Philosophical vs Substantive Polling Questions

There’s often a disconnect between what public opinion survey results suggest and the public’s true position on an issue. To understand why survey results sometime imply one outcome and reality reveals another, the consumers of public opinion data need to learn the difference and identify philosophical position and substantive survey questions.

Philosophical survey questions, as philosophical questions in general, are not designed to arrive at a definitive answer about a specific issue. While most people could identify fundamental philosophical questions (What is the meaning of life? What is good/evil? What is beauty?), survey questions designed to evaluate philosophical positions are less easily distinguished.

Philosophical survey questions often ask respondents to evaluate the priority they place on various concepts or ideas, or to indicate whether they support or oppose a plan or idea in general terms. An example of philosophical position survey question may include a priority question about whether respondents feel adding air conditioning to existing schools without air condition is a high, medium, low, or not a priority. Another example could be when respondents are asked if they agree or disagree with the statement that the government should guarantee health insurance for all Americans. These questions do not discuss a specific plan, but rather probe how respondents generally feel about these topics.

Understanding the philosophical position of the voters is critical for developing a campaign’s strategy. If it is known, following our example, that voters are generally or highly supportive of adding air conditioning to schools where it currently does not exist, then the campaign for adding air conditioning has a starting point from where to understand their position. Supporters of adding air conditioning know that they are operating in an ‘environment’ that is ‘friendly’ to their cause. Supporters can focus their campaign on issues that will retain support, instead of trying to shift opinions in favor of adding air conditioning.

Philosophical opinions are often the poll results to which campaigns try to direct the public’s attention. They are the results about which campaigns generally want their issue debated, they focus on the big idea. Substantive questions, however, often reveal a schism in respondent opinion, and are less likely to be reported or the focus of a media release. Substantive questions, following our example, would ask survey responds if they supported or opposed adding air conditioning to schools in their district where it currently does not exist if they knew the proposal would require a bond costing thirteen million dollars. A substantive question injects the facts missing from a philosophical question and allows voters to make a more informed decision about their position on the issue, concept, or idea.

There are varying degrees of substantive questions. Respondents could be asking their opinions about an issue with only being presented the basic facts (Adding X number of air conditioners at Y cost). This type of question is often called an ‘initial ask’ or an ‘uninformed ask.’ It is generally asked in neutral language and, if the issue is about something likely to appear on a ballot, worded as closely as to the questions language on the ballot as possible. This sort of question gives a clear indication if the philosophical opinions of the respondents will differ with the reality.

Substantive questions may also be included to try to simulate the dynamics of an engaged debate about an issue (along with additional message testing techniques). Questions like these are typically called ‘informed asks.’ In an informed ask, balanced messages in support and opposition of an issue are cited after a statement of facts about the issue (Adding X number of air conditioners at Y cost, supporters say… opponents say…). Respondents are then asked, based on the information presented in the question, if they support or oppose the issue.

Through examining the differences between philosophical opinions and what respondents indicate through substantive questions is where strategy is developed. Campaigns need to know where the voters stand philosophically on their issue, idea, or concept and where support (or opposition) is lost (or gained) once additional information and messages are presented about the topic. Polling offers campaigns the ability to target demographic and attitudinal groups, discover where and among which respondents changes in opinions occur, and what needs to be done to keep or convert supporters to their cause.

To genuinely understand public opinion on an issue, more than the philosophical data needs to be presented. We have observed the philosophical fallacy in the current deluge of polling results presented in the media about healthcare reform. While the polls show wide appeal for change and reforming the system, specific plans, substantive questions, show a less consensus about how to fix the system. Consumers of healthcare survey data should be careful to seek out the results for substantive questions and not be clouded by reports based on the philosophical support of a plan.

Congratulations Mark Lesko

Fako & Associates, Inc. would like to congratulate Supervisor-elect Mark Lesko in his overwhelming victory in the election for Brookhaven, New York Town Supervisor on March 31, 2009.

F&A, Inc. worked with an incredibly talented group of consultants, campaign staff, and volunteers to develop the winning strategy that guided Lesko in securing 55% percent of the vote over Republican Timothy Mazzei. Lesko put together an outstanding field campaign team and an unrivaled team of consultants whose hard work helped to elect a strong fiscal reformer who will fight corrupt practices and save taxpayers' hard earned money.

Our strategy in Brookhaven was a continuation of our work in New York State over the past several cycles, helping Democrats regain office in tough, suburban and rural parts of the State. While we have experienced several victories in urban areas, it is our work in Upstate , the North County, and the Adirondacks that helped the New York Senate to regain majority control in Albany for the first time in forty years.

Informed Polling and Getting it Right

Every pollster will tell you that political polls are a snapshot in time and at best consecutive polls can elicit trends, but should not be used to predict turnout or the outcome of an election.

While the national attention is on how the pollsters got it wrong in several contests with surveys taken only days before an election, few are talking about how early surveying often-times gets things right. We've occasionally written about the importance of informed trial heats in some of our past posts.

To refresh, an informed trial heat is designed to simulate the effects of an engaged campaign, presenting balanced positive and negative messages about each candidate (or multiple candidates). The end result of the election scenario tells a campaign what is possible with their messages and themes they plan to implement. Most importantly, this section will determine if the core message works in direct contrast to the opponent(s)' message and can identify movement among the various demographic and attitudinal groups -- helping refine strategy.

F&A recently conducted a benchmark survey for a client running in a Democratic primary for open seat in a multi-candidate open-seat race. The benchmark poll was designed to evaluate the political environment, determine voter’s top issue concerns, examine opinions of the candidates and other significant figures, and test messages in support and in opposition of various candidates. As we always do in comprehensive benchmark polls, we included an informed trial heat question in the survey. The client was on a tight budget and didn't want to include "minor" candidates in the informed trial heat. After some debate, we were able to convince the client to include the "minor" candidates in the question.

We accounted for the ability of the "minor" candidates to get their message out given their budget constraints. Their messages were curtailed in the informed scenario to a simple bio-statement, while "major" candidates received bio, supporting, and opposing information, simulating an engaged campaign.

In our poll, the results of the informed trial heat were unexpected; a "minor" candidate took a 22 percentage point lead above the assumed frontrunner in the informed scenario, an increase of 25% above the candidate's level of support in the initial trial heat (the uninformed horse race question). A "major" candidate jumped up 7% and the assumed frontrunner stalled with a gain within the margin of error. Undecided voters in the initial trial heat heavily sided with one of the "minor" candidates. The percentage of undecided voters was reduced by over 40% in the informed trial heat. At this point we recognized the minor candidate's growth potential and advised the client to pay close attention to this so-called "minor" candidate. We noted that this individual clearly had the basic background and simple message that would break through the clutter of a highly engaged multi-candidate race, despite an initial perception of not being viable.

As the campaign progressed, the so called minor candidate ended up raising some serious money and gained significant earned media attention in addition to their own paid activities. It became apparent that this minor candidate was not minor, something our polling has observed only three months before the election

The "minor" candidate ended up winning this election, slightly ahead of the "major" candidate that we also observed gaining traction through the informed trial heat. This highlighted the usefulness and importance of utilizing informed trial heat questions in polls and why clients should never ignore perceived "minor" opponents. Polls that include an informed trial heat are one of the most useful strategic planning tools available to a campaign. It gives campaigns the information needed to determine if their message works (in the above example, out client’s message was not working); provides detailed strategic planning information, particularly at the demographic sub-group level, and gives campaigns information to prepare and adjust strategy for unanticipated situation (such as an unexpectedly strong opponent).

Two common mistakes that cost local campaigns their elections.

Campaigns outside of major population centers, lower budget and local campaigns often rule out conducting strategic opinion surveys for two reasons: they think they can't afford a strategic survey and the candidate and/or staff believes they know the city or district where they are asking for votes. These mistakes set up local and lower budget campaigns for failure.

While having a comprehensive survey that allows for an in-depth evaluation of the election dynamics and gauging voters' philosophical outlook would benefit any campaign, that level of information is not necessary to develop a winning strategy for a campaign that will likely be driven by direct voter contact (field campaigning) and direct mail.

A poll, regardless of size and length, should be designed to develop a strategy. As we've discussed previously on this blog, the cost of polling is arrived at primarily though the number of interviews conducted and the length of the survey (in minutes, not necessarily the number of questions). Even a shorter poll can highlight the persuasive winning arguments and expose the strengths and weaknesses of the candidates. A modest poll doesn't mean you are only going to find out who is winning, losing and better known… All polls are about developing and refining strategies that maximize the use of a campaign's finite resources.

Localized races like District Attorney, County Board, and Alderman and Mayor (in smaller cities) can make use of smaller sample sizes (depending on the size of the voting population) and shorter surveys that last 5 to 10 minutes. Surveys of this scope can accurately evaluate the opinions of the candidates and other public figures, test the strength of support for the candidates, identify key targets, and test a limited number of messages for the campaign while not breaking the campaign's budget.

One of the main objectives of a lower budget campaign in seeking polling should be to find the message(s) that have the widest appeal to the largest amount of voters and niche messages that are targetable to subgroups of appropriate scope.

When determining which messages tend to work better with certain subgroups, it may not be possible to determine with high statistical accuracy that a particular message works better with a very specific, micro-targeted subgroup in poll with a small sample size. Larger and more practically targeted subgroups will allow for a statistically reliable reading of which message is most persuasive among broader subgroups and the sub-set will be useful and practical for your direct mail and field campaign budget.

To give an example on targeting for a lower budget campaign, let's assume the campaign's survey reveals that your candidate has weaker support among younger women. Generally speaking, it is more effective for a lower budget campaign to use a more practical, broader direct mail targeting criteria (such as "women under 50" instead of "Women 18-25 in households with children with net income over 100K") and deliver a message that tests well among many voters within a broader subgroup. Distributing your best messages to the largest amount of voters as possible is vital because campaigns at the local level already are dealing with a small numbers of voters, are lower profile, gather less earned media, and the candidates are typically less known and get less attention than other higher profile campaigns.

An experienced polling company should be able to evaluate the messages your campaign wants to test and help you refine them in a way that will maximize the quality of the data your survey will provide. Subgroup analysis to determine message targeting is still possible with smaller sample sizes, even with the inherently higher margin of error that a smaller sample size brings. The bottom line is that a campaign shouldn't spend money for a survey that provides data that the campaign’s budget won't allow it to use.

The other common mistake of lower budget and local campaigns is assuming a level of knowledge about the voters, their concerns, and what messages will work for their campaign. While it is possible to knock on every door in a smaller city or county, it is impossible for a candidate or campaign to gather the type of information that is revealed by a random sample survey of likely voters. Even the briefest of political polls will screen for likely Election Day voters and sample the electorate in proportion to the predicted Election Day turnout. The screening process helps to ensure that the opinions reflected in the survey are those of voters who will likely participate in and ultimately decide the outcome of the election. The response campaigns receive from activists, supporters, and voters at the door do not generally represent or generalize to the opinion of all those who are likely to cast their ballot on Election Day. Polling provides vital strategic data to compliment the anecdotal information that your campaign will receive.

Getting wrapped up in the cost of political polling and the need for full-blown, multiple micro-targeted direct mail campaign often reinforces the belief among candidates and their staff that their campaign doesn't need strategic research to understand their electorate. The reality is that lower budget and local campaigns can rarely afford to go into an election without the research that a basic survey can provide. Good information is the foundation of a successful campaign strategy. Thinking that political polling is too expensive and that they know their electorate well enough cost campaigns their elections.

Cross Tabulation Tables

Cross Tabulation tables (often referred to as "cross tabs" or "tabs") numerically display the results of a survey, showing how the answers to one poll question break down according to the answers to another poll question or other data derived from the sample.

Generally speaking, cross tabs are detailed statistical / numeric analysis of a survey, but in a raw, unfinished form. Your pollster will use cross tabs to generate a written analysis that will serve as your campaign's strategic guide and tactical play-book.

It is important to understand what basic components are necessary to generate a good set of cross tabs that will in turn give you the data needed to make strategic and tactical decisions. Although many campaigns have staff that understand how to read tabs, a campaign should not have to probe through cross tabs to get answers from their poll -- that's the job of the pollster.

Unlike a "topline report," which reports the breakdown of answers to a specific poll question in terms of the overall percentage of all respondents, a cross tab shows how the results of questions (such as who is winning in a trial heat) relates to a variety of demographic and attitudinal classifications.

(Figure 1, Click Image to Enlarge)

Technically speaking, a cross tabulation displays the joint distribution of at least two or more variables (e.g.: "Male" & "Candidate 1 Supporters" or "Male less than 50 years old" & "Candidate 1 Supporters") (See Figure 1).

The topline report might show Candidate 1 capturing 24% of the vote and Candidate 2 capturing 50%, with 26% undecided. A cross tab (See Figure 2), however, will show candidate 1 capturing 24% of the vote overall, but also shows candidate 1 capturing 31% among voters in the South Region and performing better among voters over 65 with 32% of their vote.

(Figure 2, Click Image to Enlarge)

Now that you have a basic understanding of how cross tabulations work and are read, it's important to know which demographics and attitudinal groups should be included in your cross tabs.

Every pollster has their preferred way of structuring questions and response items. We've created a very basic set of sample cross tabs with easy to understand response items to help show you some possible combinations withing the tables.

As you saw in Figure 2, demographic questions are among the most essential components of your cross tabs. You need to know how regional differences affect the opinion of voters; as we discussed the importance of a regional break-down in the previous section of the components of a survey. Beyond regions, your cross tabs may or should include, depending on the scope and purpose of your survey, demographic data like Gender, Age, Race, Political Party ID, Political Ideology, length of residence, education levels, income levels, marital status, if the household has school aged children and so on.

Cross tabs allow for more intricate demographic breakdown. As shown in Figure 2 and Figure 3 below, it's possible to combine variables to arrive at combinations of demographics, such as Men under Age 50 and Female African American voters. The combinations are nearly endless, but finite groups tend to statistically break down in reliability as the sample size decreases among the groupings and the margin of error increases. Your survey sample size will largely determine the level of micro-analysis capable within your cross tabs.

Beyond clearly defined demographic groups, your crosstabs should include various special attitudinal groups. The special groups may be defined through various variables based on the respondent's opinions and answers to both demographic and substantive questions in the survey. These groups can represent key targets for your campaign and a demographic picture of these voters can be developed by utilizing these groups in your cross tab analysis, allowing the pollster profile key target groups. Attitudinal groups (See Figure 3 & 4) can help you determine the demographic breakdown of various significant groups such as weak/strong supporters/opponents and undecided voters and variants of them such as undecided non-Democrats and uncommitted women, etc. They can include variables such as voters who support your party's top of the ticket, but do not support your candidate and swing / persuadable voters who in the course of the survey showed movement from undecided or supporting your opponent to supporting your candidate, etc. and numerous other possible attitudinal options. Depending on the construction of your survey, you can also determine who are single issue voters on hot button issues. There is a wide array of attitudinal possibilities, only limited by the scope of your survey.

(Figure 3, Click Image to Enlarge)

(Figure 4, Click Image to Enlarge)

In the end, a well constructed cross tab report will provide the numeric foundation and data for any campaign's strategic plans and are the root "source" for a pollster's written analysis and strategic advice. Pollsters who produce intricately detailed, useful tabs that go way beyond the standard factors will provide campaigns with higher quality strategic advice. Pollsters using this level of detail provide more precise direction on message, targets and strategy that increases the probability of developing a successful plan.

Components of Crafting a Survey

Most successful endeavors require prior planning; a successful political survey is no exception. The quality of your survey is dependent upon the value of the information gathered before it is crafted. There are generally seven research components needed to develop a strategic survey. We will briefly outline these criteria, explain their significance in surveying, and simultaneously review the various components that should be present in a good benchmark level survey.

A benchmark survey is a comprehensive evaluation of the election dynamics, candidates and issues. They are used to develop your strategy and message. They serve as your campaign's tactical guide and strategic "play book." Benchmark surveys usually are 15 - 20 minutes in length, or longer. It should be as comprehensive as your budget allows.

Unlike like Stephen Colbert’s interviews in his Better Know a District series, you have to have a basic understanding about the region where you will be polling.

What’s the population? How would you describe the various areas within where you want to poll? What are the characteristic’s of the area’s economy? Does the area have a unique history?

Population and voter registration data help determine the polling sample size needed to efficiently and conduct a statistically accurate survey. All political surveys should have multiple regions that represent something usable to your campaign. The regions that get established for your survey have to be significant for your campaign.

Is the West end of town mostly town homes? Is there a significant minority concentration on one side of town? Are there a handful of distinct townships within the region to be polled? Are these groups of precincts representative of a higher-income area? Are these groups of precincts virtually not walkable by the candidate or volunteers?

Questions like these should be asked when developing regions and the list of demographic classifications that you will find useful to have in your survey (such as any combination of gender, race, income, education, union status, political ideology, party ID, etc). Members of your campaign need to be able to look at your survey analysis and quickly determine what’s happening, among whom, and where.

Knowing population data (and vote share information) will help you determine what percentage of the vote each region represents. The campaign should work with the pollster (if needed) to find the projected turnout for the district and your winning 50% plus 1 number. Knowing where the vote is coming from and the opinion of voters in the various regions can help not only focus your campaign’s message, but also help in the allocation of the campaign’s budget.

Answering preliminary questions about the area you plan on surveying establishes the background to keep in mind throughout the survey process. Knowing a little about the district’s personality serves as a guide through the survey data collection process and makes it easier to spot any potential abnormalities early on.

When crafting a campaign, you also need the key players within the district.

Are there key officials, public figures or private individuals who would have an interest or impact on the campaign? Would an endorsement from a certain official be a help or hindrance? Is it worth spending campaign time and money to promote a particular endorsement? Will campaigning or sharing resources with another candidate hurt or help my campaign?

A proficient benchmark survey will test the personal favorability ratings of people who are of an interest to the campaign. Knowing if someone holds significant substantive recognition ratings and if they are viewed positively or negatively will help determine if valuable campaign resources should be expended in seeking or promoting endorsements and/or affiliations.

If public officials are being tested in the survey, their job performance ratings are usually also tested. When in the position of a challenger, the job performance rating and personal favorability of the incumbent is always tested. Incumbents should test their own favorability and performance ratings to help reveal any underlying potential weaknesses.

Job performance and personal favorability weaknesses oftentimes will reveal weaknesses that aren’t apparent in the benchmark survey's trial heat of candidates or issues. Sometimes voters will lean towards a known and disliked incumbent over an unknown challenger. For example, voters who responded that they are “weak supporters” in a trial heat question filtered into a group that show weak intensity toward a candidate personally or in their job performance rating through cross tabulation reports. It’s possible to narrow down the demographic list of voters that should be targeted by the campaign in this way through a variety of response position combinations.

Classifying groups of voters based on their support and feelings toward candidates or issues is not so useful in itself, but knowing what strategic message to communicate to target groups to achieve a desired effect completes the purpose of the survey.

Detailed background information on the candidates or issues must thoroughly be reviewed, prioritized and incorporated into the survey. This information could contain both helpful and damaging information. Depending on the budget of a campaign, this type of information will contain variations of the following: summaries of public records (tax liens, lawsuits, etc.), news clipping, issue positions, campaign finances, endorsements, other court/judicial records, interviews, tax records, voting records and other vital documents.

This information gathering process is called Opposition Research (aka OR, Oppo) and Vulnerability Research (research about yourself or the campaign issue). They are critical components that must be completed before a survey is crafted. There are several reputable information gathering firms who specialize in generating detailed strategic OR reports. Lower budget campaigns can conduct their own baseline research, but they need to be careful about allowing volunteers gather OR information, especially in candidate campaigns where zealous volunteers may not be diligent on fact-checking.

Opposition research has received a bad name in recent years (Negative attacks, "Swift Boating,"etc.), but OR itself is not inherently harmful to the democratic process. The ethics of OR is a discussion for a different posting, but remembered that the information tested in a survey is always left to the discretion of the pollster and the campaign. A good pollster will lend you his or her experience and guidance on whether to test controversial information and advise your campaign on utility of the information and the repercussions that could occur.

Along with knowing information about opponents, you or your issue, and, details about the area to be polled, it’s also important to understand all major or pertinent issues that may affect the campaign, including local issues. Polling allows you to test voter's opinions on important known issues.

Is a big-box retailer petitioning to move into town? Will reproductive freedom be a wedge issue brought into the campaign by your opponent? Is there a statewide debate brewing about same-sex marriage, or funding stem cell research?

Conducting a political survey about important issues will reveal not only the opinion of voters, but also the intensity they feel towards issues. You’ve probably heard the expression of a "poll driven politician," your campaign doesn’t want that association. Polling should not be used to determine your (or your campaign’s) issue positions, but rather it should help you develop an approach to talking about an issue (or whether to talk about an issue), discover natural allies (demographically, regionally), and, refine your approach in how your campaign communicates its messages about the issues.

Not every issue will be known by your campaign. A good benchmark survey will incorporate an open ended question early on to directly ask the voters what issue or problem they are most concerned about. Often times this question will reveal issues which your campaign is already aware, but open ended verbatim responses also reveal how issues are being discussed and lend insight to why voters feel the way they do. This form of open ended questioning sometimes reveals upcoming issues that haven’t boiled to the surface.

Knowing where voters stand on the issues will help refine your message. A benchmark survey can test the persuasive value of the potential arguments in support of your campaign and the arguments against your opponent(s). Testing messages isn’t so much about finding out what attacks stick against your opponent, but rather discovering which of messages hold the most persuasive value and most correlate to the concerns of the voters. Your campaign needs to speak to the concerns of the voters, testing messages through a benchmark poll keeps a campaign on target by speaking to the issues with messages that resonate with the voters.

This primer on the components of a benchmark survey is meant to familiarize information seekers on the formation process of a political benchmark survey. While in no way would we consider this discussion comprehensive, it has hopefully answered some preliminary questions about planning for a political poll. Benchmark polls should be conducted as early as possible in a campaign. To summarize, a benchmark poll’s purpose is to get the strategy and message correct early in the campaign, rather than spending the rest of the campaign fixing it.

Smoke Free Illinois

With the passage of Senate Bill 500, Fako & Associates, Inc. would like to congratulate all the organizations that have worked tirelessly to protect public health in Illinois. We especially thank the American Lung Association, who lead the Smoke Free effort around the state. We also would like to thank the American Cancer Society and the American Heart Association for letting us be a part of making Illinois a Smoke Free State.

The passage of SB500 is a strong commitment by the State of Illinois to save hundreds of thousands of lives each year from the deadly affects of Second Hand Smoke. We are proud to have been a part of this effort with our statewide work and our city-specific initiatives in the city of Chicago, Oak Park and in our state’s capital, Springfield, Illinois.