Devil in the Details

Last month I wrote an article exposing details left out of a story about a family having mortgage problems that changed how most readers would have viewed the story.  Well once again I read a story, this time about a gallop poll that is thin on the details leading its readers based on only part of a whole story.

The article posted on the Discussion Board of this site was titled "GOP Losses Span Nearly All Demographic Groups".  When you click the link and read the article they paint a very damning picture of the GOP affiliation dropping between 2001 to 2009.

The decline in Republican Party affiliation among Americans in recent years is well documented, but a Gallup analysis now shows that this movement away from the GOP has occurred among nearly every major demographic subgroup. Since the first year of George W. Bush's presidency in 2001, the Republican Party has maintained its support only among frequent churchgoers, with conservatives and senior citizens showing minimal decline.

After reading this you would think the GOP are all but through.  Nothing left but the burial and service.

Ah, but it's what you have to dig to find (or in some cases information that is totally left out) that tells the real story.

Let's start with some of the details they do give.  Here is what they report for their survey methods.

Results are based on telephone interviews with 7,139 national adults, aged 18 and older, in Gallup polls conducted January-April 2009. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of sampling error is ±1 percentage point.

Margins of error for subgroups will be larger.

Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only).

In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.

Do you see any problems with the information they give, or rather don't give?  Read it over again and look closely.

Right away thinking people should pick up on a couple key details that have been left out.

1) The "7,139 national adults" surveyed lived where?  If someone where to call 7,139 adults in inner cities you would get far different results then calling 7,139 adults along the bible belt.  The break down of where these adults live is important.

2) What exactly is a "national adult"?  "National adults" do not vote, registered voters vote.  And even then they don't all vote.  In fact in the 2008 election of the 212,720,027 eligible voters only 61.7% actually voted.  So a poll of those 38.3% who don't vote is useless.

3) When was the poll conducted as in time of day?  Let's face it, those who work hard and see huge chunks of their pay taken from then in taxes are less likely to identify with large government and higher taxes as exemplified by the Democratic Party.  I will accept that Obama did trick a lot of people into believing they would get larger tax breaks under him then they would under McCain but we're already seeing evidence showing it to be untrue and let's face it, most hard working two income families did not qualify to begin with and those who were told their taxes would go up are not likely to identify with that group.  So given that, if a poll is taken while most of the hard working Americans are off at their jobs and instead you get people who are home living off government you'll see a far different picture.

The next bit of information that jumped out at me as I read the article was a link that claimed:

The GOP may have stemmed those losses for now, as it does not appear to have lost any more support since Obama took office.

Interesting.  Following that link they show a graph of party affiliation going back to 1989.  Considering the main story only focuses on affiliation from 2001 to 2009 I did feel it's important to see the whole picture.

What is most telling about this expanded data is the you can see the cycles of ups and downs within both parties where as details only comparing 2001 to 2009 do not account for this.  For instance, in 2001 the GOP was in a climb that appears to have started around 1999 when the GOP bottomed out and the range between the two parties was similar to what is shows for 2009.  Even in 2000 when Bush first won the chart shows Democrats having the advantage in party identification by nearly the same percentage difference as we are currently seeing.  Given that fact alone the current detail showing the GOP trailing when it is a known fact they did in the past and still won elections shows how insignificant this poll actually is.  And in fact it was only around 2006 of all the years from 1989 until 2009 that Republicans showed a clear cut advantage over Democrats.

So should the GOP worry about the details of this poll?  I would say a resounding no.