Lying with statistics

This was also posted on ConyersBlog on 7/27/06 @ 3:58 PM. The context was a response to a post about the minimum wage increase. I used the term flip flop to point out a hypocrisy.

More about Ohio and the Minimum Wage. The organization I refer to here has also made essentially the same arguements about the changes to the national minimum wage, which the Republicans have flip flopped on for political gain. You may recall that I have commented on bogus use of statistics, per the quote widely attributed to Mark Twain, that "There are lies, damn lies and Statistics”. Here is an example of right wing use of statistics.

First, the issue is an increase of the Ohio minimum wage from $5.15 to $6.85, effective January 1, 2007 if the law gets passed. By my calculations with my trusty Windows calculator, that equals an increase of $1.70, or again by my Windows calculator, a 33% increase in the minimum wage. Multiplying a standard 40 hour week times 2080 hours, assuming 50 or 51 week's work, depending on if the employer gives the employee one to two week's vacation to a full-time employee. This comes out as a wage increase to this minimum wage earner of $3536.00 per year. Now, a right wing "research" organization called the "**Employment Policies Institute" uses some kind of statistical ruse (usually some kind of lumping or averaging study, which uses means or medians to shade the argument to the point under discussion). I did not have time to wade through the entire document, but this organization made a whole bunch of bogus conclusions.

Among the bogus conclusions:

1. The "average" family earning under $15,000 per year will see an "average income distribution" gain in their income of $63.00 with an income increase of $3536.00 per year. The "average" family, even though it is not stated, probably includes the elderly who do not work, the unemployed, the chronically underemployed, and some others I haven't figured out yet, which is intended to skew the results in the direction the authors of the "study" wished to take the reader, and the politicians who will vote on it. The "Average inccome gain" is listed as $1034.00, again skewing the results by lumping in part time and unemployed workers with full time workers.

2. The majority of the families involved with the increase of the minimum wage are not "poor". They mention a family income of $45,000 which is not, by definition of the poverty line as "poor". This ignores the size of the family, number of children in college or saving for college, the fact that some of the wage earners are children who are on summer break from high school or are working on stipends at college, which are limited to 10 hours per week. Again a lumping of a class into an "average" that fails to look at individuals and the impact on their personal situation. (Again, the "averages are skewed by lumping in disparate groups, this is the only way to come up with the conclusions of this study.)

There are other erroneous and misleading uses of statistics in this bogus study. My qualifications to evaluate this "study" are 6 courses in Statistics, 2 at the graduate level, and at least 4 and I think 6 courses in economics, 2 at the graduate level.

**Did'ya ever notice that Right Wing "Think Tanks" are always named some pretentious name like "Institute" or "Research Foundation" in their name, Cato Institute, etc. ad nauseum. This puts a veneer of intellectualism and respectability, such that the right wing pundits and blowhards like Rush, Glenn Beck and Hannity can say to the effect, "these guys are (you pick, scientists, economists, experts) and they have studied this for years and they are much smarter than I am, with the implied "well, if they are smarter than I am, then they must be a hell of a lot smarter than you are, and must be believed". I do not need to mention to you that a study that has a predetermined outcome is not a study, it is propaganda. The "study" is here, click on the Ohio study. Employment Policies Institute

An addition to my Conyers Post:

As an aside, our local morning drive time radio station, which puts right wing talkers on (Beck, Rush, Hannity, and a couple of locals) runs a talky moring show with a strong right leaning focus. This morning, a spokesman from the Employment Policies Institute, Mr. Mike Flint (not sure of spelling) spewed all of the same talking points. The minimum wage increase would cost jobs of the most vulnerable. The poor people, who generally pay no Federal Income tax would be better off with an Earned Income Tax Credit (EITC), than they would be with a raise. How can a gax break for someone who pays no tax be considered to benefit the person who gets no money back or who pays no Federal Income Tax anyway?

They brought up that only only 3.8 per cent of a class of female employees would benefit. They indicated that the poor should go to school to increase their "skills" the word "skills" was repeated 5 times (technically, one use of the word "skills" and 4 repeats.

The statistic of "total family income" was mentioned, and as I said above, this is a lumped statistic, that skews the result in the direction the authors of the bogus study wanted it to go.

Then they also said the majority of those families with minimum wage earners were not poor, again a conclusion that used averaging without looking at individual groups within the universe which they were averaging. It would be interesting to look at the raw statistics, whether the distribution was bimodal or trimodal, which I suspect. Multimodal distributions reflect lumping disparate populations into a single whole.

Then they said that the way out of poverty is to have a job with full time employment, but it did not matter what the wage was, just that their income would increase with more hours worked. That one deserves the proverbial, duh. This was linked to the fear mongering that if the minimum wage was raised, the loss of low skill and entry level jobs would dry up for those most in need of them.

Then they trotted out the old saw of "perspective". They actually said that poor people in America were much better off than poor people in other countries. This is no standard by which to measure poverty in America.

Let me ask the question: Would hotels and motels quit hiring house cleaners if the minimum wage increased: Would car washes and hospital house cleaners not continue to work? Thie answer is no. Would there be some minor impact on the economy and some marginal price increases for these services? Of course there will be. Will it equal a 100 per cent increase in the price of fuel that we all have to absorb? No.

The working poor have to buy gas as well.

Since perspective is a technique used to sell weak arguments, I will try to use it here to justify a strong argument. When I was a kid growing up in the 1950's and 1960's my first job as a box boy in a grocery store paid the minimum wage of $1.40 per hour. Candy bars cost a nickel. Gas cost about 27 cents per gallon. My minimum wage would buy 28 candy bars or nearly 6 gallons of gasoline. A minimum wage of the proposed $6.85 per hour will buy 12 candy bars at the current price of 55 cents, which our local stores and gas stations are charging. 28 vs 12, this is a big difference. That $6.85 will buy a little over 2 gallons of gas at the current price of $3.00. The $1.40 would buy a little over 5 gallons of gas in 1965 vs 2 gallons today. The current ,minimum wage of $5.15 will buy 1.7 gallons of gas.

The buying power of the minimum wage has dropped by well over half in the 31 years since my first job. The new Ohio minimum wage will buy 3 candy bars. Can't American business let their employees buy 3 more candy bars for their kids?

Finally, it was clear that this was a scripted conversation, and it is also clear that this is the official position of those groups who benefit by keeping the minimum wage where it is. They have a louder political voice than the working poor.

QuestionItNow Blogs