Tuesday, September 25, 2012
Polling Hijinx Explained
By Jesse Merkel
Editors' note: this piece originally appeared at PolicyMic.
Two things typically happen after the major parties conventions come to an end. People that normally do not pay attention for most of the year start to pay attention to the news and polls more, and voter enthusiasm jumps up as a result. The major polls that are appearing on the daily newscasts, in the newspapers and on the internet have become incredibly important. While meant to accurately reflect the views of the nation, these polls today have unfortunately turned into a weapon.
Many major organizations have skewed their polls as of late. They do this a couple of subtle ways that most people will not notice unless they bother to do a little digging. By over-sampling Democrat voters and under-sampling both Republicans and Independents, the poll organizations have been able to make it appear as though President Barack Obama has a decisive lead over former Governor Mitt Romney.
Aside from over-sampling, they also base their samples on previous voter turnout. Typically this is done with the last major voting year, which would have been 2010. Unfortunately for all of us, most polling agencies are basing their samples off of the 2008 turnout model. In 2008, the Republican vote was depressed, while Democrats came out in record numbers. In 2010, the Republicans returned in force, resulting in one of the most dramatic turnovers in history.
Considering that the GOP has maintained relatively high voter enthusiasm, there is no reason to believe that their numbers will be as low as in 2008. In fact, according to Rasmussen, Republican Party affiliation is at an eight year high, while Democratic Party affiliation is over four points behind. While Democrats have been registering more people over the past month, it is highly unlikely they will be able to overtake the GOP's lead.
So why should we believe that Democrats will come out in such superior numbers?
Conventional wisdom says that Obama should be trailing Mitt Romney in the polls, and the fact that he doesn’t appear to be doing so means that Romney must just be that weak of a candidate. Thankfully, some people have "unskewed" the polls, showing what looks to be a much different race than what we’ve all been fed by the mainstream media.
A look at the Unskewed polls shows us a spread of Mitt Romney being nearly 8 full points ahead of President Obama. While conservatives will naturally want to jump for joy, they should first remember the wise words made famous by President Reagan: Trust, but verify.
Rasmussen shows the current party affiliation as 37.6% Republicans, 33.3% Democrats and 29.2% Independents. However, the sample from the latest Washington Post/ABC News poll shows an over-sampling of Democrats by 8%! Instead of the plus four margin that the GOP should be enjoying to accurately reflect the current voter rolls, the Democrats are instead being reflected by a plus four margin. With the data unskewed and the appropriate number of independents reflected, Mitt Romney would actually lead by a near 7-point margin.
One sample is an outlier. Two or more? That's something else.
A recent New York Times/CBS poll came out showing 49% for Obama and 46% for Romney. Again, the data was wildly skewed in favor of Democrats. Unskewed data shows Romney with a lead again, 51% to 44%. The sample for that poll broke down with 44% Democrats, 39% Republicans and 18% Independents.
Polls have constantly shown Romney leading with Independent voters by double digits. Massive over-sampling of Democrats and under-sampling of Republicans and Independents was used again to make it appear that Obama is doing much better than he normally would be.
Each polls that the website shows leads to an Examiner article that breaks down the data. Each article links to the polls and the raw data. After looking though the data, it becomes obvious that the analyst was spot on. The links are there for everyone to see. Click on the stats if you have questions and add up the percentages for yourselves. When one takes into account the high levels of party loyalty, the truth becomes apparent. Again, trust, but verify.
The ultimate question is, why skew the data for President Obama? The answer to that is simple.
Polls can be used as a weapon. The result is a de facto means of voter suppression. Those that may be hoping for a Romney victory may see poll after poll with President Obama supposedly in the lead, and believe that it’s not worth voting. Even if it were to peel off 1% or 2%, it could end up having an effect on the election. If you don't believe so, ask Al Gore if a scant few votes can make a difference.
This is not the result of some grand liberal media conspiracy. Rather, it is a time honored technique that liberal-leaning groups have used for years. This is nothing new.
Two Gallup polls that came on Thursday should have the Obama administration worried. Even though the numbers were slightly lopsided and they were polling registered voters instead of likely voters, the polls showed bad news for President Obama. One poll showed Obama and Romney tied at 47% each, while the other showed Obamas approval rating back down to 46%.
In an age when it seems like every news company and organization has an agenda, it always helps to read between the proverbial lines. When you look through the samples that each poll puts out there, the truth beomes clear.Obama does not have an unsurpassable lead, and Romney is not a weak candidate.
Posted by Brett at 12:46 AM