«

»

Jun 05

Print this Post

Polls’ history says June polls don’t mean much…?

Gallup is recognized as one of the most prominent pollsters and is relied upon by many political analysts.  On its website, Gallup posted a summary of the history of its presidential poll results.  Since 1980, Gallup has identified the eventual presidential winner (via popular vote) only twice in its June horse race polling and even going back further does not add much more success.  Since we are now in the month of June, many political analysts are saying that current polls do not mean much. Such analysts seem to be taking the position that polls conducted five months before the election are only a snapshot in time, and are not a reliable predictor.

 

So is this reasoning valid?  Isn’t this also true in July, August, etc.?  What is unique about June?  Almost everything can and does change during the course of the Spring, Summer, and Fall that alters the presidential race (e.g., economy, international crises, scandals, etc.).  Undoubtedly, the closer to the election, (i) the less likely an unexpected event will occur and (ii) the more likely more have made up their minds on who they are going to vote for resulting in more reliable poll results.  But this does not completely explain the situation.

Gallup has been conducting polls for at least 70 years.  Initially, Gallup polled from all “National Adults”.  Beginning in the 1950s, it began polling from Registered Voters.  It was not until the mid 1990s that it began also polling from “Likely Voters”.  Looking more closely at Gallup’s methodology and results suggests that the analysts may have missed the mark entirely in blaming the month of June.   The answer may lie within the small text about who was polled that practically no one reads.   Since the 1950s, Gallup’s June polling is based on Registered Voters.  Historically, Gallup does not begin polling Likely Voters until around September/October.   A recent post here expands on differences between Registered Voters versus Likely Voters, and which polls are likely more indicative of actual results.  Had Gallup conducted polls of Likely Voters in earlier months (e.g., June), political analysts may be singing a very different tune about June polls (even though these polls are also static).

About the author

Nicole Liska

I am a Principal at Fulcrum Inquiry, an accounting and economic consulting firm that performs damage analysis for commercial litigation, forensic accountings, financial investigations, and business valuations. I hold an ABD and MA in economics from the University of California, San Diego. I perform damages analyses and serve as a damages expert witness. My resume is on Fulcrum's website.

Permanent link to this article: http://betweenthenumbers.net/2012/06/polls-history-says-june-polls-dont-mean-much/

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

*