Why we poll likely voters
For anyone spending any time at all reading SoonerPoll.com, one will quickly notice that nearly all of our polls results are of likely voters, who by definition are most likely to turnout on election day and vote. There are other populations to consider like registered voters or the general population of Oklahoma, for that matter. So, why at SoonerPoll.com do we focus so much attention on polling likely voters?
There are several reasons but, first and foremost, we poll likely voters because some of our polling are ballot questions, ranging from candidates to ballot initiatives and potential ballot initiatives. Likely voters are the logical population for this type of polling because likely voters are those who will be deciding who are elected officials will be and what ballot initiatives will be law in the state. All major pollsters in the U.S. use likely voters as their population for ballot questions.
The mere fact that we have to poll likely voters instead of registered voters or the general population is a rather sad commentary on voting in general. Oklahoma is a state of 2.7 million eligible-to-vote citizens and only 2 million (or 74%) are registered. In last year's general election, only slightly more than a million cast ballots in the governor's race, or 52% of registered voters. This means only 37% of eligible-to-vote Oklahomans cast a vote in deciding who our governor will be.
Taken a step further, only 18.5% of ALL Oklahomans (50% of voters plus one who turn out to vote) is deciding who will lead 100% of us. Those statistics were for the more popular general election that is held every other year in November, but for most other elections held throughout the year the number only gets worse.
As a pollster, I would prefer that ALL Oklahomans voted and participated in our electoral system, removing the need to estimate turnout of a particular election and collect the opinions of only those who are likely to vote. Also, polling results become less stable and less predictable as voter turnout decrease, thus creating greater problems with using a population segment such as likely voters.
So, what about questions that are not ballot questions? SoonerPoll has and will continue to ask a host of questions on a variety of issues as a way to better understand who we are and what kind of state we want to live in. But, to accomplish this, we have to make sure the appropriate questions are being posed to the correct audience.
One reader who left a comment on this website about poll results over early childhood education said, "Perhaps parents of young children should have been polled, and a poll that reflects Oklahoma�s population." On it's face, this makes sense, but the questions dealt with how our state government should deal with early childhood education and it's funding. If we are attempting to gauge the public's opinion on policy or governmental change, then shouldn't we only care about the opinion's of those who will actually play a role in any potential policy change?
Furthermore, likely voters are also much more familiar and informed on the issues because they are engaged in the political process. We have conducted extensive research among non-voters and why they do not vote. Many tell us the lack of time is the main reason for not voting, but many do not take the time to become informed on the issues and, therefore, feel unequipped to vote for candidates and ballot issues they know nothing about.
It is not because this information is hard to find. Every major newspaper covers current political issues on a daily basis, and there are three cable news channels covering all of the issues 24 hours a day. The collection of a uninformed or pseudo opinion is bad research and should be avoided in public opinion polling.
One final reason is that voters are asked to decide ALL public policy issues, not just those who would be directly affected by some particular public policy or state law. Parents would be more informed on education issues than the general public, or business owners of the state franchise tax for example, but only letting these groups decide their issues is not how our political system works.
In Oklahoma, the electorate on election day is typically more white, more female, and much older than the population at large, but right now this is the roughly 1 in 6 Oklahomans who is making all the electoral decisions for all of us.
It is my hope that those who believe their opinions are being excluded from public opinion polling such as ours will come to realize the importance of becoming engaged in the political process, learn the issues and exercise their right to vote; do so and one day they as well may be chosen at random to give us their opinion on the issues which may shape the future of this state.
Bill Shapard is the CEO of SoonerPoll