This research project examines how intolerant individuals in the United States are more likely to hold authoritarian attitudes that could be threatening to democracy compared to individuals who are tolerant of the opposing political party.
Today it is not a secret that Americans are extremely divided in their politics. At best there is mistrust, not just between political parties but for the individual people in those parties. At worst there is hatred and even threats and flashes of violence. While the United States has
mostly been dominated by two opposing political parties for decades, the pewresearch center says "on average, Democrats and Republicans are farther apart ideologically today than at any time in the past 50 years.” According to this Pew Research, both parties have moved further and further away from the center since the early 1970’s, and since 2004 any overlap between them has pretty much disappeared. Julian Zelizer, a Princeton historian and co-author of How the Tumultuous 70’s Shaped Our Political Conflicts, says the causes for this polarization have roots from 1974 when our country dealt with many major events such as the chaos in trying to end the polarizing Vietnam war, an energy crisis and the first indications of global warming, Watergate and the awareness that a president was not necessarily an honorable person, and the issue of race which boiled over with school busing riots. Zelizer says our politics began to take on a “cultural outrage” which has continued to grow and now has become routine. As a result, Americans are now becoming aware that their distrust and even hatred of the other party has grown enough to cause serious damage to the idea of democracy itself, something our founding fathers and our constitution tried so hard to preserve. As shown in the Atlantic, a 2016 survey showed that approximately 9 out of 10 of both the supporters of Trump and Biden were sure that electing the other party would result in permanent harm to the country.
As shown in the image above, recently we saw how a large division of our country was so threatened by the idea of the opposing party taking control that they were ready to follow a president with obvious authoritarian tendencies, reject the outcome of an election, and some were even willing to take part in an insurrection on our Capital to try and prevent the results of that election from being certified. It is becoming an alarming common knowledge across the country that our democracy is in danger as shown in this month’s over the counter
This Time magazine cover which is simply entitled “Democracy” shows our flag disintegrating into pieces and scribbled out with a black sharpie. The concern that we may lose our democracy seems to have had an impact on how people voted in the 2020 election where traditionalist Joe Biden won over the authoritarian leaning Donald Trump. While there are many academic studies on the separate paths our political parties have taken, we have also become aware that on an individual level, people in this country have internalized these political differences. They
are no longer just policy disagreements where one could have an honest debate, but the political parties have become intertwined in our personal identities with strong emotional attachment and no room for open minded discussion. People now tend to socialize with others who have the same political ideas and even live in homogenous areas. As shown in this article, a 2016 survey indicated that 60% of Democrats and 63% of Republicans would “balk” if their child wanted to marry someone aligned with a different party . Given the threat of increasing polarization in America, I asked the question: Do citizens who have lower levels of mutual toleration hold other sets of attitudes that might be threatening to the survival of democracy?
There is a history of political science studies about worldwide democracies including how to keep them from backsliding. Sometimes conclusions have been drawn that America appears to be an exception to the usual patterns because our democracy was standing so
strong. In considering the question that is the theme of this project, the following is just a small sample of studies that were reviewed which deal with related issues.
This is a study from Socius: Sociological Research for a Dynamic World. The authors, Christopher R. H. Garneau and Philip Schwadel show how being affiliated with certain political parties means someone will be more or less tolerant based on the political party. Although my study is not discussing whether democrats or republicans are more or less tolerant, it is still important to note that individuals in certain political parties seem to be more tolerant than others.
This is a Pew Research study from 2016 that looks at how polarized we are in this country and the “highly negative views of the opposing party—and its members.” It shows that people have feelings of anger/frustration and even fear toward the other party and see that party as “threatening.” These feelings generally make people more likely to support an authoritarian attitude in order to ensure their party wins. Interestingly in some cases those who have friends within the opposing party are more likely to have toleration by viewing them less “coldly.” Overall, this study shows that Americans have extremely negative and fearful feelings about the opposing political party.
This is a Pew Research study from 2019 which shows even more alarming results of the polarization and lack of tolerance as described in the first two examples. It describes how some people in the U.S. are actually moving toward the idea of being more willing to expand presidential power, even becoming more authoritarian which threatens an end to our democracy. The survey shows that Republicans increasingly believe that problems in the USA could be better solved “if presidents didn’t have to worry so much about Congress or the courts,” which is actually a question I used in my study. Nearly half of general republicans in 2019 felt like presidents could get much more done if they didn’t have to deal with the other branches of government, a fundamental part of our democracy. This shows that many people in America today are holding authoritarian attitudes.
Given that many individuals are more likely to support extremist candidates to ensure that their party wins, I expect that the more intolerant a civilian is, the more likely they are to hold an authoritarian attitude. Moreover, this means that my null hypothesis is that there is no relationship between being intolerant and being more likely to hold an authoritarian attitude. In this case my independent variable, mutual tolerance, means the extent to which citizens view the opposing political party or other people with opposite political views as legitimate and whether they have the right to exist. My dependent variable, an authoritarian attitude, essentially means an attitude that threatens or impedes the progress of democracy. These attitudes can include many things, such as views about democracy itself, beliefs about how power should be distributed in politics, the importance of laws/policies, etc. Although citizens do not have the power themselves to propose, regulate, enforce, etc. certain policies in the government, they can have attitudes toward democracy that are powerful and promote or possibly even cause backsliding. Additionally, this means that the more tolerant a respondent is, the more likely they are to support a democratic attitude, a positive relationship.
The sample I used for this study was the American National Election Survey Time Series data. This is a study that interviewed over thousands of participants in recent years and is a random sampling of the United States. The ANES has a variety of questions that measure individuals’ democratic and authoritarian attitudes, feelings toward organizations, feelings toward groups of people, and much more. For this study, I used a variety of questions that measure a different type of attitude that can threaten democracy as well as a group of questions to establish how tolerant or intolerant an individual was. While there is no specific survey question that explicitly has “mutual toleration,” there are many different questions in the ANES that can help measure people's views of the opposite political party that relates to violating mutual tolerance. In order to measure mutual tolerance, I first I had to find out which party the respondents affiliated with, as without knowing the respondent’s party ID, it would be impossible to say what party they “oppose”. This means that party ID is used to show how Democrats feel about Republicans and how Republicans feel about Democrats. After taking into account the respondent’s political party, I used four questions to measure how tolerant or intolerant a respondent was. For example, one the questions asked was “is there anything you like about the Democratic party?” and the respondent either answered with a yes or a no. The other questions I used were “is there anything you dislike about the Democratic party?” “Is there anything you like about the Republican party?, and “is there anything you dislike about the Republican party?” Again, all of these questions the respondent had to answer either with a yes or no. In order to measure the respondent’s tolerance, I coded a binary variable that took into account the respondent’s party ID while also measuring if there was anything they disliked about their own party, anything they liked about the opposing party, or anything they disliked about the opposing party. The intolerant group, which is called 1 in the coding, is for respondents who disliked something about the opposing party and liked nothing about the opposing party. Tolerant respondents, which were coded as 2, were respondents who like something about the opposing party but also dislike something about the opposing party and for respondents who like something about the opposing party and disliked nothing about the opposing. I feel that this is a good way to measure mutual tolerance, as it shows that if the respondent likes nothing about the opposing party and only dislikes it, they are more likely to find the opposing party as not legitimate and are less likely to tolerate the other party. Moreover, if the respondent dislikes the opposing party but also likes something about it, they are more likely to see the opposing party as legitimate and are more likely to tolerate the other party. Therefore, my mutual tolerance measure takes in to account the respondents ID and uses the questions already discussed in order to see whether a respondent is tolerant or intolerant.
My dependent variable is the attitudes citizens have that are authoritarian and overall threaten democracy. In order to measure this phenomenon, I used four different survey questions, all of which measure a different type of authoritarian attitude. The questions measure people's trust with democratic institutions, rights of journalists, and beliefs about what politicians should or should not be able to do. The first question was “Would it be helpful, harmful, or neither helpful nor harmful if U.S. presidents could work on the country’s problems without paying attention to what Congress and the courts say? and the respondent had to reply with helpful, harmful, or neither helpful nor harmful. I recoded this variable in order to better visualize whether being more intolerant makes the respondent more likely to hold an authoritarian attitude. “Helpful” was coded as 0, 1 was coded as neither helpful nor harmful, and 2 was coded harmful. I feel this is a good measure of an authoritarian attitude that threatens democracy, as it is measuring the extent to which the respondent feels getting rid of a democratic institutions (congress and the courts) and letting a certain branch of government have more power than the rest is a good idea. My second question I used was “Would you prefer a government official who compromises to get things done, or who sticks to their principles no matter what? “And the respondent either answered with compromises or sticks. I coded the variable so that 1 is for respondents who said sticks and 2 is for respondents who said compromises. This is a good measure of an attitude that threatens democracy, as it is measuring how willing people would be to support a president that breaks the “rules” or follows their own rules in order to get what they want at the expense of democracy. The third question I used was “do you favor, oppose, or neither favor nor oppose elected officials restricting journalists’ access to information about government decision-making?” and the respondent had to answer with either favor, oppose, or neither favor nor oppose. I coded 0 as favor, 1 as neither favor nor oppose, and 2 is oppose. I feel this is a good measure of an attitude that threatens democracy, as it is measuring the extent to which the respondent feels getting rid of a democratic right would be a good idea. The final question I used to measure whether a respondent holds an authoritarian attitude is “In general, how appropriate or inappropriate is it for the president of the United States to ask leaders of foreign countries to investigate political rivals?” and the respondent had to respond with either appropriate, inappropriate, or neither appropriate nor inappropriate. I coded this variable so that 0 is appropriate, 1 is neither appropriate nor inappropriate, and 2 is inappropriate. I feel this is a good measure of an authoritarian attitude that threatens democracy, as it is measuring the extent to which the respondent is willing to let foreign countries threaten democracy in order to make sure the political candidate they support wins. Overall, I feel that these four questions helped best capture an authoritarian attitude, as although all of them are different, they do test if a respondent has an attitude that threatens democracy.
This histogram is a visual representation of the question “Do you prefer a leader who compromises to get things done or sticks to their own principles no matter what?” Here, the minimum is 1 since it represents preferring a leader who sticks to their own principles no matter what and the maximum is 2 since it represents preferring a leader that compromises to get things done.
This histogram is a visual representation of the question “Do you favor, oppose, or neither favor nor oppose elected officials restricting journalists’ access to information about government decision-making?” Here, the minimum is 0 which represents favor and the maximum is 2 which represents oppose.
This histogram is a visual representation of the question “Would it be helpful, harmful, or neither helpful nor harmful if U.S. presidents could work on the country’s problems without paying attention to what Congress and the courts say?” Here, the minimum is 0 which represents it is helpful and the maximum is 2 which represents it is harmful.
This histogram is a visual representation of the question “In general, how appropriate or inappropriate is it for the president of the United States to ask leaders of foreign countries to investigate political rivals?” Here, the minimum is 0 which represents it is appropriate and the maximum is 2 which represents it is inappropriate.
The histograms above show how the democratic and authoritarian views are distributed among respondents. As shown, a large majority of respondents hold a democratic attitude. Even so, at least above 10% of respondents in each histogram hold an authoritarian attitude, which is very large considering the ANES is representative of the entire United States population.
Finally, I decided to control for three variables about the respondent’s demographics. My control variables (independent variables) all could potentially influence or cause the relationship between mutual tolerance and holding an authoritarian attitude. The three demographics I decided to control for were the respondent’s level of education, yearly family income, and ethnicity. In order to control for these variables, I also used survey questions from the ANES Time Series data. When controlling for education, the question I used was “What is the highest level of school you have completed or the highest degree you have received?”. I coded this in a way so that 0 represents if the respondent had an education of high school or less, 1 represents if the respondent had an education of some college to an associate’s degree, and 2 represents if the respondent had an education of a bachelor’s degree and above. I decided to code the variable in this way because it showed the effect if the respondent had a low education (0), a moderate amount of education (1) and a high level of education (2). When controlling for yearly family income, the question I used included asking the respondent which range of income do they have with over 22 different ranges all spanning between under $9,999 to over $250,000. I coded this variable so that 0 would be respondents who had a yearly family income of under $9,999 to $49,999, 1 would be respondents who had a yearly family income of $50,000 to $74,999, and 2 would be respondents who had a yearly family income of $75,000 to $250,000 or more. I coded income this way so that 0 represents a respondent with a low family income, 1 represents a respondent with a moderate family income, and 2 represents a respondent with a high family income. When controlling for ethnicity, the question I used asked for the respondent’s race, which gave nine different ethnicities to choose from. I coded this variable so that 1 represents if the respondent was white and 2 represents if the respondent did not identify as white. The reason I coded this variable like this is because I mainly wanted to see the effects of being white vs not being white. Therefore, the way I coded the variables will effectively show if ethnicity has an effect on the relationship between mutual tolerance and holding an authoritarian attitude.
In order to effectively measure the relationship between mutual tolerance and authoritarian attitudes, I used contingency tables and linear regressions. I used the contingency tables to show whether a respondent who was intolerant or tolerant was more likely to hold an authoritarian attitude or a democratic attitude and I used the linear regressions to show if the relationship found in the contingency tables (which was the expected effect) still held true even while controlling for level of education, yearly family income, and ethnicity.
Results
Based on the contingency table, my predicted hypothesis was correct. When looking under category 1 of the contingency table (leader who sticks to their own principles no matter what), I observed the expected effect. The intolerant (1) group had a total of 36.8% of respondents say they prefer a leader who sticks to their own principles, while the tolerant group(2) had only 30.5% of respondents say they prefer a leader who sticks to their principles. This is what I expected, as the more tolerant someone was, the less likely they were to say they prefer a leader who sticks to their own principles (authoritarian attitude) and the more intolerant someone was, the more likely they were to say they prefer a leader who sticks to their own principles (authoritarian attitude). Additionally, under category 2 of the contingency table (prefer a leader that compromises to get things done), I also observed the expected effect. The more tolerant a respondent was, the more likely they were to prefer a leader that compromises to get things done (democratic attitude). Still looking at category 2, we can see that the intolerant group (1) had a total of 63.2% of respondents that said they preferred a leader who compromises while 69.5% of respondents in the tolerant group said they preferred a leader who compromises. This was my expected effect, as I predicted that the more intolerant a respondent was, the less likely they would support a democratic attitude and the more tolerant a respondent was, the more likely they would support a democratic attitude. Moreover, the relationship is statistically significant, as there is a less than .1% chance that the relationship was due to random sampling error. Ultimately, the more intolerant a respondent was, the more likely they were to prefer a leader who sticks to their own principles no matter what, that is, an authoritarian attitude. Interestingly, whether a respondent was intolerant or not, most respondents said they preferred a leader who compromises to get things done.
Overall, my predicted hypothesis was correct. Under category 0 of the contingency table, that is, favoring restricting journalist access, I received the expected effect. The intolerant group had a total of 18% of respondents favor restricting journalist access while the tolerant group had a total of 14.5% of respondents favor restricting journalist access. This was my expected effect, as the more intolerant someone was, the more likely they were to support an authoritarian attitude, that is, favoring restricting journalist access and the more tolerant a respondent was, the less likely they would support an authoritarian attitude. When looking under category 2 of the contingency table (opposing restricting journalist access), I observed the expected effect. The intolerant group had a total of 59.7% of respondents say they oppose restricting journalist access, while the tolerant group had a total of 62.5% of respondents say they oppose restricting journalist access. This is what I expected, as the more tolerant someone was, the more likely they were to oppose restricting journalist access (which is a democratic attitude) and the more intolerant someone was, the less likely they were to oppose restricting journalist access. The relationship shown here is statistically significant, as there is a .5% chance that the relationship was due to random sampling error. Ultimately, the more intolerant a respondent was, the more likely they were to favor restricting journalist access.
As shown above, my predicted hypothesis was correct. When looking at the 0 category, that is, not paying attention is helpful, I observed the expected effect. The intolerant group had a total of 12.7% of respondents say they found not paying attention to be helpful while the tolerant group had a total of 10.8% of respondents said they found not paying attention to be helpful. This was my expected effect, as the more intolerant someone was, the more likely they were to support an authoritarian attitude, that is, it is helpful for U.S. presidents to work without paying attention to the congress/courts and the more tolerant a respondent was, the less likely they would support an authoritarian attitude, not paying attention is helpful. When looking under category 2 of the contingency table (not paying attention is harmful), I observed the expected effect. The intolerant group had a total of 68.6% of respondents say they found not paying attention to be harmful, while the tolerant group had a total of 74.5% of respondents that said they found not paying attention to be harmful. This is what I expected, as the more intolerant someone was, the less likely they were to support a democratic attitude, that is, it is harmful for U.S. presidents to work without paying attention to the congress/courts, and the more tolerant someone was, the more likely they were to support a democratic attitude, saying it is harmful. The relationship shown here is statistically significant, as there is a less than .1% chance that the relationship was due to random sampling error. Ultimately, the more intolerant a respondent was, the more likely they were to say it was helpful for presidents to work without having to pay attention to the congress/courts.
As shown above, my predicted hypothesis was correct. Under category 0 of the contingency table (it is appropriate) the more tolerant a respondent was, the less likely they were to say it is appropriate for the president to ask foreign countries to investigate political rivals. Still looking at category 0, we can see that the intolerant group (1) had a total of 13.2% of respondents say it was appropriate while 10% of respondents in the tolerant group said it was appropriate. This was my expected effect, as I predicted that the more intolerant a respondent was, the more likely they would support an authoritarian attitude, in this case, saying it is appropriate for the president to ask foreign countries to investigate political rivals and the more tolerant a respondent was, the less likely they would support an authoritarian attitude (it is appropriate). When looking under category 1 of the contingency table (it is neither appropriate nor inappropriate) I observed the expected effect. Finally, when looking at the 2 category, that is, it is inappropriate for presidents to ask, I observed the expected effect. The intolerant group had a total of 68.4% of respondents say it is inappropriate for a president to ask while the tolerant group had a total of 75.4% of respondents say it is inappropriate for a president to ask. This was my expected effect, as the more intolerant someone was, the less likely they were to support a democratic attitude, that is, saying it is inappropriate for a president to ask and the more tolerant a respondent was, the more likely they would support a democratic attitude. The relationship shown here is statistically significant, as there is a less than .1% chance that the relationship was due to random sampling error. Ultimately, the more intolerant a respondent was, the more likely they were to say it is appropriate for a president to ask foreign countries to investigate political rivals.
As shown in the bar plot, when looking at preferring a leader who compromises to get things done (2), there was increase when moving from being intolerant to being tolerant. This shows that the more tolerant the respondent was, the more likely they were supported a democratic attitude, that is, compromising to get things done. This also shows that the intolerant group was less likely to support a democratic attitude, that is, less likely to support a leader compromising to get things done. This is shown further when looking at preferring a leader who sticks to their own principles no matter what (1), as there was a decrease when moving from being intolerant to being tolerant. This shows that the more intolerant a respondent was, the more likely they were to support an authoritarian attitude, that is, supporting a leader who sticks to their own principles no matter what. This also shows that the tolerant group was less likely to support an authoritarian attitude, that is, supporting a leader who sticks to their own principles no matter what.
As shown in the bar plot, when looking at favoring restricting journalist access (0), there was a decrease when moving from being intolerant to being tolerant. This means that the more intolerant a respondent was, the more likely they were to support an authoritarian attitude, that is, favoring restricting journalist access compared to the tolerant respondents, who were less likely to support the authoritarian attitude. When looking at opposing journalist access (2) there was an increase when moving from being intolerant to being tolerant. This means that the more tolerant the respondent was, the more likely they supported a democratic attitude, that is, opposing restricting journalist access. Moreover, the more intolerant the respondent was, the less likely they supported a democratic attitude.
As shown in the bar plot, when looking at the president not having to pay attention to be helpful (0), there was a decrease when moving from being intolerant to being tolerant. As shown, the more intolerant a respondent was, the more likely they were to support an authoritarian attitude, that is, finding the president not having to pay attention to the congress or courts to be helpful, compared to the tolerant respondents, who were less likely to support the authoritarian attitude. When looking at finding not paying attention to be harmful (2) there is an increase when moving from being intolerant to being tolerant. This means that the more tolerant the respondent was, the more likely they supported a democratic attitude, that is, finding the president not having to pay attention to be harmful. Moreover, the more intolerant the respondent was, the less likely they support a democratic attitude.
As shown in the bar plot, when looking at, it is appropriate (0), there was a decrease when moving from being intolerant to being tolerant. As shown, the more intolerant a respondent was, the more likely they were to support an authoritarian attitude, that is, saying it was appropriate for presidents to ask compared to the tolerant respondents, who were less likely to support the authoritarian attitude. When looking at, it is inappropriate for presidents to ask (2), there was an increase when moving from being intolerant to being tolerant. As shown, the more intolerant a respondent was, the less likely they were to support a democratic attitude, that is, saying it is inappropriate for presidents to ask foreign countries to investigate rivals and the more tolerant a respondent was, the more likely they were to support the democratic attitude.
As explained earlier, I had a total of three control/independent variables, all of which are demographics. The first is level of education, the second is family income, and the third is ethnicity. Overall, while controlling for these variables, I found that all the relationships were still significant except for restricting journalist access while controlling for education. When looking the column “compromises or sticks” and the row “mutual tolerance,” every time I had a one unit increase in the mutual tolerance scale; I had an increase in the respondent preferring a leader who compromises to get things done, which was my expected effect(.0631). Since the scale moves from intolerant to tolerant, this essentially means that the more tolerant someone became, the more likely they were to prefer a leader who compromises. Additionally, respondents who were intolerant (1) were more likely to prefer a leader who sticks to their own principles no matter what than the respondents who were tolerant. The relationship produced here is statistically significant, as there is a less than .1% that the relationship was caused by random sampling. Ultimately, the less tolerant/more intolerant a respondent was, the more likely they were to hold an authoritarian attitude under the column of “compromises or sticks” (preferring a leader who sticks to their own principles no matter what). While controlling for the respondents’ level of education, yearly family income, and ethnicity under the column “compromises or sticks, “the original relationship was still significant, that is, as the more intolerant a respondent became, the more likely they were to support an authoritarian attitude (preferring a leader who sticks to their own principles no matter what). When controlling for level of education, the slope was still positive (.0555) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. When controlling for yearly family income, the slope was still positive (.0524) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. Finally, when controlling for the ethnicity of the respondent, the slope was still positive (.0623) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. Therefore, the more intolerant a respondent was, the more likely they were to have an authoritarian attitude (preferring a leader who sticks to their own principles no matter what) while controlling for education, family income, and ethnicity.
When looking at the “restrict journalist access” column and the row “mutual tolerance” I observed the expected effect. Every time I had a one unit increase in the mutual tolerance scale; I had an increase in the respondent opposing restricting journalist access to government decision making, which is a positive relationship(.0629). Since the scale moves from intolerant to tolerant, this essentially means that the more tolerant someone became the more likely they were to oppose restricting journalist access to government decision making. Additionally, respondents who were intolerant (1) were more likely to favor restricting journalist access compared to the respondents who were tolerant. The relationship produced here is statistically significant, as there was a .5% that the relationship was caused by random sampling. Ultimately, the less tolerant/more intolerant a respondent is, the more likely they are to hold an authoritarian attitude under the column “restrict journalist access” (favoring restricting journalist access to government decision making). While controlling for the respondents’ level of education, yearly family income, and ethnicity under the column “restrict journalist access”, the original relationship was still significant, that is, as the more intolerant a respondent became, the more likely they were to support an authoritarian attitude (preferring a leader who sticks to their own principles no matter what) when controlling for family income and ethnicity. However, when controlling for education, the original relationship was not statistically significant, as there was a 6.7% chance that the original relationship was due to random sampling error. Even though this is technically not statistically significant, there is still a 93.3% chance that the more intolerant a respondent was, the more likely they were to favor restricting journalist access. Moreover, when controlling for level of education, the slope was still positive (.0403). When controlling for yearly family income, the slope was still positive (.0539) and the original relationship was still statistically significant, as there was a less than 1.8% chance that the results were due to random sampling error. Finally, when controlling for the ethnicity of the respondent, the slope was still positive (.0663) and the original relationship was still statistically significant, as there was a .3% chance that the results were due to random sampling error. When it comes to education, there was an interaction effect, meaning the relationship between being intolerant making someone more likely to favor restricting journalist access (authoritarian attitude) is dependent on that individual’s level of education. Therefore, the more intolerant a respondent was, the more likely they were to have an authoritarian attitude (preferring a leader who sticks to their own principles no matter what) while controlling for family income and ethnicity.
When looking at the column “political rivals” and the row “mutual tolerance” I observed the expected effect. Every time I had a one unit increase in the mutual tolerance scale; I had an increase in the respondent finding it harmful for the president to work without having to pay attention to the congress and courts, which is a positive relationship(.101). Since the scale moves from intolerant to tolerant, this essentially means that the more tolerant someone became the more likely they were to say it is harmful for the president to not pay attention. Additionally, respondents who were intolerant (1) were more likely to say it is helpful if the president could work without having to pay attention to the congress or courts compared to the respondents who were tolerant. The relationship produced here was statistically significant, as there was a less than .1% that the relationship was caused by random sampling. Ultimately, the less tolerant or more intolerant a respondent was, the more likely they were to hold an authoritarian attitude under the column “political rivals” (it is helpful if the president could work without having to pay attention to the congress or courts). While controlling for the respondents’ level of education, yearly family income, and ethnicity under the column “political rivals” the original relationship was still significant, that is, as the more intolerant a respondent became, the more likely they were to support an authoritarian attitude (saying it is appropriate for presidents to ask foreign countries to investigate political rivals). When controlling for level of education, the slope was still positive (.0864) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. When controlling for yearly family income, the slope was still positive (.0849) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. Finally, when controlling for the ethnicity of the respondent, the slope was still positive (.0969) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. Therefore, the more intolerant a respondent was, the more likely they were to have an authoritarian attitude (saying it appropriate for presidents to ask foreign countries to investigate political rivals) while controlling for education, family income, and ethnicity.
When looking at the column “pres paying no attention” and the row “mutual tolerance”, I observed the expected effect. Every time I had a one unit increase in the mutual tolerance scale, meaning going from intolerant to tolerant, I had an increase in the respondent saying it was inappropriate for presidents to ask foreign countries to investigate political rivals, which is a positive relationship(.0784). Since the scale moves from intolerant to tolerant, this essentially means that as the respondent became more tolerant, the more likely they were to say it is inappropriate for the president to ask foreign countries to investigate political rivals. Additionally, the respondents who were considered intolerant (1) were more likely to say it is appropriate for presidents to ask foreign countries to investigate political rivals and the respondents who were tolerant (2) were more likely than the intolerant group to say it is inappropriate. This was my expected effect, as the less tolerant a respondent was or the more intolerant a respondent was the more likely they were to support an authoritarian attitude, that is, saying it is appropriate for the president to ask foreign countries to investigate political rivals. The relationship shown here (being intolerant makes someone more like to support a leader who sticks to their own principles no matter what) was statistically significant, as there was a less than one .1% that the relationship was caused by random sampling. While controlling for the respondents’ level of education, yearly family income, and ethnicity under the column “political rivals” the original relationship was still significant, that is, as the more intolerant a respondent became, the more likely they were to support an authoritarian attitude (saying it would be helpful if the president could work without having to worry about the congress or the courts). When controlling for level of education, the slope was still positive (.0577) and the original relationship was still statistically significant, as there was a less than .4% chance that the results were due to random sampling error. When controlling for yearly family income, the slope was still positive (.0686) and the original relationship was still statistically significant, as there was a .1% chance that the results were due to random sampling error. Finally, when controlling for the ethnicity of the respondent, the slope was still positive (.0790) and the original relationship was still statistically significant, as there was a less than .1% chance that the results were due to random sampling error. Therefore, the more intolerant a respondent was, the more likely they were to have an authoritarian attitude (saying it would be helpful if the president could work without having to pay attention to the congress or courts while controlling for education, family income, and ethnicity.
Based on the contingency tables and linear regression, there is clearly a significant relationship between being intolerant and being more likely to hold an authoritarian attitude. The contingency tables show that there is an increase when moving from being intolerant to tolerant and responding with a non- authoritarian/ democratic response. Additionally, when moving from being tolerant to intolerant, there is in an increase in holding an authoritarian attitude. As the linear regressions show, the hypothesis still holds true even while controlling for the respondents’ level of education, yearly family income, and ethnicity. Therefore, I reject the null hypothesis, that there is no relationship between being intolerant and being more likely to hold an authoritarian attitude.
Ultimately, we can conclude that my results support my hypothesis, that is, individuals who violate mutual tolerance are more likely to hold an authoritarian attitude that threatens democracy. Interestingly, as shown by the histograms, frequency tables, and contingency tables, there was still a large majority of people that preferred the democratic attitude rather than the authoritarian, even when the respondent was intolerant. This may be due to how the questions were worded, as a respondent may be inclined to hold a democratic attitude because it sounds more politically correct. Additionally, Milan W. Svolik discusses in the article "Polarization versus Democracy" that a majority of people still value democratic norms and democracy as a whole. However, many voters are willing to drop democratic norms for their own partisan interests. This could explain why so many people answered with a democratic attitude, as the majority of people clearly still valued democracy. However, if the questions presented to the respondents were worded in way that made the respondent choose between an authoritarian attitude that supported their partisan interest versus holding up a democratic principle, we might have seen even more respondents hold up an authoritarian attitude. In future studies, this would be an interesting phenomenon to study and take into account for when measuring respondents’ tolerance and holding up an authoritarian attitude.
If other scholars are interested in studying this phenomenon in the future, an improvement they could make on my design would be how I measured mutual tolerance. I wanted to work in the realm of recent data, so I had to use the question “is there anything you like or dislike about the Republican/Democratic party, yes, or no?” questions. However, in the ANES , there are different questions that could better capture someone’s mutual tolerance that use older data. For example, there was a question that asked the number of likes/dislikes someone had towards the Democratic/Republican party. This could better capture the idea of mutual tolerance, as rather than just having a binary measure where the respondent is either tolerant or intolerant, someone could make a measure where, for example, a respondent having 5 dislikes means they are more intolerant than a respondent with 3 dislikes. This better captures the idea of tolerance, as some people are more tolerant than others and it would be better representation of reality rather than just putting people into 2 separate categories. As always, there are many more survey questions that capture authoritarian attitudes that threaten democracy, and it would be interesting to explore more. Obviously, I was not able to control for every rival explanation, so it would be interesting to see if being intolerant makes someone more likely to hold an authoritarian attitude is still true when controlling for the age of the respondent and the gender of the respondent.
Looking at these conclusions in light of today’s political differences and growing polarization, we can see a little more deeply into the notion of tolerance and what attitudes people with differing levels of tolerance hold. By learning more about these and related ideas we may come to a better understanding what can be done to halt and reverse the increasing intolerance that is a part of our country’s polarized populace and the damage it is taking on our democracy. Americans have shown over our history, a unique ability of overcoming chaotic times as we did at the end of the Civil War. Today there is an indication in the data that even people with low tolerance still claim to want to live in a democracy. Because it is probably unlikely that people will release their strongly held beliefs, it is important to continue the study of how to promote tolerance. According to John Avalon, Senior Political Analyst at CNN, there is more agreement in this country than we think, even while other studies claim no overlap exists anymore. One of the ways to increase tolerance is to find and shine a light on any areas of agreement. Avalon cites studies that show that our opinions of the other party are “wildly off-base.” For example Democrats hold the idea that only half of all Republicans believe that racism exists in our country and Republicans say that only half of the Democrats are even proud to be an American. The statistics in the study, however, show that 80% of Republicans know that racism is still an issue here and 80% of Democrats identify as being proud to be an American. Obviously, we need to find ways to better highlight the positive side of our opponents, even without agreeing with them. But first we have to find ways to increase our desire to do that. It seems our social media is overwhelmingly negative and hate speech is rampant. We need to identify ways to minimize this, perhaps by better regulating these platforms to tamp down the damage. Avalon identifies other solutions such as working on our civic education in schools. We need everyone to learn clearly why and how democracy works and to understand our struggle to maintain it throughout our history. He suggests there should be a test, the same test for everyone, before graduating high school to ensure our young people have grasped this basic knowledge of our country as they go out into the adult world. This common foundation might prevent susceptibility of falling prey to conspiracy theorists and disinformation and again, increase levels of understanding. Avalon also says it is important to return to the work of solidifying our middle class to ensure equal opportunity to social mobility in order to prevent desperate people from turning to populist ideas that have proven dangerous and potentially violent. We need, he says, a broad coalition of people who understand these things, and to take steps to implement solutions.
There are many ideas about ways to restructure our electoral system to change from our politically biased and winner-take-all elections which exaggerate our differences and inflame people. One idea is to increase the number of political parties so that the growing hatred between the two major parties and the way they see one another as the ultimate enemy, could be defused. Another common idea is to do something about the extreme gerrymandering that increases frustration and tension in our primaries and leaves us with many extreme leaders whose primary goal is to denigrate the other party. Parties drawn on district lines based on nothing but political advantage do not equate to fairness in elections and many feel their vote is being stolen. Yasha Mounk, author of The Great Experiment: Why Diverse Democracies Fall Apart and How They Can Endure, says we need institutional reform and that our current excessive gerrymandering allows the most extreme members of the party to have considerable control of primary elections, resulting in more divisive leaders in Congress. Divisive leaders obviously find ways to divide people, and discourage highlighting anything good about their opponent, and of course toleration suffers. Mounk suggests that districting should be taken out of party control as has been done in several states. Producing more centrist candidates, willing to listen to both sides of the political aisle, might be able to drown out the current loud voices of the extremes.
In addition to the ideas political scientists are putting forward to change our electoral systems to try reducing polarization, there are also social sciences that are working in that regard. For instance, the Greater Good Science Center at the University of California, Berkley discusses related ideas from a social psychology viewpoint. They believe the way to increase toleration in our culture is to work on developing a system that provides more meaningful intergroup contact. There are so many ways on social media to anonymously spew hatred, they say there needs to be a place for groups of people to come together. They propose something called “Citizen Assemblies" to discuss difficult political issues in a format that encourages finding common ground. This was done in Ireland and has been credited with advances in the way the country approaches climate change. “Citizen Assembly on Brexit" managed by the University College of London decided on a compromise that might have influenced the effort of UK to break from the EU. We need to be better able to see things from another’s perspective. According to this same article, Pinker, who wrote The Better Angels of our Nature, says books that were written showing the perspective of black slaves, may actually have had an effect during the abolition of slavery. Some of our current media outlets promote the outrage in our society because it is good for sales and ratings. Maybe journalists and authors could be encouraged to provide powerful stories that highlight the humanity of our people and increases our tolerance of one another. This article also suggests we should have more referendums on our electoral ballots where people can choose to voice their actual opinions on issues rather than just voting for a politician who may have more self-serving ideas. This may take the partisan divide out of the picture a bit and give voters a more thoughtful, less toxic way to vote. It would potentially encourage more ways to agree with a variety of people over issues across the political divide. None of these methods to change the destructive polarized path we are on would be easy but understanding some of the issues relating to toleration may help. Even though the research in this paper is not particularly hopeful and may even show that at the current time intolerance is a major problem, we still need to try to find solution to polarization. Based on the increasing public knowledge that we are in danger of losing our democracy, perhaps people will be alarmed enough to want to find another way to live. The research in this paper is just small attempt to look closer at toleration and the attitudes attached to it. Hopefully as more people understand the need to change, more focus will be put on this kind of research. Tolerance is a subject worth studying and may, in the end, save us from ourselves.