1. Opinions on the Fairness of Income
Inequality: Does Education Matter?
Kendra Peterson
December 15, 2015
SOC 307
2. Introduction
Income Inequality runs rampant through this country. It is defined as the uneven
distribution of income across a population, or the extent to which it is uneven. In other words it is
the gap between those with the most money in a society and those with the least. With income
gaps higher than they have ever been, the rich are getting much richer and the poor are creeping
along far behind (Kelly 2010). Many wonder, speculate and theorize about what is causing this
inequality. Many have ideas of how to fix it. Another way to view it is to look at what kinds of
people even care about income inequality. What affects a person’s opinion on income inequality?
For the purpose of this paper I will be answering this question in the context of education. Does
education make a person more sympathetic towards income inequality? I posit that it does. This
is demonstrated in the conceptual model below.
If any of these claims turn out to not be true, then I will fail to reject the null hypothesis. I
will be unable to show evidence that education does make a person more sympathetic towards
income inequality.
Literature Review
Does education in itself lead to sympathy towards income inequality? Many claim
it does due to a curriculum that promotes the values discussed earlier: liberty, respect, interest in
Education Sympathy towards
income inequality
Controls
Gender
Income
Race
Age
3. and sensitivity towards others. This curriculum is based on the compartmentalizing of academic
subjects. This helps to foster positive attitudes towards equality (Harðarson). These values
stimulate egalitarian opinions.
One form of opinion is known as Social Dominance Orientation (SDO) (Pratto et al 756).
This is essentially the range of attitudes about whether or not your in-group should be dominant
over out-groups. SDO suggests that if you believe your in-group should be dominant, then you
likely prefer inequality. However, those on the other end of the spectrum, feel that relations
should be equal, not hierarchical. They do not believe that one group should be “at the top of
society”. Instead they hold that each group matters equally and thus should share equal access to
resources. Just as SDO is negatively correlated with education level, it is also negatively
correlated with values such as concern for others, communality, tolerance and altruism (Pratto et
al 754). An opinion that income inequality in the US is fair and just would be contradictory to a
low SDO. Generally speaking, those at the top of society, or those that are in the dominant in-
group tend to have a higher SDO, yet research suggests that SDO decreases as education
increases (Pratto et al 756). This negative correlation is puzzling when considered in context to
the likely wealth that is achieved by higher educated people. These people would be at the
aforementioned “top of society” and would thus have every reason to want to maintain their
status. However, it could be the result of the kind of education they are experiencing.
In general higher education liberates people from their immediate culture, gives wide
exposure to different characters, and opens the mind to be receiving of new ideas (Corey). The
very act of getting an education does this for you. With the institution of affirmative action, the
increasing popularity of semesters abroad and a curriculum that teaches about many different
modes of thought, it stands to reason that a highly educated person would also have had a decent
4. amount of exposure to different kinds of people. This is the objective especially in Liberal Arts
colleges. They create an educational environment based on human interactions and
intellectualism rather than vocational skills (Seifert 2007). At these institutions individuals are
encouraged to broaden their minds to the world around them. They are expected to look beyond
their usual environments for answers, to consider how someone else might think of the situation
and to search out alternative solutions. Even without going to a Liberal Arts college, however,
students enjoy a curriculum that is designed to promote liberal values: liberty, respect, interest in
and sensitivity towards others (Hollway). With so much exposure to the outside world, many
may find themselves more sympathetic when they leave college than when they entered.
Education not only makes people more aware of those less fortunate than them, it
also lends them to be more sympathetic of those less fortunate as well. Countries with more
equality tend to have more progressive education models that lend toward sympathy (Wraga
2014). Those countries with progressive education models tend to be radical in that they have
non-selective schools, mixed ability classes, late subject specialization and measures to equalize
resources between schools. These differences instill in the educated a sense that life should be
fair, resources should be divided fairly (Wraga 2014).
One possible mediator for the hypothesis that education makes people more
sympathetic towards income inequality is the idea that education may be making people more
liberal, which in turn makes them more sympathetic. Liberalism is defined as "the genuine
concern for the welfare of genetically unrelated others and the willingness to contribute larger
proportions of private resources for the welfare of such others" (Kanazawa 2010). Preferences in
regard to economic issues are related closely to political party affiliation, with liberal Democrats
being the most concerned with economic equality (Page, Bartels and Seawright). Democrats tend
5. to be more liberal than Republicans, and Republican status is positively correlated with SDO
(Pratto et al). Thus we would expect a liberal person to have a low SDO and a person with a low
SDO to be more sympathetic towards income inequality. However, for the purposes of this paper
we will simply be studying the direct relationship between education and sympathy.
Method
Sample
In an effort to answer the thesis question, I used data extracted from the General Social
Survey (GSS). The GSS has been conducted from 1972-2014. It compiles responses in a random
sample of the U.S. population. The responses were gained via person-to-person interviews and
Computer Assisted Personal Interviewing (CAPI). Some interviews have been conducted over
the phone. Generally speaking, the sample used by the GSS is not representative, despite 57,062
respondents. The average age of respondents are 45 years old, while the average age in the US is
37 years old. 81% of respondents were white, while the U.S. is 63% white. 50.4% of the U.S. is
female, yet 56% of respondents were female. Respondents also had an average family income of
$31,481, yet the U.S. average family income is $50,500, meaning this variable is also, not
representative. The only representative variable found was education. Respondents had an
average of 12 years of education, the same average found in the U.S.
Variables
Working with the same formula, it was necessary to code our questions to fit the model.
The original question of “Years of school completed” was coded with each year representing
itself, making a continuous variable. For the purposes of this study, I posited that a person with
16 years or more could count as having completed a bachelor’s degree or more. Thus I created a
6. dummy variable called “college” in order to distinguish the group. To operationalize the concept
of sympathy towards income inequality I chose the question, “How much do you agree with the
following statement? Differences in social standing between people are acceptable because they
basically reflect what people made out of the opportunities they had.” This question was
answered on a Likert Scale with “Strongly Disagree” coded as 4, “Somewhat Disagree” coded as
3, “Somewhat Agree” coded as 2, and “Strongly Agree” coded as 1. For the control variables, I
made a dummy variable for female in place of sex, white in place of race, and over50 in place of
age.
Missing Data
In order to understand patterns in the missing data, I ran an mdesc test in STATA. This
test revealed that the sympathy variable, had 55,667 missing responses. Education level, recoded
as college, only had 165 missing variables. Our controls, sex and race only had one missing
variable each. The other control variable, age, had 203 missing responses. An nmis test found
that the vast majority of respondents, 49,885 didn’t answer only one question so I can assume
that those respondents didn’t answer the question about sympathy, but did answer the rest of the
questions. Because of the results of the nmis test, it can be assumed that the missing data had
little to no effect on the significance of those variables. For this reason, I decided not to change
any of the missing data, other than code it as missing.
Analysis
Education was a continuous variable, while the answers for the question about sympathy
were categorical. To gain a rudimentary understanding of these variables, I ran a sum test. This
gave me information about the means, standard deviations and ranges of the variables. This
information is included in Table 1. I then ran a regression analysis in order to understand how
7. these variable worked together. I also included my controls, age, race and sex in the regression
analysis in order to understand whether or not they were lurking variables. This is represented in
the OLS Model labeled Table 2. Once I understood how the variables worked together I needed a
visual to accompany them, to show what that analysis represented. I chose to use a bar graph to
represent the variables in graph-form because, once again, one was continuous and one was
categorical. This is included in Figure 1.
Results
While these results may be limited, the findings have somewhat followed the literature.
The results indicate that education is related being sympathetic towards income inequality. When
interpreting the results it is important to remember that the question operationalized for
sympathy towards income inequality, “How much do you agree with the following statement?
Differences in social standing between people are acceptable because they basically reflect what
people made out of the opportunities they had,” was answered on a Likert Scale with “Strongly
Disagree” coded as 4, “Somewhat Disagree” coded as 3, “Somewhat Agree” coded as 2, and
“Strongly Agree” coded as 1. For the purposes of this study, those that disagree with the
statement are considered more sympathetic towards inequality than those that agreed with it. So,
when we see in Table 1 that the mean response for the sympathy question is 2.132, it is
interpreted to mean that most respondents agree with the statement and are not sympathetic
towards inequality. However, the fact that this mean is greater than 2 means that enough people
disagreed with the statement to pull the mean towards 3. We can also gather from Table1 that
22% of respondents were college educated.
8. As shown in Table 2 and Figure 1, college educated respondents were more likely to
disagree with the statement. In Table 2 we see that college educated respondents had a positive
association with the question about sympathy. This means that the average answer for college
educated respondents was higher, or closer to “Somewhat Disagree” than less educated
respondents. This association could have been skewed by other variables, however, according to
Table 2, college educated respondents continue to be more likely to disagree even when
controlling for sex, race, and age. Age was the only control variable that could have been a
lurking variable. According to the results shown on Table 2, age had a negative association with
the statement, meaning that those respondents over 50 years old were more likely to agree with
the statement than those under 50.
Table 1: Descriptive Statistics, General Social Survey.
Mean SD Range
Sympathy 2.132 .750 1-4
College .221 .492 0-1
Model 1
college .262 ***
(.052)
Male .053
(.041)
White .015
(.057)
Over50 -.150 ***
(.042)
_cons 2.142 ***
(.562)
***p<.001 **p<.01 *p<.05
Note: Standard errors in paratheses.
N=1,390
Table 2. OLS Model of the Association
between Education Level and Sympathy
toward Income Inequality. General Social
Survey.
9. Figure 1 shows a little more clearly the difference between college educated respondents
and less educated respondents. The higher mean means that college educated respondents were
more likely to disagree with the statement than less educated respondents. This also means that
college educated respondents were more sympathetic towards inequality.
Conclusion
The claim that the educated care more about inequality is a bold one. It is a dispute that
can make logical sense no matter what side of the fence you sit on. We need more information
on the topic in order to fully understand it. One issue with claiming that people with a higher
education are more sympathetic towards inequality is that people with a higher education tend to
be or consider themselves to be more affluent. There are multiple reasons for this. First, those
that make it into college are generally those that had the time and resources in their younger
0
.5
1
1.5
2
2.5
Never Finished College Finished College
Sympathy towards Income Inequlity by Level of Education
10. years to dedicate to academia and extracurricular activities. At the same time, college is
expensive so most college students show the ability to afford higher education by simply being
there. Finally, those with a college education are likely to make more money later in life and thus
be more affluent. Presumably, a person who is more affluent would be less likely to be
sympathetic towards inequality. This is a topic that would need further investigation in the future
as it was not addressed in this study.
Regardless, it is important to note what this study has revealed. Those with more
education are more likely to be sympathetic toward inequality. From this we learn that we can do
more as a country to continue this trend. We can do more to teach people about inequality and
help them to be more sympathetic. If more educated people are sympathetic towards inequality,
there will be more drive to resolve it. Educated people have long been considered the most able
to make changes in society and government. Those are the people that we need on the side of
those that cannot defend themselves. Those are the people that will stand up for the little guy.
Ultimately it is now in the hands of the educators to teach the upcoming generations about
inequality and how to fix it.
11. References
Corey, David C. n.d. “Liberal Education: It's Conditions and Ends.” Perspectives on Political
Science 42(4):195–200.
Harðarson, Atli. n.d. “Equality and Academic Subjects.” Journal of Curriculum Studies
45(2):119–31.
Hollway, Michael C. n.d. “A Comparison Of the Impact of Two Liberal Arts General Education
Core Curricula on Student Humanitarian Values.” The Journal of General Education
54(3):237–66.
Kanazawa, S. 2010. “Why Liberals And Atheists Are More Intelligent.” Social Psychology
Quarterly 73(1):33–57.
Kelly, Nathan J. and Peter K. Enns. 2010. “Inequality and the Dynamics of Public Opinion: The
Self-Reinforcing Link between Economic Inequality and Mass Preferences.” American
Journal of Political Science 54(4):855–70.
Page, Benjamin I., Larry M. Bartels, and Jason Seawright. n.d. “Democracy and the Policy
Preferences of Wealthy Americans.” Perspectives on Politics 51–73.
Pratto, Felicia, Jim Sidanius, Lisa M. Stallworth, and Bertram F. Malle. n.d. “Social Dominance
Orientation: A Personality Variable Predicting Social and Political Attitudes.” Journal of
Personality and Social Psychology 741–63.
Seifert, Tricia A. et al. 2007. “The Effects of Liberal Arts Experiences on Liberal Arts
Outcomes.” Research in Higher Education 49(2):107–25.
Wraga, William G. 2014. “Condescension and Critical Sympathy: Historians of Education on
Progressive Education in the United States and England.” Paedagogica Historica
50(1/2):59–75.