If you read it on the internet, it must be true

Dusting off the old blog again.  This time, I got called out for calling out the verisimilitude of an article on breitbart.com regarding the declining unemployment rate and newly elected republican governors.  The article is here. It's a two minute read, I encourage you to check it out before continuing.  The "calling out" is here:

I'll begin by restating that I don't doubt any of the unemployment statistics given.  I searched unsuccessfully for the numbers because I wanted to see ALL of the unemployment stats but I have no reason to believe the ones presented aren't correct.  I do however have numerous problems with the presentation, analysis and conclusion.   

1. Flawed- The article argues that the 17 states are reducing unemployment 50% faster than the rest of the country.  The U.S average decline is .9%.  Add up the total percentage declines and divide by 17 and you get 1.35%, 50% better than the country.  If you can't see how this math is wrong, I'll offer up a baseball analogy: Say your first-baseman has been slumping all season; he is batting .200 (20%).  At some point mid-season, you bring in a pinch hitter who in his first and only at-bat gets a hit.  What is the collective batting average of your first-basemen?  By the article's math, your first-basemen are collectively batting .600 ([.200+1.000]/2).  Obviously wrong.  When I look at the list of states, I see three states that are squarely in the top half of the country by size (Pennsylvania, Florida & Ohio with possible inclusion of Michigan and Georgia).  So what we have is a collection of states dominated by the little guys.  Maine anyone?  And while those three big states did have big declines, they were already outliers with high unemployment that were reverting to the mean.
2. Lazy- If you click the link in the article to the examiner.com you'll see that this is merely a summing up of another article with most of the same problems that I am pointing out here.  The breitbart piece does manage to omit the performance of the new democratic governors.
3. Incomplete- What would make this story stronger?  Maybe the whole story.  Whatever that is.  How about looking at all the states?  States with incumbent republicans and democrats or where new leaders have taken over for the incumbent party.  How about just a glance at the numbers from the other side?  Unemployment fell in all states with new-party governors except one.  I would also point out, from the numbers provided, that 29% of the new republicans have seen unemployment drops below the national rate while democrats are a bit worse at 38%.  Clearly electing a republican over a democrat is not the sure shot to reducing unemployment... if the governor or their policies are indeed what is driving down unemployment.  Which leads me to my last point...
4. Correlation does not prove causation-  It's the cardinal rule when using statistics.  Just because X frequently appears with Y, does not mean X causes Y.  But read the last line of the story: "This is yet another example of how the so-called “blue state” model is not working." What ever that model is.  What exactly is the red-state model that IS working?  I think the issues here are a bit more complicated than an imagined cookie-cutter model of governance.  

The story certainly provides a neat and intriguing headline and a seemingly concise take-away, just as long as you don't think about it too much.