work or invest to increase ones income, as well as lead to attempts to earn money in ways that were tax sheltered or off the books entirely. This was roundly ridiculed by many at the time.

This debate went round and round until by the late eighties the data was pretty clear, though often ignored at the time, that that was exactly what had happened. Whereupon the argument went from it was a bad idea to lower marginal tax rates (I am not arguing by the way that any lowering accomplishes this, just from the rates at the time) because it wouldn't work, to the argument that it is a bad idea because it does work. What was irritating was in order to not admit they were right Liberals (sic) also claimed tax rates were never sold on the basis of increasing the income the taxes were subject to. Talk about petty. You got to give them their due even if you decry that they were right.

Michael Kinsley was the first to make the argument I believe, sometime in the late 80's (I think.) He spent a large amount of ink and time arguing that the Reaganites were wrong because the reason the rich were paying a much larger absolute amount and share of taxes was because they were making more! Now this resulted not in him arguing that the tax cuts were a bad thing because they did exactly what it was claimed they would do, but actually seemed to be making the claim that the tax cuts were an intellectual error because, well I guess because he set up a straw man where they were advocated as causing increased tax revenues not by encouraging people to earn more and thus expose more to the tax man, but that "supply siders," or whatever the diverse group of tax cutters can be mislabeled as, believed in some kind of voo doo.

I remember reading his essay and saying to myself, oh so they were wrong because things worked exactly as they said they would! What idiots those Reaganites were to actually be right! Therefore no longer do we hear the argument that the tax cuts led to the rich paying less as was originally argued, we get to hear they pay more because they make more and that is bad because we don't want them to make more, and we still get to say Reagan and Co. were all delusional idiots because they didn't argue what they actually argued, but instead that lower taxes raised revenue because of some bizarre faith. Quite an accomplishment.

Maybe it is a bad thing that the rich pay more if it means they earn more (though exactly how it is worth arguing that if rates are cut and earnings don't go up, revenues will go down is beyond me. It seems to assume those one disagrees with are morons and why bother convincing morons?) It would be nice to point out though, that the original argument was one made by the Wall Street Journal at the time and later and that that is what they were talking about. They are crowing they were right. I doubt it ever occurred to them that you or anyone else would assume they were morons and didn't realize that lower rates only work if the income the rates are applied to go up. How else would it occur? Are you claiming that they are implying it was because revenues magically rise due to some unknown math? Talk about spin!