This question may seem a little confusing, but not when you actually stop and think about it.
Presidential elections are won by having a majority of the Electoral College votes, which are accumulated on a state by state basis. So, a poll which represents support on an overall level, a national level, may actually be misleading. It may be interesting, to a certain extent, but not definitive, election-wise.
A candidate may win the overall “popular” vote by many millions of votes, yet lose the election.
Our Constitution was specifically designed this way, to make sure that a few highly populated states could not basically take control of the entire country. The authors of our Constitution, Thomas Jefferson, James Madison, Thomas Paine, and John Adams, were quite wise in this respect.
So, even though a national presidential poll does not really reflect how a President actually gets elected, how do these polls work? How are they conducted?
First, some facts to help us gain some perspective.
Regarding national elections:
235 million people in our country are of age to vote.
There are 200 million of us who are registered voters.
Only 60% (120 million) of registered voters actually vote.
Regarding Polls and phones:
Polls are conducted solely over the phone now.
237.7 million people own a cell phone.
91% of adults have a cell phone.
Near or below poverty level people are more likely to be wireless.
45.9% of households still have a landline.
Landlines are more likely to be targets for telemarketers.
Many of us don’t see how the views of a 1,000 people or so can accurately predict the views of hundreds of millions.
Many of us think we have an idea of how the polling process works.
Well, let’s see.
Probability sampling is the fundamental basis for all survey research. The basic principle is, if selected correctly, a randomly selected small sample of a population of people can represent the attitudes, opinions, or projected behavior of all of the people from which the sample is obtained.
The key to reaching this objective is a fundamental principle called equal probability of selection, which states that if every member of a population has an equal probability of being selected in a sample, then that sample will be representative of the population.
In the case of national polls that track elections and other issues, the target audience is all adults who have a landline phone or a cell phone.
Here’s where it gets tricky. According to the Gallup polling company, “Findings from telephone surveys are based on standard national telephone samples, consisting of directory-assisted random-digit telephone samples using a proportionate, stratified sampling design. This complicated process starts with a computerized list of all telephone exchanges in America, residential and cellular, along with estimates of the number of phones these exchanges have attached to them. The computer, using a procedure called random-digit-dialing (RDD), actually creates phone numbers from those exchanges and then generates telephone samples from them. In essence, this procedure creates a list of all possible household phone numbers and all possible cell phone numbers in America and then selects a subset of numbers from that list for interviewers to call.”
Then, “Within each contacted household, reached via landline, an interview is sought with the adult 18 years of age or older, living in the household, who has had the most recent birthday. Calls to cell phones are not handled this way because they are typically associated with one individual rather than shared among several members of a household.”
You got that?
The fact is, polling companies have been forced to use a procedure like this because polling via phone is really the only practical option and many residential (landline) phones are unlisted, and almost all cell phone numbers are unlisted.
I understand the general process being used by polling companies, I’m just not sure that I agree that their processes generate a sampling of the population that is representative of that population on a national scale.