WHAT DO YOUNG PEOPLE BELIEVE ABOUT ECONOMIC INEQUALITY?

We begin our analysis by investigating what youth believe about recent trends in economic inequality. Specifically, the YPP survey asked, “Do you think the difference in incomes between rich people and poor people in the United States today is larger, smaller, or about the same as it was 20 years ago?” This question has long been asked on national surveys, so it enables comparisons to other periods and national norms. In addition, we chose this question because it addresses the central concern that drove expanded interest in income inequality as a political topic. As a factual matter, the evidence is also clear. The real incomes of the top 1% of US households increased by 62% between 1993 and 2013, whereas incomes of the bottom 99% increased by 7%. This means that 59% of total growth during the period was captured by only 1% of all households (Piketty and Saez Reference Piketty and Saez2003; 2015).

As shown in table 1, most students believe that differences in incomes have grown “somewhat larger” (41%) or “much larger” (30%). This percentage is only slightly lower than that observed for the adult population: 78% of adults asked the same question in the 2012 American National Election Study stated that inequality had increased in the last 20 years.

Table 1

Students’ Beliefs and Learning about Economic Inequality

WHO HAS CLASSES ON ECONOMIC INEQUALITY?

The survey also asked students how frequently they had a class in which issues of economic inequality were discussed. The responses indicated that many young people had these classes but few had many discussions. Sixty-six percent of students reported having at least one class period in which issues of economic inequality were discussed, but only 12% reported having three or more class periods devoted to the subject.

Table 2 presents the results of a logistic regression model of whether a student reported having a class in which the issue of economic inequality was discussed. ^{Footnote 3} To test whether students self-select into these classes based on their political predispositions, the model includes four variables relating to their political orientations: political interest, political knowledge, and ratings of the Democratic and Republican parties. The model also includes a set of control variables: race, age, gender, household income, mother’s education level, survey mode (telephone or online), and survey language (English or Spanish). ^{Footnote 4}

Table 2

Logit Model of Having a Class Discussion about Economic Inequality

Note: *p<0.05; **p<0.01 for a two-tail t-test.

Of all the political and demographic variables, the only one with a statistically significant effect on the likelihood of having a discussion about inequality is political interest. Thus, it appears that to the extent that there is any self-selection of students into classes in which economic inequality is discussed, it is most likely to be based on interest in politics. ^{Footnote 5} We found no evidence that exposure to these classes was based on partisan preferences or political knowledge. This finding is important. If, for example, Democrats were more likely than Republicans to take classes in which economic inequality was discussed, then partisan polarization could explain the relationship we found between taking a course and having knowledge related to the topic. Similarly, if students who were more knowledgeable about politics were especially likely to take or to recall a class in which the topic was discussed, then political knowledge might explain the apparent association with knowledge specifically related to economic inequality. Because no relationship with political knowledge or partisanship was found, we are more confident that any relationship between taking a class and having knowledge is the result of learning that occurred through classroom instruction.

HOW DO CLASSES AFFECT KNOWLEDGE OF TRENDS IN ECONOMIC INEQUALITY?

Finally, we address the question of whether classroom discussions affect students’ factual knowledge regarding economic inequality. Table 3 shows that having even one class in which these issues were discussed is associated with a substantial increase in the probability of recognizing that economic inequality has increased in recent years. Of those students who did not have any classes in which this topic was discussed, 55% stated that economic inequality was larger. This percentage increased to 79% among those who had one or two class discussions about the issue and to 87% among those who discussed it in six or more class periods.

Table 3

Beliefs about Economic Inequality Trends by Learning about Economic Inequality in School

To isolate the effects of learning about economic inequality in school from other variables that might affect knowledge regarding the issue, table 4 displays results of a logit model regarding the belief that income differences have increased. The model includes as controls the same set of variables in the model regarding class discussions about economic inequality. Consistent with Bartels’s (2009) findings for adults, we found that partisanship is related to high school students’ perceptions about economic inequality, as the Democratic Party rating has a positive, statistically significant relationship with knowledge of the trend. Moreover, even after controlling for these political orientations and other control variables, we found that having at least one class in which economic inequality was discussed has a statistically significant, positive effect on the likelihood of stating that it has increased.

Table 4

Logit Model of Knowledge about Economic Inequality

Note: *p<0.05; **p<0.01 for a two-tail t-test.

In part, the robustness of this result depends on ruling out alternative explanations. The relationship described previously between political interest and the reported number of class discussions about inequality raises the possibility that only more interested students learned factual information from these discussions, or that they remembered taking the class and also already knew this information. To test this possibility, an additional specification of the model included an interaction between a class discussion of economic inequality and political interest. This interaction term did not have a statistically significant effect on knowledge that inequality had increased, as expected if interest were driving the relationship between class discussions and previous knowledge. We also tested similar interactions between a class discussion and three other political variables: political knowledge, Democratic Party rating, and Republican Party rating. We wanted to test, for example, whether classes on economic inequality increased knowledge of its growth only among Democrats. None of these interaction terms was statistically significant. ^{Footnote 6} Thus, the effect of discussing economic inequality in school apparently increased knowledge about the issue for all students, irrespective of their interest in politics, general political knowledge, or partisanship.

The substantive implications of the relationship between classroom experiences and knowledge of economic inequality were illustrated by calculating predicted probabilities for hypothetical students. All variables were held constant except whether they had a class in which economic inequality was discussed. The first row of table 5 shows these expectations for a “typical” student in our sample ^{Footnote 7} : the predicted probability that a student who did not have a class about economic inequality would report that it had increased in the last 20 years is 0.54. This probability increased 50% to 0.81 for an otherwise identical student who had a class in which economic inequality was discussed.

Table 5

Predicted Probabilities of Stating that Economic Inequality Has Increased by Classes about Economic Inequality

Note: Entries are predicted probabilities based on the model of answering that economic inequality has increased reported in table 4.

The substantive significance becomes even starker when one compares the effects of a classroom experience addressing economic inequality on the interaction that might occur between individuals of opposite partisanship. The last two rows of table 5 display these same calculations for a “typical Democrat” and a “typical Republican,” respectively. ^{Footnote 8} For two otherwise typical students who had not taken a class in which economic inequality was discussed, there is only a 0.28 probability (0.64 x 0.44) that a Democrat and a Republican would agree that inequality had increased in the last 20 years. However, if these same two students had each taken a class about economic inequality, the joint probability that they would both be aware of the trend increases to 0.64 (0.87 x 0.74). That is, the likelihood that two students with opposing partisan attachments would have a shared understanding of this basic economic fact more than doubles if they both had been in a class in which economic inequality was discussed.