About Me

Name: RightThinking
Biography
Loading...

Create Your Own Blog Find Other Townhall Blogs

Comments

Archives

Blog Roll

 

Lies, damned lies, and Obama's statistics

You would think there's enough real statistics about our current misfortunes that no one would need to make up worse ones.  You would be wrong.  Yesterday, in kicking off his healthcare reform campaign, Obama slung this doozy:  "The cost of health care now causes a bankruptcy in America every thirty seconds."  (Speech text here.)  A few seconds' calculation is enough to cast serious doubt on this one; yet most people will accept it without a blink, except to feel the intended emotional jolt of surprise and horror at the social injustice -- which is exactly what Obama needs you to feel, to set up the rest of his speech.  So I was delighted to read an unusually thorough and clear explanation of why this statistic is wildly exaggerated -- and more surprising yet, it's from mainstream outlet ABC, to give credit where due.  This is a perfect opportunity for me to detour from my in-progress series of articles to highlight this concrete example of the Right-Thinking principles I'm talking about.  So here is our first category of Wrong-Thinking.

Crimes Against Truth:  Bad Statistics

In the ABC article referenced above, they describe the following convolutions taken to get the "once every thirty seconds" number:
  • The data is several years old, before there was a sharp drop in bankruptcies due to changes in federal law.
  • They include all bankruptcies where medical bills was given as a cause (among a list of causes), not the primary cause.
  • Beyond the above, they added anyone with more than $1,000 in unpaid medical bills at the time of bankruptcy.
  • And my favorite, they consider the death of a wage-earning spouse (even if they get hit by a bus) as a "medical" reason!
If you read my first post which defined the goals of this blog," you saw the trio of categories I use to define Right Thinking:  Truth, Sound Reasoning, and Values.  My examples will always fall into one of these three.  This example from Obama is clearly an abuse of Truth -- a type of abuse so common that its description has entered the popular language:  "lies, damned lies, and statistics."

There is a widespread suspicion of statistics, which is well-founded.  I agree that both Republicans and Democrats are frequently guilty in this area.  However, I would argue that statistics abuse is especially rampant among liberals -- indeed, "social crisis statistics" are an essential part of the liberal's tool chest, because many liberal factions are tightly tied to "advocacy" politics based on proving the existence of some horrible situation in order to get a mandate to fix it.

The problem is, as suspicious as most people are, they've never been given the tools to tell a good statistic from a bad one.  So they can never defend or repudiate a suspect assertion -- and thus have to revert back to their emotion-based like or dislike of the person giving the speech.  What we need is simply a basic understanding of how statistics are formed, and what questions to ask to determine if a statistic is valid or not.  Not a math degree -- just some basic tips.

And here it is:  This article by Joel Best, "Telling the Truth About Damned Lies and Statistics," gives delightful examples (including the worst social statistic of all time) and describes a thoughtful commonsense approach the reader can take to evaluate statistics in the media.  The most important point is just to ask some simple questions.  Obama's medical bankruptcy number is a great example where some basic questioning can expose a statistic as poorly defined and exaggerated to try to prove a desired point.   Some of the types of questions it's important to ask of any statistic include:
  • How is the category defined?  (Obama's medical-caused bankruptcies is a case in point)
  • What's the source of the numbers?  (e.g. police statistics vs. self-surveys)
  • Were the numbers generated by someone with an agenda?
  • Are there estimates involved?  How were the estimates made?
  • Does the statistic mesh with common sense and our understanding of the world?
If this topic and Joel Best's article excites you and you want to delve deeper, I highly recommend his full book on the topic that the above article is drawn from:  Damned Lies and Statistics: Untangling Numbers from the Media, Politicians, and Activists."  It's a practical toolkit to help the average thoughtful adult evaluate what you hear, and is chock-full of examples.
Email ItEmail It | Print ItPrint It | CommentsComments (1) | TrackbacksTrackbacks (0) | Flag as offensiveFlag as Offensive