Polling is one of the most useful tools for politicians. When running a campaign, many politicians hire pollsters to work for the campaign. These campaign pollsters help the politicians measure the campaign’s effectiveness and what the candidate needs to focus on throughout the campaign. Campaign pollsters are most often associated with big statewide or national races, like a senator or a president. However, all elections can have campaign pollsters, even local town or city council races. If you are thinking about starting a political career, it is essential that you have a basic understanding of the world of political polling.
Like all pollsters, a campaign pollster measures a random population of voters with a set of questions. These questions are valuable to ascertain if particular methods on the campaign trail are working. The random sample is chosen within the particular geography chosen. So, for a presidential, governor, or senate race in a state, the whole state may be chosen with a particular random pool. National polling data, then, comes from a random pool of voters from across the country.
To get this population, the most common tool that pollsters use to find out what voters think of a candidate or policy is through direct calling. Random households are called often from a random phone number list from a particular region selected by computer. This is called random digit dialing.
The heads of the household or the registered voters in the household are asked several questions that the campaign pollsters establish. The questions need to be developed properly so an accurate answer may be found. Many questions are multiple choice with answers ranging from “completely agree” to “completely disagree”. Allowing for a wide-spectrum of answers allows respondents to be more truthful. For example, if a candidate’s likeability is an issue, answers like “very likeable/unlikeable”, “likeable/unlikeable”, or “somewhat likeable/unlikeable” helps respondents choose a better answer than “likeable” versus “unlikeable”. In that example, when all the data is collected, all the different answers with “likeable” may be added together, as well as all the different answers with “unlikeable”. This, then, provides the pollster a very accurate sample of likeable versus unlikeable traits compared to if respondents were just given two choices, likeable or unlikeable, in the question.
The results of those phone calls are then pooled and a statistical result is established. The statistics come from the results of the polling weighed against the census figures of the population. Results may be that more than 65% of respondents support the candidate on issue X while 25% of respondents over the age of 65 disagree on policy Y, which the candidate supports.
Many pollsters may want random data from a particular population. For example, if a campaign pollster wants to know how a candidate is doing with African Americans, women, or adults over the age of 65, the pollster will use a sample from the random calling that are self-identified as African American, women, or adults over the age of 65 households. This gives the pollster reputable polling sampling for how the candidate is performing with those set of voters.
Campaign polling results may be used in a number of ways. If the candidate’s policy recommendations or profile is asked to respondents, the result helps the campaign determine if clarification is needed or if the candidate is or is not winning over voters from their policy stances. On the other hand, many campaign pollsters test the waters of public discourse. Campaign pollsters may ask respondents how they feel about a particular issue. Depending on how the majority of the random sample answered, the candidate may decided to take on that issue on the campaign trail. This helps the candidate capitalize on what the public wants to talk about or what they find as favorable policy.
The best campaign polling is polling that is random with clear questions. The best questions with the most random pool can help a candidate’s election strategy and possibly win an election.