Wednesday, June 22, 2011

National Organization for Marriage Admits Flawed Methodology, Fails to Justify It

By Nick Sementelli
Cross-posted from Faith in Public Life

In the final days before the New York State Senate vote on expanding marriage rights to same-sex couples, the National Organization for Marriage (NOM) released a "flash survey" claiming to show that 57% of New Yorkers think "marriage should only be between a man and a woman."

As Dan explained last week, this poll makes the same mistake as the Alliance Defense Fund's recent poll on marriage which asked about opinions on the definition of marriage rather than the more relevant issue of legality.

But that's not the only problem with this poll. As others have already noted, its sample size is tiny (302 people out of a state population of 19.3 million), and its respondents aren't very representative of the state population, skewing older, more conservative, and more likely to be married (all demographic indicators of lower support for same-sex marriage).

Attempting to respond to the criticism, NOM took to its blog to justify its flawed methodology. Admitting the poll's sample is skewed, they rationalized that their findings should still be taken seriously because their sample matches the age demographics of nationwide voters in the 2010 mid-term elections.

NOM doesn't explain why they think a nationwide demographic is applicable to a New York-specific issue, especially when the state's demographic turnout in the 2010 mid-term elections wasn't as conservative as the nationwide average. New York exit polls showed that 28% of New York voters self-identified as liberal and 32% identified as conservative. In the nationwide House exit poll, 20% self-id'ed as liberal and 42% identified as conservative.

And, as an off-year election in a Republican wave year, the 2010 voter pool is not a very good predictor of future election demographics, particularly for the next election New York legislators will face in 2012 with an incumbent Democratic president back on the ticket in a reliably blue state.

Of course, all this raises two important points:
  1. If NOM wants elected leaders to truly act in the best interest of all New Yorkers, shouldn't they base their argument on something other than a tiny, unrepresentative sample of voters? NOM's suggestion that this skewed sample's opinions constitute a compelling argument shows that they're appealing to political calculation rather than moral principles.
  2. If NOM does want to make a purely political argument for state senators to vote against the marriage bill, they're not giving very good political advice, given that their skewed sample doesn't map onto the New York landscape very well and has questionable relevance going into the 2012 election.
As New York senators decide how to vote on this bill, it's pretty clear they should ignore this poll.

Nick blogs on religion and politics at Faith in Public Life.

No comments:

Post a Comment