Did a New York Times article use deceptive statistics?

In my post about David Cay Johnston’s NY Times article about the top 0.1 percent, I wrote:

People will criticize the article and say it reads more like an editorial, and it does. It also uses some deceptive statistics that are designed bolster the point of the writer rather than shed any light. I give a big thumbs down to deceptive statistics. They are especially uncalled for considering that the non-deceptive statistics, like the one quoted above, are so illuminating.

David Cay Johnston didn’t like my characterization of his statistics as “deceptive.” He left a long comment on my blog. Most of his comment explains where he got his statistics from, which misses entirely the point of my comment. I didn’t say anything in the article was untrue, I was expressing my dissatisfaction with some of the statistics which were cherry-picked to overemphasize his point.

Johnston wrote in his comment, “For example? You don't cite a single example of a supposedly deceptive statistic.” The nature of blogs is that we don’t write huge essays all the time, it’s not what people want to read, nor do I have the time. But to set the record straight, this is probably the part of the article that made my deceptive statistics radar go off:

One way to understand the growing gap is to compare earnings increases over time by the vast majority of taxpayers - say, everyone in the lower 90 percent - with those at the top, say, in the uppermost 0.01 percent (now about 14,000 households, each with $5.5 million or more in income last year).

From 1950 to 1970, for example, for every additional dollar earned by the bottom 90 percent, those in the top 0.01 percent earned an additional $162, according to the Times analysis. From 1990 to 2002, for every extra dollar earned by those in the bottom 90 percent, each taxpayer at the top brought in an extra $18,000.

Using absolute dollars to compare rich and poor is an immediate red flag because it easily distorts the true picture. Furthermore, from the quote above, it’s hard to even understand exactly what he’s writing about. I presume he’s comparing the change in the arithmetic average income of the bottom 90% to the arithmetic average income of the top 0.01%, but he never tells us what these averages are nor how much they changed.

Imagine if, in 1950, the average for the top 0.01% was 200 times higher than for the bottom 90%. In that case, the top 0.01% earning an extra $162 for every extra $1 earned by the bottom 90% would mean that the bottom 90% were actually catching up, yet the way it’s phrased it appears that the top 0.01% are pulling away.

What about the $18,000 figure? Well, suppose that during a period of time, incomes were nearly flat, with the bottom 90%’s income increasing by 0.01% but the top 0.01%’s income increasing by 1%. Hardly a very big deal, only a 1% increase in income, but if the 0.01% started out with 180 times as much income, then the scenario above would equate to that $18,000 figure.

Now my numbers above are just made up to illustrate a point, but the point is that these comparisons from the article are meaningless because they tell us very little about the differences between the two groups, yet they give the appearance of the better off group being outrageously better off.

At the end of his comment, Johnston seems to be asking for an apology:

If you don't have an example of unethical conduct -- and that is what you accused me of -- but rather blogged without carefully thinking first then a corrective post would be the honorable response.

I don’t see why I need to apologize for accusing him of “unethical conduct,” because I never said he was unethical. Deceptive statistics are used all the time in the MSM (as well as the blogosphere), so they seem to be an accepted practice. Nor do I need to apologize for not “carefully thinking.” Even if true, it’s the nature of blogging that you can’t carefully think about everything you write, but in this case my original impression of the article most certainly does stand up to closer scrutiny.

Comments

[Luke Lea]

I'm not sure where I come out on this one. I like your blog a lot for its uncommon sense, and I like David Johnston's work for its uncommon information. I don't think you showed an example of his using misleading statistics so much as a case of un-illuminating statistics; but then your critique was equally unilluminating, so maybe it is just my own obtuseness.

In any case, as Per Bak showed a few years back (How Nature Works) the natural distribution of income in any market economy, like many natural phenomena (the distribution of earthquakes by magnitude for example) follows a power law, which is to say, it is highly skewed with a very tiny number of really big winners: it seems to be a characteristic of nature curiously.

Even so, even power laws can vary; not all exponential processes are equal, or something like that.

There's no question, to choose one egregious example, that free trade with low-wage goliaths like China and India has greatly lowers the average marginal utility of labor of the average Joe in America, even as it has greatly increased the average marginal utility of the last dollar of capital invested. Which is no problem except that the economists lied to the American people (Samuelson especially) when they said free trade with poor countries was going make everybody better off, including American wage workers (actually Samuelson, in the East Room of the Whitehouse, was a little more disengenuous: he said protection "had never caused wages to go up", insinuating that a relaxation of tariffs would not cause them to go down.)

Since the truth is that income redistribution is required to make everybody better off under free trade (because of "factor price equalization" in the jargon), there was an element of deception, or fraud, that was perpetrated upon the American public in this particular instance -- and, for all I know, upon the president of the United States as well. In a fit of political correctness, the whole economics profession was to blame.

Of course massive third world immigration is also operating here, plus the effect of the introduction of a lot of new labor saving technology as a result of the computer and IT revolutions: they all tend in the same direction, causing capital's share of the growing economic pie to explode, while the average worker's real hourly wage is not much different than it was back in the 1960's.

A rising tide lifts only large yachts nowadays it seems, while the average joe has a dingy that is stuck in the mud. His kids go to suck schools, his wife puts their six-week old in daycare, and, if he is lucky, he gets to die in a nursing home alone, cared for by poor immigrants who don't know him from Adam and probably can't speak English. This is America?

To get wages up we need to either subsidize them by taxing capital, or else artificially restrict the supply of labor. Since unions are out, the only alternative would seem to be a much reduced work week, with time and half pay mandated for overtime. The four day week and six hour day would be family friendly.

BTW, check out Johston's book Perfectly Legal: that .01 of one percent also doesn't pay its taxes, to the tune of some $300 billion a year. That's enough to reduce everybody else's tax bill by roughly a third, assuming he's got his facts right. Apparently he does.

[Half Sigma]

I never said I disagreed with the point of the article, which is that there is an extremely unequal distribution of wealth.

I also believe that the wealthy are getting wealthier in comparison because of the devaluation of labor with respect to capital, a trend which benefits those get their income from owning capital (the rich) vs. those who get their income from working (the rest of us).

However I was accused of falsely accusing the article of using deceptive statistics, and by that I meant a statistic which is quite unilluminating upon careful observation, but at casual observation seems to be saying something substantial.

[spencer]

I do not understand your problem. He simply says that
since 1970 the income of the top relative to the bottom has grown faster then it did before 1970. What is so difficult about that? I find it easy to understand and it is not misleding.

I suggest you show the paragraphs to a few people and see if they have problems understanding them.

Alternatively, how would you express the idea?

[Half Sigma]

I suppose it would be best expressed with a graph, showing the percentage distribution of income over time.

[Luke Lea]

One key statistic, which is hard to come by, is real take home pay per hour worked, by quintiles of the population. This is the the best measure of "the standard of living" of working Americans. It is hard to come by, unfortunately, due to the fact that most reports on income refer to family income, which may inclue one, two, three or more people working various numbers of hours (and jobs). The shift to the two-earner marriage has particularly obscured this data.