Received date: December 17, 2015; Accepted date: January 11, 2016; Published date: January 20, 2016
Citation: Anderson JO, Crooku ST. Does the Media Have Some Explaining to Do? Distinguishing Between the Effects of Acquired and Applied Knowledget on Public Opinion of the National Debt. Global Media Journal. 2016, S1:3.
Visit for more related articles at Global Media Journal
Does acquiring facts about political issues necessarily give individuals the information they need to create an opinion about support for public policy? This study seeks to refine the effects of political knowledge on public opinion, differentiating between types of information that do and do not elicit opinion change. We distinguish two categories of political knowledge: acquired knowledge, a surface awareness of factual information, and applied knowledge, knowledge that can be transferred to new concepts and beliefs. Using a survey experiment and information about the amount of the national debt as a test issue, we find modest support for the hypothesis that political knowledge is more likely to affect the opinions of those individuals who can apply the information to a political context than the opinions of those individuals unable to apply information to political realities. The implication of this study suggests that media explanations of some complex political issues are necessary in order to move public opinion.
Political knowledge; National debt; National deficit; Public opinion, News
In today’s multi-faceted information environment, citizens have omnipresent opportunities to gain political knowledge. But do individuals use the information they accumulate in a manner that will affect their political behaviors? Most models describing the relationship between political knowledge and political behavior assume so. Whether described as “knowledge” , “education” [2,3], “sophistication” [4,5], or “awareness” , studies tend to assume that information learned by a subject or political participant is inherently understood and utilized in the decision making process. Nevertheless, a smaller but important body of political knowledge research addresses the concern that opinion surveys can misrepresent collective opinion because respondents lack full comprehension of their acquired knowledge [7-9]. The contrast between the two sets of findings raises the question: when and how does political knowledge inform the behavior of individuals who possess it? This study seeks to refine the effects of political knowledge on public opinion, differentiating between types of information that do and do not elicit opinion change. We suggest that individuals may harbor certain pieces of policy-specific political information that have no consequence to their political behaviors. In particularly complex or technical political issue areas, it is plausible that individuals retain facts but do not have the necessary contextual information to apply those facts. We use the topic of the national debt as a testing ground for our theory. With the size of America’s debt far above the conceptualization of the average citizen, policy-specific knowledge of the amount of the debt may be inadequate to move opinions. An individual that has contextual information to ground her policy-specific knowledge may exhibit different levels of policy support, in this case support for government spending, than a person without knowledge of context, even if both individuals are aware of the debt amount. In other words, in some cases political knowledge in and of itself may not be enough to change political behaviors and beliefs.
The distinction between general acquired knowledge and complex applicable knowledge is essential to the theoretical underpinnings of this study. Scholars of educational instruction techniques found a significant difference between knowledge that is limited to “superficial awareness of key concepts and facts” (what this study calls acquired knowledge) and knowledge that can be transferred to new situations (what this study calls applied knowledge) . Political science primarily studies political knowledge under the assumption that the accumulation of political knowledge is a linear progression that grows with each addition [6,11,12]. However, if individuals acquire facts but cannot apply them to political policy, the conceptualization of political knowledge and its effects should change.
By analyzing responses to different types of information about the national debt, this study gives support to the idea that only certain types of political knowledge are useful in predicting and shaping political behavior. We find that individuals with knowledge of the amount of the national debt behave no differently from individuals without knowledge of the debt amount. However, individuals with knowledge of the amount AND context to help them understand the amount demonstrate modest but statistically significant differences in their levels of support for government spending. The paper proceeds as follows. First, we offer a summary of the debate of the effects of political knowledge on political behavior, highlighting gaps in the research that offer the opportunity to test for the effects (and lack of effects) of varying types of political knowledge. Next, we present our survey experiment and discuss specifics of the treatment conditions, participants and results. Finally, we examine the implications of this research, its limitations and future research possibilities.
Political knowledge affects political behavior in a number of ways. Individuals with higher levels of political knowledge think, vote and participate in politics differently than those who are illinformed [13-15]. Research on the effects of political knowledge shows that the unknowledgeable citizenry uses heuristics to overcome much of their information deficit [6,11,16,17], but fully informed individuals are more likely to form stable and efficacious political opinions [1,18]. John Zaller’s  model of how people express political opinions indicates that if people are sophisticated enough to receive political information, then they must understand it. Otherwise, their lack of political sophistication blocks reception of political messages before the messages have a chance to resonate or affect opinions. Furthermore, through over three dozen political opinion surveys, Jerit and Barabas  found that respondents were dependent on recent exposure to information from the media in order to correctly answer questions designed to gauge specific knowledge of policy. If an individual has a limited understanding of the information they receive, then they will likely not retain the information for later use. Both of these findings suggest that individuals have very limited political knowledge that they fail to understanding. Though these findings are important (indeed, Zaller’s work in seminal to the field of public opinion research) and undoubtedly true for a large portion of issues, we question if the framework holds for all types of political information. In an age where media and technology widely disseminate many kinds of knowledge, it may be possible for individuals to acquire political facts, and even retain them, without fully understanding them. With the complexities of the modern political world, it is conceivable that the recipients of political knowledge are given facts with such frequency that they retain the facts while failing to fully comprehend them. In such an environment, individuals are often presented with bullet point versions of facts and tables that would take much longer to digest than the attention span of the individual will allow. Armed with pieces of political knowledge and perhaps a phrase of jargon of a technical nature (that is not likely understood), many individuals may retain some policy-specific knowledge without an ability to apply it to policy beliefs. The longer an issue cycles through the national news or social media sphere, the more likely is an individual to glean bits of political information. The media’s stratification of information opportunities further provides for the possibility that individuals will gain only bits of political knowledge that do not affect behaviors. Individuals have the opportunity to select the coverage they receive, and greater media choice makes it easier for people to find their preferred content, often absent of substantive political coverage . Searching for the most entertaining version of information, individuals are likely obtaining knowledge on a baseline level that does not truly represent the complexity of the subject being reported. Conversely, some knowledge may be achieved while not searching for it at all, such as news article headlines in a Facebook or Twitter feed.
A prime example of such a public policy with facts and phrases that a voter would know without understanding might be sequestration. With so much politicization of the term, it is conceivable very few members of the public actually understand the mechanism of budget reduction, but rather repeat the political talking points they hear through their chosen news media outlet. The individuals may have seen their favorite talk show host use the term, or know that it was designed to account for $4 trillion dollars in debt reduction, but without information on how the tool is supposed to be implemented, they are unable to form a complete opinion on the mechanism .
Upon implementation, the general public had very little awareness of sequestration. This notion is supported by the findings of a poll conducted by Gallup on May 2-3, 2013 . In this poll, the Roper Center for Public Opinion research sought to evaluate the public’s support for sequestration. In spite of giving respondents a definition of sequestration within the question, investigators still received a "do not know enough to say" response at a rate of 51%. Furthermore, a poll conducted by the Washington Post on February 26th, 2013, found that 46% of respondents did not understand what would happen if sequestration were implemented. Surprisingly, however, was the fact that 51% of respondents described themselves as following the news about sequestration . This disconnect between the amount of time individuals spend following the news and their confidence in their understanding of specific information supports the theory that individuals in the current political climate accept information but do not understand it.
As discussed above, conventional wisdom about political knowledge is that individuals with knowledge of political affairs behave differently than those without such knowledge [1,2,4,6]. However, in today’s increasingly complex political climate, it is conceivable that individuals may retain facts about political issues and yet not truly understanding their meaning. Some pieces of political knowledge, especially those technical or unfamiliar in nature, may function much like an unrecognized word in a foreign language. If the word is emphasized strongly enough, or often enough, we are likely to remember it. But without some contextual clues or a specific definition, the word is essentially meaningless. Given the public’s very low levels of overall political knowledge [24,25], it is plausible that individuals have difficulty interpreting certain types of newly acquired information, and consequently, this type of new knowledge may not affect subsequent political behaviors. This hypothesis follows:
H1: Individuals who learn the informational context to apply policy-specific political knowledge are more likely to shift their policy views than those who acquire political knowledge without informational context.
According to this hypothesis, political knowledge is likely to affect public opinion for those individuals able to apply that knowledge to policy areas. Political knowledge is unlikely to make a difference in opinions or behavior if the knowledge cannot be applied to policy contexts. The debate surrounding the growing amount of the national debt is a prime example of such a complex issue, and it will also serve as the testing ground for our study distinguishing between the effects of acquired and applied knowledge. The national debt, at more than $18 trillion dollars (roughly $15 trillion at the time of this study), is an amount of money incomprehensible to the average American. The average college graduate earns $2.27 million dollars over his or her entire lifetime , yet the lifetime earnings of 100,000 college graduates together barely account for one per cent of the $18 trillion dollar debt. Thus, the issue of budget deficit is alien to most United States citizens. The accumulated debt is an amount of money so large that it does not fit into the average citizen’s view of the world. Nevertheless, many Americans are familiar with the national debt as a political issue because of its prevalence in the 2012 Presidential campaign. Mitt Romney repeatedly cited the amount of the debt as he attempted to persuade the public to oust President Obama for economic reasons. Arguments against the high yearly deficit arguably hallmarked the Romney campaign. A Google search for “Mitt Romney” + “Presidential campaign” + “national deficit” in April 2013 yielded over 91,700 results. Thus, the national deficit and total debt as a test issue for this study exhibits high external validity. Americans reasonably may know something of substance about the debt amount because of repeated exposure to Governor Romney’s appeals, and yet there is reason to believe that understanding of the information may be incomplete. In this study, we examine if policy-specific knowledge of a particular political issue- the amount of the national debtnecessarily affects public policy support. The results of the study will yield one of three possibilities in regard to the relationship between the independent variable, knowledge of the amount of the national debt, and the dependent variable, support for government spending. The first potential result of the study is that knowledge of the amount of the national debt does not matter. That is, individuals possessing knowledge of the amount of the national debt do not have preferences for government spending distinguishable from those without knowledge of the debt. This prospect goes against an underlying tenant of a large body of political knowledge research that political knowledge affects individuals’ behavior [1,2,4,6]. Nevertheless, given some findings that even large increases in political knowledge can have limited effects on behavior , we must allow for this contingency.
The second possible finding is that knowledge of the amount of the debt, in and of itself, affects individuals’ levels of support for government spending. This prospect fits with the existing literature on political knowledge that suggests that knowledge changes citizen behavior. If this is the case, we would expect citizens with knowledge of the amount of the debt to show different levels of support for government spending than those without knowledge of the debt amount. The third possible relationship between the variables is conditional: political knowledge of the amount of the national debt will affect individuals’ support for government spending if and only if they have additional contextual information that helps them apply the knowledge of the amount of the debt to their view of the world
To test for differences between the effects of acquired and applied knowledge on support for government spending, we conduct a survey experiment. Experiments offer an appealing method to test for the effects of political knowledge. Experiments eliminate the problem of self-reported overestimations of media exposure to relevant policy areas , and they allow for control of both content and levels of exposure. With existing knowledge and opinions randomly distributed and theoretically equalized across groups, differences in levels of policy support will be the result of the experimental manipulations. By randomly assigning participants to conditions and systematically manipulating the independent variable, we maintain control of external sources of variance . The study utilizes a convenience sample of undergraduates and a smaller nationally-representative sample to test for the effects of political knowledge. We will discuss the specifics of the research and then report results.
In the experimental study, conducted in April 2012 (undergraduate sample) and October 2012 (national sample), participants were 264 undergraduate students at a Southern state university and 200 adults from across the United States. The undergraduate students were recruited through lower-level general education core courses and received extra course credit in exchange for strictly voluntary participation. The national study was conducted through Survey Sampling International’s Quick Take program. SSI Quick Take is an opt-in survey panel comprised of a representative sample of U.S. residents. It uses the Internet as a platform for polling rather than traditional polling methods. Participants are recruited over the Internet as well, primarily through web advertising campaigns that appear based on keyword searches. In exchange for their participation, participants earn incentives. By maintaining a database of each participant’s demographic information, SSI employs block randomization to approximate a random sample of all U.S. residents. Subjects in the undergraduate study were randomly assigned to one of three conditions: no debt amount, debt amount, and debt amount + context. We describe the treatments in detail below. Subjects took the study using pen and paper in their undergraduate classrooms. Subjects in the national study were randomly assigned to one of two conditions, deficit amount and deficit amount + context. They took the study on their own computers at a location of their choosing, linking to the online study through an email invitation. For undergraduate and national subjects, as they began the study, they answered a series of questions on demographic information (gender, age, race) and political predispositions (gauging partisanship, ideology and knowledge). Subjects were then asked to read a news article that contained the experimental treatments. Following reading the article, subjects evaluated a series of questions about support for government spending that will make up the key dependent variables for the data analysis, as well as a question about their Presidential vote choice. To conclude, subjects were debriefed and thanked.
The experimental manipulations occurred within an edited news article entitled, “Obama and Romney Talk Deficits on the Campaign Trail.” The article was a compilation of three different actual news articles published between November 2011 and April 2012, with minor editing changes [29-31]. The content of the articles varied by specific numerical information of the debt amount and by contextual information that helped participants understand the amount of the debt. Below is an excerpt of the news article that illustrates the differences between the three treatments.
Excerpt from Treatment 1: No debt amount “You realize, of course, that number up there you are going to be paying that back,” Mr. Romney said to a group of students, while pointing toward a clock calculating the federal debt. “Our economy cannot sustain the current level of spending. Difficult cuts must be made.” “We must fight the debt by having those Americans who can afford it to pay their part of taxes,” said Alan Krueger, Obama’s chief economic adviser. “Millionaires like Romney need to pull their own weight.”
Excerpt from Treatment 2: debt amount “You realize, of course, that number up there that $15 trillion you are going to be paying that back,” Mr. Romney said to a group of students, while pointing toward a clock calculating the federal debt. “Our economy cannot sustain the current level of spending. Difficult cuts must be made.” “We must fight the $15 trillion debt by having those Americans who can afford it to pay their part of taxes,” said Alan Krueger, Obama’s chief economic adviser. “Millionaires like Romney need to pull their own weight.”
Excerpt from Treatment 3: debt amount + context “You realize, of course, that number up there that $15 trillion is difficult for the human mind to comprehend. If someone wanted to pay it off, he would have to pay $10 million a day, every day, for 4,100 years. Put another way, 15 trillion one-dollar bills would wrap around the earth’s equator 58,000 times. You are going to be paying that back,” Mr. Romney said to a group of students, while pointing toward a clock calculating the federal debt. “Our economy cannot sustain the current level of spending. Difficult cuts must be made.” “We must fight the $15 trillion debt by having those Americans who can afford it to pay their part of taxes. Some people describe the deficit as a mountain of debt. Let's start there. The tallest mountain in the world is Mount Everest, at just over 29,000 feet. To get $15 trillion, you'd need well over 185 thousand stacks of dollar bills as tall as Mt. Everest. It’s not a mountain, it’s a mountain chain,” said Alan Krueger, Obama’s chief economic adviser. “Millionaires like Romney need to pull their own weight.”
The full news articles can be found in the appendix. Table 1 shows treatment conditions and number of subjects within each, as well as the average demographics. The student sample was slightly female and Republican, and very white. The national sample was slightly Democratic and more racially diverse.
|Undergraduate sample||National Sample|
no debt amount
Mean partisanship: 5
debt amount only
Mean partisanship: 4.89
Mean partisanship: 3.74
Debt amount + context
Mean partisanship: 5.08
Mean partisanship: 3.39
Table 1: Treatments and subjects of experimental study.Dependent variables
Subjects were asked to answer several questions designed to gauge their level of support for government spending. Responses were coded to run from most supportive of spending increases to most supportive of spending cuts. Because preferences for government spending are linked to partisan preferences, the study utilized questions about both health care spending and military spending, issues more closely linked to the Democratic Party and the Republican Party, respectively. These questions were worded as follows: Should the winner of the 2012 Presidential election increase or decrease national government spending on health care? Should the winner of the 2012 Presidential election increase or decrease national government spending on the military? To indicate level of support for spending, respondents answer on a five-point scale anchored by “increase significantly” and “decrease significantly.” Subjects may wish to cut spending on the opposing party’s issue for strictly partisan reasons, so the dependent variable was measured by the subjects’ responses to spending questions about their own party’s issue area. Republicans were measured by their willingness to cut spending for the military, while Democrats by their willingness to defund health care programs. Republican and Democratic leaners, or those independents identifying more closely with one party than another (three’s and five’s on the standard seven-point partisan scale), were grouped with their closer partisans, as research suggests little difference in behavior between weak partisans and partisan leaners . For true independents, or those who do not see themselves as closer to one party than another, we employed the average of the spending preference for the health care spending and military spending questions. A secondary dependent variable measures the length of time subjects believe will transpire before the United States eliminates its national debt, a measure that should indicate the perceived seriousness of the national debt as problem. The question, “When will it be realistic for the United States to eliminate its national debt?” is coded on a 4-point scale, from shortest to longest length of time until elimination. The time frame is by generation, with choices of within my lifetime, within my children’s lifetime, within my grandchildren’s lifetime, and none of the above .
A comparison of means of the dependent variables across treatment conditions reveals a modest but statistically significant effect between acquired knowledge and applied knowledge on the level of support for spending cuts of study participants in the student sample. Contextual examples that help individuals apply knowledge increase levels of support for spending cuts. Those receiving information of the debt amount plus contextual information to help them apply knowledge of the national debt (treatment 3) were more likely to support spending cuts to their own issue areas than participants in the other groups. Participants receiving treatment 1, who gained no numerical information on the amount of the debt, were indistinguishable for participants in treatment 2, who gained numerical knowledge only of the amount of the debt, suggesting that in the case of the national debt, knowledge alone is not a sufficient condition to elicit opinion change. The mean level of support for participants in the applied knowledge treatment (treatment 3) was 3.03, while the average support for treatments 1 and 2 was 2.82 and 2.79, respectively. This difference of 0.24 represents 6% of the possible movement in the scale. In the national sample, expressed levels of support of participants in treatment groups 2 and 3 follow a similar trend, but the differences are quite small. Numerically, those participants who gained numerical information plus contextual information average higher support for spending cuts compared to those who gained only numerical knowledge of the debt, but the differences, 2.52 for participants in treatment group 3 and 2.37 for those in treatment group 2, do not achieve typically accepted levels of statistical significance.
For the second dependent variable, length of time required for the elimination of the national debt, a comparison of means reveals that treatment 3, the condition of numerical information and context, again affected opinions of the issue more strongly than did treatment 2 of numerical information only (for the student sample). Participants in treatment 3 answered the question with a mean score of 3.27, while those in treatment 2 gave a mean score of 3.05, a difference of 0.22, or 7.3% of the possible movement on the 4-point scale. The difference approaches statistical significance at t < 0.08. As above, the national sample follows a similar pattern compared to the student sample, with subjects in treatment 3 scoring 2.58 compared to 2.43 in treatment 2. Nevertheless, the national sample means are not statistically different from each other. Mean scores for each treatment group are found in Table 2. In addition to producing differences in levels of support for spending cuts, we expect measures of applied political knowledge to further our overall understanding of individual determinants of public support for government spending on one’s own party issue areas. If such is the case, reactions to exposure to policy-specific information and contextual examples should behave in an expected fashion with other known individual-level predictors of fiscal policy beliefs. To test this, we regress support for government spending on one’s own party’s issue on a set of predictor variables. The independent variable of most interest is the dummy variable representing the experimental treatment for numerical debt knowledge plus contextual information. As controls, the model also employs independent variables for gender (male), strong partisanship and ideology. It also includes a number of control variables regarding political knowledge to ensure that preexisting information about the national debt is not driving the model’s effects. These questions were asked prior to the treatment conditions and include general political knowledge, as measured by the number of correct answers on an 8-question political knowledge battery; preexisting debt knowledge, as measured as a correct response to a multiple-choice question about the amount of the national debt; self-assessed debt knowledge, as measured by an individual’s response to a multiple-choice question about how much she knows about the debt issue; and prioritization of the national debt, as measured by a multiple-choice question on the importance of the debt the new President should place on the issue. The results of these analyses for the student sample are displayed in Table 3. The results of analyses of the national sample were not statistically significant Table 3. Entries are OLS regression coefficients with standard errors in the following column. The dependent variable is coded so that a higher score indicates a greater level of support for reduced government spending. *p < 0.05; ** p < 0.01 key finding is that the experimental treatment providing participants with both numerical knowledge of the debt and contextual information had observable effects on participants’ levels of support for government spending cuts. Analyses using a variable capturing these effects on participants provided with deficit knowledge only show no effects. There is modest support for the hypothesis that individuals who learn the informational context to apply policy-specific political knowledge are more likely to shift their policy views than those who acquire political knowledge without informational context. Other key comparisons among independent variables merit further explanation below. The control variable ideology behaves as expected, with more conservative individuals being more likely to support spending cuts. Males also were more likely to support spending cuts compared to females. This finding may be driven by the selection of a social issue, health care, as the democratic issue, as women are more likely to favor social issues than men. Importantly, the knowledge variables fail to affect support for government spending cuts. In the complex issue of the national debt, knowledge must have a context for application before it impacts policy positions.
|Support for spending cuts:
|Time until debt elimination:
|Time until debt elimination:
no debt amount
debt amount only
Debt amount + context
**Statistically distinguishable at t < 0.05; * Statistically distinguishable at t < 0.1.
Table 2: Differences of means of key variables.
|Treatment: Numerical information plus context||0.26*||0.13|
|General political knowledge||-0.03||0.09|
|Preexisting debt knowledge||0.05||0.14|
|Self-assessed debt knowledge||0.05||0.06|
|Prioritization of debt||0.03||0.07|
Table 3: Differences of means of key variables.
Because political knowledge is thought to affect so many types of political behaviors, the results of this project merit further study. In particular, new work must be done in two areas. First, though the small size of the nationally-representative sample may explain why statistically significant findings were not found in its data analyses, there is a chance that the results of the student sample were driven, at least in part, by the age of those in the sample. Additional work must to done to exclude (or explore) the possibility that “youthful ignorance,” so to speak, causes effects (and lack of effects of knowledge) unseen among the general population. Nevertheless, because the direction of the results of the national sample pointed in the same direction as the student sample, there is reason believe results may translate across age groups. Second, new subject areas and issues must be tested. Is the issue of the national debt an exception to the rule, and thus of little concern, or is it just one of many areas where individuals may know policy-specific information without that information causing an observable effect on their behaviors? If subsequent research demonstrates further differences between the effects of applied knowledge and acquired knowledge only, political scientists must take increased care that political knowledge batteries gauge applied political knowledge rather than acquired knowledge only. This is particularly important for surveys utilizing political knowledge as a control variable and not explicitly testing its effects. If knowledge questions errantly tap into a dimension that does not affect political behavior, a wide range of political studies could suffer. The results may also be applicable to media outlets in decisions about how much explanation of facts to give to consumers. This research in no way invalidates earlier research that suggests political knowledge matters. Just as knowledge informs good decision-making in every area of life, political knowledge will continue to guide and shape political behaviors. Nevertheless, as technology floods individuals’ minds with tidbits of potentially useless information, a distinction between types of political knowledge that do and do not matter may become increasingly important.