« It's Not Really Clear To Me What A Blog Is For | Main | Insurance Matters »

July 24, 2007

So Are Real Average Wages Up?

This is the central contention of the Brooks' article. And yes, if you consider "lower than they were five years ago" up, then yes, real wages are up. But I don't imagine most people think like that. And as Brad DeLong notes, ask yourself why we're looking at real average wages, rather than the median wage. Could it be because "averages include--and give high weight to--what is going on at the top of the income distribution?" I imagine so. Here's how Daniel Gross explains the difference:

If Goldman, Sachs CEO Lloyd Blankfein gets a $54 million bonus, and 53,999,999 other workers get nothing, then on average, 54 million people have received a $1 bonus. In reality, however, only one person has more money in his pocket. Crudely speaking, that's what has been happening in the U.S. economy.

Now why would David Brooks want to use such an inaccurate and even misleading measure? Particularly when median household income shows a 3% fall between 2000 and 2004? And has completely divorced itself from productivity gains? Doesn't Brooks find any of that worth mentioning?

July 24, 2007 | Permalink

Comments

I posted about two other misleading points and the mysterious $10000 per year income of US households due to globalization in this diary at DKos. What is it with Brooks and 10000?

Posted by: dlcox1958 | Jul 24, 2007 3:44:18 PM

I posted about two other misleading points and the mysterious $10000 per year income of US households due to globalization in this diary at DKos. What is it with Brooks and 10000?

Posted by: dlcox1958 | Jul 24, 2007 4:03:04 PM

The usual HS stats classroom example: If Bill Gates goes into Moe's Tavern, the average/mean customer is a billionaire. The median customer is still Barney.

Posted by: Davis X. Machina | Jul 24, 2007 4:28:42 PM

Doesn't Brooks find any of that worth mentioning?

No. But not because he's dishonest or anything. Heavens to betsy, perish the thought!

Posted by: W.B. Reeves | Jul 24, 2007 4:41:48 PM

the brooks article is awful, for sure, but, looking at the wages of production/non-supervisory workers (which i guess he's doing, and, which is the focus of the WSJ piece cited by Gross) actually does generally control for inequality.

production and ns workers are essentially the bottom 80% of the wage distribution (roughly), so, haven't seen the huge gains that have accrued to the top.

this series actually tracks median wages pretty well.

that said, they also earn less than they did 5 years ago, so, there's that.

the $10,000 per household number comes from a terrible IIE study (Peterson Institute of International Economics). i've mocked another piece of the same study at the link below, and, i've got a fuller (and more serious) critique of the larger study i've been meaning to finish up.

http://www.epi.org/content.cfm/ib233

Posted by: josh bivens | Jul 24, 2007 4:45:03 PM

Now why would David Brooks want to use such an inaccurate and even misleading measure?

Obviously he shouldn't have without qualification, but who knows why he did? Sloppiness? Bad advice? Insufficient research? Why did you, the other day, say that the bottom 90% of the income distribution lost ground last year, when (as JasonR pointed out), your source was about 2005, and was only about market income, while total income apparently actually went up (a little) across the board (Jason's source)? I don't know, but it wasn't because you were trying to mislead people.

Posted by: Sanpete | Jul 24, 2007 10:01:54 PM

The other dishonesty is using time periods that include the Clinton years. I've seen right-wing hacks do this repeatedly, trusting (usually accurately) that our idiot MSM won't bother unpacking the numbers. On the rare occasions that someone does, what's invariably revealed is that all the economic progress being cited actually took place under Clinton, and that the same numbers stagnated or regressed under Bush. But most people don't see that. Fortunately, most people make their economic judgments based more on their own and their relatives/friends/neighbors' circumstances then on statistics, and thus most people seem to know that the Bush years have stunk for most people, despite the GOP's and the MSM's best efforts. Most of us aren't related to Bill Gates, after all.

Posted by: beckya57 | Jul 24, 2007 10:15:56 PM

In Sanpete's world a bee sting is the same as a beheading, stealing a pen is the same as stealing millions from the office, ordering a veal cutlet is the same as cannibalism...it really boggles the mind.

Posted by: Ricky | Jul 25, 2007 9:53:48 AM

Ricky: you just have to remember Sanpete's only goal in these discussion is to prove that he is more noble than everyone else here...

Posted by: Meh | Jul 25, 2007 10:04:43 AM

In Sanpete's world a bee sting is the same as a beheading

No, Ricky, but a lie is the same as a lie. Again, just apply the same standards all around.

Meh, I state my purposes and give reasons. A rational response would address those rather than just make something up.

Posted by: Sanpete | Jul 25, 2007 11:47:13 AM

"No, Ricky, but a lie is the same as a lie. Again, just apply the same standards all around."

But this isn't a site about tracking the lies of those on the left. When the rightwing noise machine gets worked up about something said by a lefty, then the lefty blogs take a look and generally call a spade a spade and move on or they defend that lefty if the attack is unwarranted. I imagine Ezra has disagreed with something Krugman has said on occasion, but in my several months reading this blog, Ezra has never shown himself to be about fact-checking people who lean left.

This isn't a forum for you to equate one possible Krugman misrepresentation as equaling Brooks's weekly mendacity and expect people to think you are anything but an asshole. Start your own damn blog...I have a feeling it would probably just end up being another Instapundit. An 'independent' who happens to only hold the left to any standards while defending the right's lies by saying the left does it too.

Posted by: Ricky | Jul 25, 2007 1:33:58 PM

Ricky, Krugman's egregious detour around the truth is an example, not the sum of all such problems from him or others. I see nothing in what you say that justifies a double standard. If you're going to call what Brooks said lies, then everything that meets the same standard ought to be called lies, including a good deal of what posters on both sides say here all day long. If Ezra disagrees far more with Brooks than with Krugman, or has other reasons not to want to critique both, that makes perfect sense and I don't object at all. But we still need to call both sides by the same standards. And if we apply that standard to ourselves first, I don't think we'll often be calling others liars.

Posted by: Sanpete | Jul 26, 2007 2:31:26 AM

"But we still need to call both sides by the same standards. And if we apply that standard to ourselves first, I don't think we'll often be calling others liars."

Such as it should be in life, but this is a blog with limited space and readers come here not to you derail thread on tangential points. If I wanted to read a site that excoriates Krugman for once saying something that was inaccurate or was editorial in nature, I can go to any number of right wing sites that will take that and run with it.

You always start with the assumption that we are so ideologically skewed that we can't tolerate one of our own caught in a lie or making unsubstantiated comments. We do it on a daily basis here in the comments where people not only critique Ezra's posts they also critique each other's points. I have never read anywhere on this board where someone deified Krugman or anyone else on the left. The point of this blog doesn't seem to me to have anything to do with keeping score on left vs. right pundits.

Why don't you start your own blog? I would read it, just to see how tortured your arguments where in trying to create moral equivalences for the lies on both sides.

Posted by: Ricky | Jul 26, 2007 10:42:26 AM

There is a fundamental mistake in the entire concept of average, which is both misused by tools like Brooks and also critics, like Gross. How you calculate the "average" or more appropriately the expectation value of a statistical sample is fundamentally dependent on the distribution of the sample. What Gross is calculating in the example he gives is the arithmetic mean. However the arithmetic mean would certainly not produce the most probable measure called the maximum likelihood estimate(MLE) of the expectation value. In other words, unless your data is approximately distributed according to a statistical distribution whose MLE for the expectation value is the arithmetic mean(normal, poisson, exponential), than you probably should not be using it as a descriptor for the average.
However, I should add, that the one thing the critics always get correct, is that we should be looking at the median. There is an important statistical reason why we should, its because the median is a proper measure of your sample. In the same way that the MLE of the expectation value is a proper measure of your sample.

Here's some good reading for you:
http://en.wikipedia.org/wiki/Pareto_distribution
http://en.wikipedia.org/wiki/Maximum_likelihood

Posted by: Zaid Khalil | Jul 26, 2007 11:02:06 AM

Ricky, you're making excuses for the inexcusable and, not surprisingly, it's consequently you making the tortured arguments. Nothing I've said requires more detail or space or tangential points or excoriating Krugman or keeping score, as I've clearly explained. It only requires using the same standards all around about what constitutes lying. Stop squirming and just admit that there's a double standard, and that there's no excuse for it.

You always start with the assumption that we are so ideologically skewed that we can't tolerate one of our own caught in a lie or making unsubstantiated comments.

I make no such assumption. I observe actual behavior.

Posted by: Sanpete | Jul 26, 2007 1:35:45 PM

In one way I'm greatful to Sanpete for flogging this particular dead horse. It led me to read Krugman’s essay.

Having done so, I have to say that I found Sanpete’s initial criticism laughable. After attacking Krugman for providing insufficient evidence for his thesis, Sanpete commits the same “crime” himself by citing a single ambiguous sentence by Friedman as definitive evidence that Krugman was playing fast and loose with the record.

The sentence in question, “The elementary truth is that the Great Depression was produced by government mismanagement. It was not produced by the failure of private enterprise, it was produced by the failure of government to perform a function which had been assigned to it. supposedly refutes Krugman’s characterization that Friedman was blaming activist government for the Great Depression. Of course, the validity of Sanpete’s charge turns on what Friedman meant by “a function assigned to it.”

As Krugman cogently argues, Friedman’s notion of this function was the Monetarist theory that the government should have increased the money supply rather than pursuing overt economic interventions of the Keynesian type. He points to data showing that the money supply did in fact increase during the period, with no discernable ameliorative effect on the economic crisis.

Obviously, Friedman did not equate his Monetarism with Keynesian style Government economic activism since his championing of that theory was an attempt at refuting Keynes. The suggestion that Friedman, in this single sentence, was repudiating the fundamental premises of his life’s work is, as I said above, laughable.

Since the sentence does not mean what Sanpete seems to think it means, ie, a general condemnation of government inaction and an implicit endorsement of government intervention beyond Monetarism, it really has no significance to the thrust of Krugman’s thesis. Or, more precisely, since Krugman has demonstrated that the data shows that the money supply did increase and Friedman’s contention was that the government should have increased the money supply, including the sentence in context would have buttressed Krugman’s argument rather than undermining it.

In short, Sanpete’s "evidence" of Krugman’s “dishonesty” is no evidence at all.

Posted by: WB Reeves | Jul 26, 2007 2:03:09 PM

WBR, I respond to the basic misapprehensions in your account here.

Posted by: Sanpete | Jul 26, 2007 6:26:44 PM

The comments to this entry are closed.