This article explains numerous ways in which we can be misled by measurements.

When these types of measurement results are the subject of media stories, it leads to another example of Greenberg’s Law Of The Media.

Follow this link to the USA Today story Medical expenses have ‘very steep rate of growth’

I followed the suggested action and put the following comment on the Worcester T & G comment board for a news story about health care reform:

Health care spending rose to an estimated $2.5 trillion in 2009, or $8,047 per person — and is now projected to nearly double by 2019.

So the government’s spending of $1 trillion over ten years to get some control of health care spending seems a lot smaller when you consider that we are already on a path to spend $25 trillion to $50 trillion over the next 10 years on health care.

When anybody touts a single number with the intention of getting you to gasp at how large it is, you always have to ask, ‘Compared to what?’

The headline is “NYC takes lead in setting next food target _ salt”.  In this article by Stephanie Nano of the Associated Press I find the following statement:

A recent analysis showed that for every gram of salt cut, as many as 250,000 cases of heart disease and 200,000 deaths could be prevented over a decade.

As I said in the title to this post, this must make common table salt the most deadly poison in nature.  Imagine a gram of salt causing 250,000 cases of heart disease and 200,000 deaths. If the 1 gram of salt can be ingested over an entire decade and still cause this effect, you have to wonder what the 1100 grams of salt in a single serving of canned soup can do to you.  How does anyone survive?

You have to wonder who edits or fact checks these articles.

I don’t know whether this falls under Greenberg’s Law of the Media or not.  This statement is so blantantly ridiculous that it cannot really be misleading.  It just makes you want to laugh at the author and her editors.

[See Steve's comment on interpreting the monthly changes.]

Derek Lowe of Corante’s ‘In the Pipeline’ (a drug-discovery blog) points to this graph in an article by Bristol-Myers Squibb’s Stephen Johnson, titled, The Trouble with QSAR (OR How I Stopped Worrying and Embrace Fallacy).

Lowe writes, ‘The most arresting part of the article is the graph found in its abstract. No mention is made of it in the text, but none has to be. It’s a plot of the US highway fatality rate versus the tonnage of fresh lemons imported from Mexico, and I have to say, it’s a pretty darn straight line. I’ve seen a lot shakier plots used to justify some sweeping conclusions, and if those were justified, well, then I’m forced to conclude that Mexican lemons have improved highway safety a great deal. The vitamin C, maybe? The fragrance? Bioflavanoids?

‘None of the above, of course. Correlation, tiresomely, once again refuses to imply causation, even when you ask it nicely.’

I’m sure the readers of Steve’s Politics Blog know the difference between correlation and causality but it is always nice to have an amusing refresher.

BTW, In The Pipeline is one of the many blogs in the Corante family. Check them out; you may find a few that interest you. [I stumbled upon Corante years ago and then met the founder, Hylton Jolliffe, at the tennis courts where, to my surprise, I learned he is the son of one of my tennis buddies.]

I now return you to your hard-core politics and economics.

-RichardH

The author says he believes in Keynesian macroeconomic policy.  However, he provides some critiques of that policy and only weak rebutals.

One part that he presented stopped me dead in my tracks:

Economists Harold L. Cole and Lee E. Ohanian argue Keynesian elements of the New Deal failed to stimulate growth, by failing to put Americans back to work. This conclusion is shocking when it is realized that the NIRA created millions of jobs to rebuild America’s infrastructure, e.g. roadways, bridges and schools. Their data are unambiguous: total hours worked per adult fell 18 percent during Roosevelt’s first three years and 23 percent from 1933-39, after the NIRA was passed. We could call this the New Deal paradox: an increase in federal public work projects led to a decrease in total hours worked.

In an online response to this part of his article, I posted the following:

You are aware of the passage in 1938 of the “Fair Labor Standards Act”?

This was the culmination of the Labor movement’s 100 year effort to obtain the 8-hour work day and 40 hour work week.

So the effort to cut the average number of hours worked was deliberate. The purpose was to spread the work over a larger number of workers.

I find that when a statistic makes shockingly little sense and there is no explanation for it, then usually your instinct is right and the statistic is wrong.

Isn’t it ironic that George Bush was able to overturn some of the effect of the “Fair Labor Standards Act”, and we now find ourselves in this economic mess?

I’d say the original article might be another example of Greenberg’s Law of the Media.

I also wonder how Despain reconciles his statistics with the possibly equally misleading chart shown by Rachel Maddow. Her chart may be a stronger refutation of the straw man that he put up than his own efforts to refute it.

Apparently the above graph was used on the Rachel Maddow show this past Monday. This seems to be a pretty strong statement about the history of the depression years.  I await to see what the opposition to economic stimulus can come up with to refute the implications of this graph. (I must admit the scale to the right of the graph makes no sense that I can divine.  However, pay no attention to the numbers that prove Greenberg’s Law of the Media.)

Follow this link to the Huffington Post story about how Representative Steve Austria had to retract his claim that President Roosevelt caused the depression.

As of Feb 13, 2009, I have sent an email to Rachel Maddow asking her to explain the graph. The more I view other statistics about the depression the more I grow suspicious about the meaning of the graph.

In a scathing rejoinder there was this comment:

Bill08 (not verified) says:

Uh, Joe, I know you wrote your article almost verbatim from a “Media Matters” press release (“Media still wedded to $70+ per hour autoworker falsehood despite GM’s recent statements to the contrary”), which in itself is nothing more than a regurgitation of a bunch of phony UAW claims.

But there is a complete refutation of your entire article, using empirical evidence like SEC filings and how the automakers are required to report costs by the FASB (Financial Accounting Standards Board). It is extensively referenced with footnotes, unlike your Media Matters screed.

http://www.heritage.org/Research/Economy/wm2162.cfm

The fact is that the $73.26 hourly cost as reported by GM includes only CURRENT employees, not retired ones. If you have the intellectual integrity, please read the article and attempt to refute any of its positions.

Noticing the link was to the Heritage Foundation, I could not help taking the bait. Here is what I found.

Their chart http://www.heritage.org/Research/Economy/images/wm2162_chart1.gif is as phony as a $3 bill.

Their reference at http://chryslerlabortalks07.com/Media_Briefing_Book.pdf

Shows the following
GM Assembler Hourly Rate $26.09
COLA 1.77
Total $27.86

FORD Assembler Hourly Rate $26.10

COLA 1.83
Total $27.93

DAIMLERCHRYSLER Hourly Rate $26.86
Assembler COLA 1.89
Total $28.75

The section labeled

2006 Average Labor Costs — UAW represented (per hour worked)

DaimlerChrysler $75.86
Ford $70.51
General Motors $73.26

Does not specify what other factors go to make up this number until you get to page 41.

In 2006

$29.15 (38%) went to Base wage and COLA
$20.14 (27%) Went to Health Care includes incremental FAS beginning with 1993
$26.57 (35%) All other elements
$75.86 total

I didn’t use a fine toothed comb, but I did not see anywhere “The fact is that the $73.26 hourly cost as reported by GM includes only CURRENT employees, not retired ones.”  I wonder what the commenter thinks he knows is in the 35% of the cost labeled as “All other elements”.  One also has to wonder how Chrysler calculated the 27% that went into health care. With these two pieces we are unsure of how 62% of the costs were calculated.

If you go on to read the rest of the Heritage Foundation’s report, they also mention:

The hourly benefits figure includes payments into defined benefit pension plans to provide future pensions to current workers. It also includes the estimated costs of future retirement health benefits that current workers earn today.

Read the rest of the report yourself and judge for yourself.  Of course I have just cherry picked a few paragraphs from the report.

According to corrolaries of Greenberg’s Law Of The Media, you would have to see comparable calculations for the transplant auto companies to make any judgments about these numbers.  These comparable numbers are never presented.  Instead we are to take Heritage Foundation’s word for what they all mean in the greater scheme of things.

Follow this link to the Huffington Post story where I first saw the above video

My initial comment on this story was:

I have come to the conclusion that charging Fox News with bias and using sarcasm may be warranted, but it is not the best strategy.

A better strategy might be to pretend Fox News’ (and Florida TV reporter’s) questions were serious questions and address them.

Burton let slide her factoid that the top 25% pay 67% of the taxes. He should have countered that Fox news could instead quote the percentage of income the top 25% get instead of just mentioning that they are the top 25%. In order to judge whether or not they are paying a fair amount of taxes, you have to know how much income they are making. To hide the size of their income is to mislead the public.

It would also be interesting to look at the details of the Rasmussen poll. If you ask the viewers of a network whether or not their channel is biased, this is a measure of how rabid the viewers are. It is not a measure of how biased the network actually is. The viewers of other networks show a reasonable degree of skepticism unlike the viewers of FOX.

I have edited the above quote to get the exact percentages mentioned in the video. A previous post on this blog, United States Income Distribution 1967-2003, shows income distribution, but does not answer the question of taxes versus income. The Tax Foundation has published the tax versus income data for 2002. Back then the top 25% paid 83.6% of income taxes and earned 64.7% of the income. Of course income taxes are not the full measure of taxes.  The lower income groups pay a larger percentage of their income in Social Security, Medicare, and other federal taxes than do the wealthy.  In the really low income groups people pay more for Social Security and Medicare taxes than they pay income taxes. (I know, Social Security is not a tax by some people’s standards depending on what argument you are trying to make.)

The Center on Budget and Policy Priorities has an article, Recent Tax And Income Trends Among High-income Taxpayers, that discusses the total tax picture as of 2003.

The Treasury analysis shows that the one percent of taxpayers with the highest incomes paid 34.3 percent of federal individual income taxes in 2003.  However, an analysis by the Congressional Budget Office shows that this group paid a substantially smaller proportion — 22.6 percent — of total federal taxes, including payroll, excise, and other taxes.[2]

Even this report does not take into account that higher income groups have more money left over after meeting their basic needs than do lower income people.  I think it only fair that higher income people pay a larger share of the taxes than is represented by their share of income.  The question is, how much?  Given the recent cut in income tax rates for higher income groups and yet they still pay a larger fraction of the total income taxes than they used to shows just how much of the economic pie the top earners are getting.  They must be getting more than 100% of each dollar added to the pie.

Another comment on the Huffington Post found the link to the Rasmussen Poll mentioned in the video.

I found the following statement in the Rasmussen report to be most supportive of my bias:

Other data showed that voters tended to select news sources based upon their political preferences.

It gets better.  The Rasmussen report on selection of news sources has this gem:

Given this polarized environment, we are especially pleased with the bi-partisan audience that visits RasmussenReports.com. Forty-three percent (43%) of our visitors are Republicans while 38% are Democrats.

I don’t think those numbers were typical of the American electorate in 2004 when this report was written.

In my online discussions on the Worcester Telegram & Gazette, I find that most readers find the newspaper biased to the left, whereas I find it biased toward the right. Read this letter to the editor and the ensuing comments. I see more published letters to the editor complaining about left bias than right bias.  Of course the letters that get published are not random representations of the letters received by the newspaper.  Nevertheless, when I see some of the specific complaints about leftward bias, I see some of them have merit.  Maybe the newspaper is just careless in its news analysis pieces.

Talk about Greenberg’s Law of the Media, all these numbers are perfect examples, even the ones that I write.

If you cannot detect the application of Greenberg’s Law of the Media, I will clarify by including my letter to the editor of the Worcester Telegram & Gazette.

Has the Associated Press no shame whatsoever?

On page 3 of today’s paper, you have a substantial article by the Associated Press that goes into great detail as to why polls may have large variations in results.

In this article they mention their own poll which had Obama’s lead at only 1 point.  This poll was a far outside the results of all other polls.

The article even mentions the use of weighting to “correct” raw results.

Nowhere in the article do they mention that their own weighting included weighting evangelical voters at twice the rate that they have voted in the past.

Now that they have been caught doctoring the polls, they have the gall to come out and try to obfuscate what they have done even  further.