Polls Are For Dancing: The Problems of Network Polls

In recent years, political outlets have found it increasingly easy to reinforce what their follower’s value by using their own polling centers and controlled polls for their programs. I believe that because of this “quick polling” via outlet websites, that truthful and accurate polling have fallen by the wayside. As an outcome of this, Americans are more misinformed about the political numbers than ever before and it’s causing skewed and biased opinion to dominate popular opinion, in turn never fully achieving their goals of voter understanding. According to Thomas Hollihan’s book Uncivil Wars: Political Campaigns in a Media Age “Members of the news media seem to believe that their audiences are more interested in stories reporting who is ahead in the race… As a result, in recent years there has been a dramatic proliferation both in the number of public opinion polls and in the degree to which they dominate electoral news coverage.” (Page 168)

Increasingly, more and more polls are becoming popular and with up to “273 national polls” (Page 169) being conducted before the 2004 election, the numbers are becoming more and more contradictory. This trend in most prevalent on networks that poll on opposite party candidates. Popular opinion would say that Conservatives/ Republicans tend to watch FOX Network. Whereas voters whom identify with the Liberals/ Democrats tend to gravitate towards MSNBC for their political news, because these networks are aware of this split, they tend to answer loaded questions that reinforce their demo-graph’s beliefs. A result of this is a lack of new information being found, and more times than not, completely ignoring the moderate opinion.

When networks like these conduct these polls in such a manner they are often inaccurate in the real numbers because of their use of stratification in polling. When a news outlet allows any user to take part in their online poll, they are essentially allowing their key viewers to participate, not considering people who do not usually watch the network for their news. For example, in recent poll conducted in September of 2009, FOX News asked 900 voters via phone “Do you think the government's recent takeovers of sections of private industries such as in the banking industry, the insurance industry and the auto industry were a good idea or a bad idea?” The results were as fallows;

Party affiliation	Good idea	Bad Idea	Mixed	Don’t Know
Democrats	55%	31	8	6
Republicans	12%	80	7	1
Independents	23%	58	15	4

As Hollihan states in his book, the independent numbers are often skewed especially when “Push Polling” takes place. A push poll is considered to be a strategy in which a poll attempts to reinforce numbers people want to hear or see. The wording of “government take-over” versus what another outlet may word as “Increased regulation of” is a loaded questioning strategy. By shaping the opinion of party being polled by giving a question a connotation, the poll can receive the answer it prefers to receive.

Questions seven and eight of that same survey conducted are even more intriguing to analyze. Question seven asks; “Since January of this year, do you think the country has become more united or more divided?”

Party Affiliation	More United	More divided	No Change	Don’t Know
Democrats	38%	48	11	3
Republicans	13%	76	8	3
Independents	23%	60	14	3

Looking at this question at first glance, it seems harmless and straight forward, but if we examine it closer we see that the question is indeed a bit loaded and bias. It appears that popular opinion would have the majority of the country in agreement that the country is less divided since January. In-fact, the question is never followed up with the more important question, “Why do you think the country is more divided?” Without asking a question like this, it would be impossible to truly say the country is in agreement based on all the possible answers. Perhaps the reason a Democrat believes the country is more divided differs greatly from the opinion of a Republican. By asking a loaded question like this, Fox News is able to legitimize their stance, as well as reinforce their viewer’s beliefs. The majority in every demographic thinks the country is more divided, an opinion that carries the connotation that our Democratic president is not doing his job as most republicans would believe.

Question eight is especially interesting, the question asks “In the next congressional election in 2010, are you more likely to: SCALE: 1. Vote for the Democrat to help Barack Obama pass his policies and programs 2. Vote for the Republican to provide a check on Obama's power 3. (Too soon to say) 4. (Don't know/Refused).”

Party Affiliation	Democratic	Republican	Too soon to say	Don’t know/ Refused
Democrats	77%	7	10	6
Republicans	5%	79	12	5
Independents	29%	28	36	7

Again, by fallowing up with a question like this that again is able to shape the opinion of the party being questioned before they have time to develop their own opinion is a way FOX is able to reinforce Conservative opinion. It completely disregards the idea of opinion on a candidate, whether a voter may see a worthy candidate in their district that stems from the other party, although it’s naïve to think that would be the case.

Perhaps the greatest issue with this poll is that it was conducted to using 900 people via phone. The first problem with this method is that a poll of 900 people, 738 of which were white and only 103 of which were black, in a country with over 260,000,000 people can be troublesome. It’s substantially less than one percent of the population, so in order to stratify the poll properly would mean the call list would have had to select every 288,888 person within the United States. Even then, it is impossible to guarantee that you are receiving a fair piece of the population with every demographic being properly represented. The second major issue with the manner in which this poll was conducted stems from the process of phone polling. This assumes that everyone has a landline. Most people these days are beginning to drop their landline service, and the numbers of college age students and young voters that own landlines are minimal. The makes the numbers for the cohort almost impossible to examine.

Now if we turn our attention to MSNBC and their polling techniques, we find similar strategies that FOX uses, although there are some differences as well. The first major difference between the two cable networks use of polls is the use of NBC polls in the MSNBC polling results. FOX news conducts its own polls through its private company, whereas MSNBC’s numbers come from NBC “the network.” Another major difference in the polls comes from the cohorts involved in the polls. The NBC poll is much more thorough in sampling the demographic of the poll. Another major difference within how the polls are conducted is how the NBC poll frames responses. The poll does not group the responses of the group as either Republican or Democrat.

One question that can be seen from this poll that is especially interesting is number twelve; “Now I am going to tell you more about the health care plan that President Obama supports and please tell me whether you would favor or oppose it.

The plan requires that health insurance companies cover people with pre-existing medical conditions. It also requires all but the smallest employers to provide health coverage for their employees, or pay a percentage of their payroll to help fund coverage for the uninsured. Families and individuals with lower- and middle-incomes would receive tax credits to help them afford insurance coverage. Some of the funding for this plan would come from raising taxes on wealthier Americans. Do you favor or oppose this plan? (IF "FAVOR/OPPOSE," ASK:) And do you strongly or only somewhat (favor/oppose) this plan?” The answers ranged from 53 percent of Americans supporting this and 43 opposing this.

Overall, much of the NBC poll is conducted in this manner, which can be seen as problematic in a couple ways. The answers offer more variation within each question, allowing for a broader range of feelings. Another issue some might have with this polling method is the lack of political affiliation within each question, by saving that towards the end; it’s harder to identify what could be considered a certain affiliation’s opinion. The questions, much like twelve, could be seen as helpful by some because it attempts to explain the question at hand, instead of assuming those being polled are informed. Perhaps the most interesting part of the demographic explanation was the numbers of minorities used in this poll as opposed to those used in the FOX poll. Again, a bit broader range is used, as well as despite having a smaller cohort, the NBC poll was able to select a fair range of voters, although dominantly having white cohorts like FOX. The “cell-phone” question is interesting has of 138 polled, 100 had only cell phones, showing an increasing number of cell only use as stated early.

With all of these factors taken into account, it is easy to see how networks are able to shape their polling, sometimes to shape their viewers opinions on a hot button subject, or sometimes to reinforce the beliefs of those viewers. Either way you look at it, it is evident the use of so many polls makes it extremely difficult to have a completely accurate poll.