The days after the election on very few hours of sleep, I made up a survey to figure out what I missed to try and make sense of what happened. Thank you to the respondents for humoring me and taking time out of their days to answer it. As promised, I’m posting the data and results. I hope others find them as helpful as I have. I realize it it lengthy (on Google Docs, it’s 18 pages long single spaced… whoops) so I’m going to break this up into a few posts. Below you will find the pseudo-introduction, links to the survey results, and my brief discussion of the results. Really I could talk for pages but I tried to restrain myself to what was reasonable to extract. Also, I apologize for any redundancies or misspellings. I tried to copy and paste as much as possible moving over from the collection form, but this has been edited and rearranged several times so it’s possible something may have been messed up. And now on to the main event…
Why a survey?
I created this survey the day after the election, because I wanted to answer the question of “Why Trump?”. For me and many people I know, the result of this election just blew us away. I never actually considered this a possibility, never mind think that it would become my reality. But yet here we are. So to help me, and other people like me, figure out what we missed throughout the election cycle, and how to move forward, I made this survey.
“Why didn’t you talk to people?” I recently moved away from home to one of the most blue areas in the nation. I can’t sit down and talk to people that I know as I am hundreds of miles away, and I can’t talk to people here because a) everyone I know is either from a feminist book club or my job at a company that exists to support progressive campaigns and are unlikely to have voted from Trump in the first place and b) the area where I live went 9 to 1 for Hillary Clinton, so even sitting somewhere with a sign that says “let’s talk about the President elect” probably wouldn’t yield the results I’d hoped for. Further, when I was working on a campaign this summer, I encountered many people that did not want to discuss politics at all because they did not want to have to defend their choices. I get that, and especially because I was a vocal Clinton supporter, I understand why someone would not want to reveal their opposing opinion to me and then feel like they have to fight me about it. But I still wanted to hear from that sort of person, because their thoughts are important too. So I made an anonymous survey, and posted it on Facebook.
Why this survey?
The questions I asked are the questions I wanted an answer to, basically. How do people who voted for Trump feel about some of the more offensive things he has said on the campaign trail? To what degree do they support the policies that have people across the country afraid for their loved ones? Do they even need to be afraid? You may have heard the phrase that the media and Democrats took Trump literally but not seriously, whereas his supporters took him seriously but not literally. Is that true? Is this a matter of mere difference of opinion that has been amped up by poor wording and selective bias in the media, or is there something more sinister afoot?
I asked why people voted for Trump in many ways. Was it partisan support? Did he perfectly capture their beliefs? I also asked to what degree people support some of the things he mentioned as proposals along the way. Do they really want to build a wall? I also asked about general beliefs, with questions on race, class, ability, gender, sex, and more. How do they see the world, and the people in it? Are we really that different? Finally, I asked questions about Trump himself. Was he ever alienating? Were they ever offended by something he said or did? Why, or why not?
I cannot emphasize this enough: THIS WAS NOT SCIENTIFICALLY DONE. No one reviewed these questions before I posted the survey, there is not an ethics board overseeing my work, nothing. I tried to uphold the principles of research that I have worked with before, mainly that the subjects are above the age of 18 (they must be by definition if they voted), they were not coerced into participation and freely gave consent in participation, they are able to remain anonymous, there should have been minimal to no risk in their participation, and they were free to end participation at any time. But, as one participant pointed out, the questions were not written or reviewed to prevent skew in collection, and were asked from a place of bias. I have tried and will continue to try to be as objective as possible, but I also freely disclose that I voted for Hillary Clinton and wrote this from a Democrat’s position.
About the participants
Participants were recruited through Facebook. I posted a status on November 10th, 2016 containing a link to the survey, and again posted a status on November 15th, 2016, and specifically asked that people who voted for Trump to complete it (rather than Trump supporters, as I believe that you can vote for someone and not consider yourself to be a supporter). I had a couple friends share the link as well. I kept the survey open until December 4th, as I had a couple family members mention to me that they shared the link with a friend recently, but all responses were collected between November 10th and November 17th. I will talk about how the recruitment process affected the survey design and information collected a bit later.
So let’s talk demographics. There were 23 respondents to this survey (Note: one person may have double-submitted their answers, but they were timestamped several minutes apart and different slightly, so I am leaving them in as two respondents). 13 respondents were female, and 10 were male. There were no transgender or nonbinary respondents. 100% of respondents self-identified as white, and no other race or ethnicity. 22 of the respondents identified as straight, while 1 identified as bisexual.
In terms of religious or spiritual identity, respondents were mostly Catholic or Christian. I left this field as a fill-in-the-blank, so responses were not uniform. 6 respondents identify as Catholic, and 5 others identified as being either Christian, formerly Christian, non-practicing Christian, or born again. 6 indicated they either were atheist, agnostic, or did not have a spiritual or religious identity. One person identified as conflicted. The remaining 6 people either left the question blank or wrote they chose not to answer.
I chose to collect socioeconomic identity rather than having participants select their annual income. Most of the respondents identified as middle class/middle income, with 17 respondents choosing that option. 4 people identified as being upper income/upper class, and 2 identified as being low income/working class.
All of the participants have obtained a high school diploma. 7 have some amount of college education without a degree, 2 hold an Associate’s degree, 6 hold a Bachelor’s degree, and 5 hold a Master’s degree or higher.
There was also the option to self-identify with any group I had not provided the option to identify with previously in the survey, specifically naming veterans, disabled, and immigrants, and including an “other” field where people could write in an identity they hold. 3 veterans and an active service member participated in the survey and identified themselves here, as well as the spouse of a veteran. One person identified as disabled. 2 people identified as immigrants. One person identified as “American born and bred”.
So what does all of this demographic information this tell us? Well, not much. This is a function of a small sample size, the design of the survey, and the recruitment process. I knew that I would recruit for participation in this survey via Facebook, which meant that all of the respondents would be people I know, or someone someone I know knows. Given that I wanted to offer anonymity to respondents, I did not want to have someone reveal themselves to me through their demographic data. As such, I did not ask for age or location of a participant, and tried to keep most demographic information vague. If this survey were to have been done on a larger, less personal scale, it would have been interesting to see what differences there were between generations, between geographical regions, between income groups based on actual numbers rather than a class identity, etc.
Further, because this sample is so small, even though I have some of this information, it wouldn’t be enough to make any generalizations about it. Is it telling that this survey had respondents that were 100% white? Maybe, but does that tell you more about the target audience (Trump voters) or who I am friends with on facebook? Does more women responding to this survey mean that more women voted for Trump than men, or that I have more female friends who voted for Trump than male friends who voted for Trump?
What conclusions can be made?
I found the data I collected absolutely fascinating. For example, among participants there was a universal distrust of government, and an overwhelming belief that people on welfare should be drug tested. Is this representative of all people who voted for President-elect Trump? Who knows, maybe. Is it fascinating? Absolutely.
Why isn’t it representative of all people who voted for President-elect Trump? As I’ve mentioned before, this survey was not scientifically done. Not only was I not approved to do this research from anyone and therefore did not have any checks or balances in the creation of the survey, the population from which I drew participants is not representative and therefore no generalizations can be made from the data. Because of this, I will be displaying only what was reported to me, and not doing much analysis.
Alright, now that I have the background information covered, let’s dig in. Remember, this is NOT representative and we cannot use this information to make assumptions about Trump voters on any scale.
- Why did you vote for Trump?
- Support for Policies
- General political attitudes
- How do you respond?
- What am I missing?
What did I learn?
Well, as I mentioned before, I am not willing to make broad generalizations because I believe it to be irresponsible with the little data that I have available. That said, I still found this very interesting. Of course, I never expected the answers to be universal. Obviously there is variation within a population of voters and people are drawn to a candidate for different reasons, so I wasn’t too surprised that there was disagreement on several topics and issues.
Keeping in mind that we can’t generalize this much, we can still explore what data we have. The central question I had (and still have) is “why Trump” but as one of the respondents pointed out, another way to ask this is “why not Clinton”.
There’s a bit more information in here to answer that question than I anticipated. For example, over half of respondents (56.5%) answered that they believe Hillary Clinton has committed crimes against the nation when asked why they voted for Donald Trump. Even more people (65.2% of respondents) answered that they wanted a political outsider, which a former Secretary of State, First Lady of the Nation, and multiple-time candidate for president is most certainly not.
Of course, this is highly partisan and structured language. If I admire Trump for anything, he certainly knows how to choose a team that is incredibly good at framing a conversation.
While less than half of respondents selected that identifying as a Republican or conservative and desiring a president to uphold these values was a reason that they voted for Trump, the majority of people named Republicans when asked who they supported in the primary (though that may be the fault of how the question was asked and where it was ordered).
This raises the question for me though: while Hillary Clinton does not meet the wants for a president as framed in the first question, would any Democratic candidate be able to? After all, President Obama was criticized as not having enough experience when he ran (of course, different political climate, etc.). But still, I completely neglected to try and find if these voters would have voted Republican regardless of who the candidate was, or if they were intrinsically motivated by Trump. My assumption here is that these groups of people are separate and motivated by different things, and would thus have different opinions on the issues, and may prioritize issues differently.
Yet another way of asking this questions is, are the respondents hard-line Republican voters, or moderates that were swayed by Trump’s rhetoric and campaign promises, or former Democrats that were turned away from the party due to Clinton being the nominee?
We can perhaps use the degree of agreement with “I was conflicted on how to vote for a long time” to help uncover this. I include this data below for ease of reference:
|I was conflicted on how to vote for a long time|
|Do not agree or disagree||4|
People who disagree with this statement, I assume, likely decided to either vote for Trump in the primaries, or decided to vote for the Republican nominee regardless of who it was and stuck by that decision, leading to little or no conflict on how to vote. I also assume people who chose they do not agree or disagree with the statement were pretty comfortable either with Trump himself or voting for whomever the Republican nominee was. That leaves us with 9 “unknowns” of the 23 respondents- approximately 39%. Were these people conflicted as a moderate Republican, and were uncomfortable with Trump’s rhetoric? Or were they people conflicted as moderate Democrats, and were uncomfortable with Clinton as the nominee?
We can turn to some of the answers to the survey question, “If you have been upset or offended by anything Trump has said through his campaign, how do you reconcile your beliefs with your vote choice?” to help to answer this. For example, the two people who wrote the following likely never were to have voted for Clinton in the first place based on issues:
- His strengths align with my major concerns, and although I may not agree with everything he says I agree more with him on important issues that I believe are important
- Of course. Both candidates have character flaws. I voted based on issues
This person on the other hand, may have been persuaded to the Democratic side had Clinton’s messaging not been “look how bad this guy is” and was instead “look what I can do for you!” OR if someone else had been the nominee (Would they have voted for Clinton? Probably not, looks like they wanted an outsider, but may have gone for Bernie. Hypotheticals, yes, but it’s still something to think about).
- I believe he has attacked Clinton harshly, but I believe the media has blown up the election. Do I think Trump is perfect, gosh no. But what candidate stands for EVERYTHING you believe in? No one. Many say we had to pick between the lesser of two evils, but I felt it incumbent I made my decision on what would affect me in the long run and what can help my family and friends. We need a change, not another pretty politician to promise something and get nothing from it. The past 8 years my family has benefited nothing, but rather watch other Americans benefit from programs. We want a change.
All in all, I can’t answer why these people chose to vote for Trump more than they have already. They have answered a ton of my questions, and I am grateful to all of the respondents for taking their time to help satiate my curiosity. I wish I had thought to collect political identity and a few other tidbits, and gotten many more responses, but I guess this is a pretty good start at understanding what we missed with election forecasts, and what we can expect from our fellow Americans in the coming four years.
Happy holidays one and all!