Lying With Statistics

When doing some research for the previous blog, I came across a New York Times article that cited a very reputable survey [the General Social Survey] that released data showing that the percentage of U.S. households owning firearms has decreased from fifty percent to thirty five percent over the past forty years. Needless to say, the National Rifle Association and others disputed this finding, claiming that since the number of firearms held Americans has increased from something like two hundred million to over three hundred million, there was no way that the number of households holding firearms could have decreased.

And… both the GSS and the NRA happen to be right.  Why?  Because although the percentages are correct, the U.S. population has increased by more than thirty percent, and the number of households has almost doubled [families having fewer children and more single person households are why the percentage increases in population and number of households don’t correspond].  That means that eight million more households have firearms than in 1970, but that the total percentage of households having them is less, and interestingly enough the youngest households have the lowest percentage of firearms, well under thirty percent, while older Republican households have the highest percentage (which just might suggest that Republicans either have more to steal or worry more about it being stolen, if not both).

The same kinds of claims go on in advertising as well.  One truck manufacturer asserts that it has more trucks of a certain type on the road than any other manufacturer, proving durability, while its rival asserts that its trucks stay on the road longer.  The first manufacturer produces and sells more trucks so that in fact it does have more trucks on the road, but the second manufacturer’s trucks do last longer. Assuming the statistics are correct, it does suggest that the first manufacturer is misrepresenting the meaning of accurate facts.

Inflation is measured in a number of different ways, using everything from a Gross National Product (GNP) deflator to various forms of Consumer Price Indices.  In recent years, the government has also used a Core CPI, which excludes “highly volatile” goods, such as food and energy, on the grounds that these items distort overall price patterns.  That may be, but over the last twelve months the price of food has increased [according to BLS measurements] 1.6%, while energy prices have declined.  The problem, as I see it, is that overall food prices, over time, at least over the last seventy years, seldom decline, while energy prices are currently low because the economy is depressed.  As the economy recovers energy prices will rise, and food prices certainly won’t go down, especially given the drought still pervading many U.S. agricultural regions. Likewise, a number of politicians are proposing that Social Security Cost-of-Living Adjustments [COLA] be linked to a Chained CPI instead of the current CPI-W, on the grounds that the Chained CPI is more accurate.  It likely is, but that would mean that benefits would be decreased by some 3% over the next ten years, and three percent adds up to thousands of dollars for each beneficiary. The government is doing its best to present this proposed change in terms of increased “accuracy,” but it would mean Social Security benefits over time will be lower than they would have been.  Of course, opponents claim that the change cuts benefits, stopping there, and ignoring the fact that, over time, the system cannot pay those benefits unless the benefit levels are decreased or taxes are increased.

I could go on, but it all brings to mind the observation attributed to  Mark Twain Benjamin Disraeli:  “There are lies, damned lies, and statistics.”


8 thoughts on “Lying With Statistics”

  1. Bob Walters says:

    Added to the idea of using statistics in a misleading way are the studies people frequently site to prove a point. Perhaps most of groups sole reason for existence is to prove some point or other. Usually these groups have an agenda and are funded by people who seek to gain financially from the point they are trying to make. The best way to wade through this morass is to follow trail of cash to see where it originates. This will tell you if you trust what is said. As to social security the simplest way to ensure solvency is to raise the cap on deductions.

  2. Brian says:

    There are three types of lies — lies, damn lies, and statistics.

    –attributed to Benjamin Disraeli

    1. I stand corrected. That’s what I get for relying on memory. I’ve corrected the blog as well.

      1. Bob Walters says:

        Now if we could just edit our replies if we make a mistake.

      2. Brian says:

        Truth be told…I never thought of Disraeli. I thought it was Twain, too. In fact, I’ve used the same quote myself elsewhere and cited him. This time, though, I decided to look it up to see in what context he originally used it and found the following:

        “Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: ‘There are three kinds of lies: lies, damned lies and statistics.'”
        – Mark Twain’s Own Autobiography: The Chapters from the North American Review

  3. Graham says:

    Wasn’t the Disraeli version “Lies, damned lies and official statistics”? No idea of nor interest in the attribution…

    1. I looked into this a bit more and discovered that no one can find the quote in any of Disraeli’s works or papers, but that that Mark Twain attributed it to the British Prime Minister in “Chapters from My Autobiography”, published in the North American Review in 1906, where he wrote. “Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: ‘There are three kinds of lies: lies, damned lies, and statistics.'”

  4. Tim says:

    Benjamin Disraeli died in 1881. Mark Twain published his attribution in 1906. No one could challenge it.

    The Wikipedia site for Lies, damned lies, and statistics states that Mark Twain may indeed be the originator.

Leave a Reply

Your email address will not be published. Required fields are marked *