My research: Nationalistic Nostalgia predicts support for Trump & Brexit
Nationalistic nostalgia has been demonstrated to predict support of Trump and Brexit better than demographic factors, such as gender, age, education, ethnicity, and even political party. In order to better understand and react to the results of the EU referendum in the UK and Donald Trump’s election in the US, this summary will illustrate the concept of nationalistic nostalgia, describe how it functions and influences our voting decisions, summarize the results of this research, and provide recommendations for Democratic and left-leaning messaging.
What is this research about?
Nationalistic nostalgia is the general sentiment that the country used to be better off in the past. This sentiment has a long history in the US and the UK, but the election cycles of 2016 have put it front and center. Two well-known slogans articulate this attitude and belief perfectly: Donald Trump’s slogan “Make America Great Again” and the slogan of the campaign for the UK to leave the EU “Take Back Control”.
I set out to understand what impact nationalistic nostalgia could have on voters’ electoral decisions and perhaps more importantly, where it comes from.
In the academic world and specifically psychology, we look to previous research on either the same or related topics in order to guide our research, and so I looked to general nostalgia, conspiratorial thinking (belief in conspiracy theories), and religious thinking as examples of related types of thinking or cognition. All of these forms of cognition appear to respond to threats. When something is threatened – our livelihood, our identity, our family or group – we are compelled to act. Without a threat, we’re not going to react.
It’s also worth pointing out that when looking at “threats”, we’re talking about someone’s perception of a threat. Whether or not there is a “true” threat doesn’t matter. If we feel or perceive a threat to some key things – our life, our identity, our family or group – again, we are compelled to act.
So this research investigates how the perception of various threats can increase nationalistic nostalgia in someone’s thinking, and then, evaluates whether or not that attitude impacted their vote in the 2016 US Presidential election and the UK’s referendum on EU membership.
Why does it matter?
We currently have major election results in two of the most powerful Western democracies that were not predicted by the majority of quantitative research (polling). Understanding the underlying psychology of the US and UK electorates has national implications in both countries as well as international implications. To understand why and how these results transpired, I argue that we need to consider the cognitive processes that contribute to our electoral decision-making.
If you’re trying to understand the mindset (or the cognitive processes) that contributed to voting to leave the EU or voting for Donald Trump, then this research is for you.
What went into this research?
I conducted three studies: 1 experiment in the UK and 2 surveys in the UK and US respectively.
For the experiment, I went out into 4 small towns outside of London in the southwest of England and asked people on the street to fill out a 15 minute survey.
The two surveys were conducted during the final days leading up to the EU referendum and the US presidential election,[1] and these surveys collected responses online from voters in the respective countries. Each survey had about 120 respondents, which is a large enough sample to observe statistically significant correlations and associations, but not large enough to generalize specific proportional statistics to the entire UK or US.[2]
What does it tell us?
Generally, the results of these studies revealed a lot about how nationalistic nostalgia functions, and it confirmed that it does indeed play a role in how we make voting decisions. In fact, it confirmed that nationalistic nostalgia was a better predictor of voting to leave the EU or voting for Trump than almost all of the demographic factors that are often credited with explaining why someone votes the way they do, such as gender, age, education, ethnicity, and even political party.[3]
The experiment told me that hearing or reading nationalistic nostalgic slogans did not generally impact someone’s thinking or how they were going to vote (in the EU referendum). Not too shocking, but the surveys provided much more insight.
What surprised me were the causes behind nationalistic nostalgia. The two surveys in the US and UK investigated five threats that could potentially increase nationalistic nostalgia in someone’s thinking.
Threats to:
- Personal safety and security
- Personal finances
- National safety and security
- National economy
- National ethnic identity
In the British data, the only threat that correlated significantly with nationalistic nostalgia was the threat to national ethnic identity, meaning as someone’s perception of a threat to that identity increases, their nationalistic nostalgic thinking also increases as a result.[4]
In the American data, there were three threats that correlated significantly with nationalistic nostalgia: threats to national safety and security, national economy, and national ethnic identity.
One way to understand these threats is to group them into “realistic” and “symbolic” threats.[5] Realistic threats are those concrete things that threaten your livelihood, i.e. your income, your health, etc. Symbolic threats, on the other hand, refer to your (or rather your group’s) worldview or way of life. So it’s these symbolic threats to national security, the economy, and national ethnic identity that produce nationalistic nostalgia, and then in turn increase the likelihood of voting to leave the EU or voting for Trump.
What do we do with these results? (Or so what?)
So this study offers insight and texture to the current political landscape in the US and UK, but it doesn’t test messages to see how they impact someone’s thinking or voting decisions. That’s a major distinction to be made. In terms of recommendations, I’m also taking the perspective of a progressive and/or Democratic operative. Again, an important qualification to be made.
My recommendations pertain to what threats or anxieties are shaping voters’ decisions in the current political environment. Put another way, these are recommendations on who progressives could hope to win over and how.
If the aim of political messaging is to connect with and alleviate someone’s perceived threats and resulting anxiety, then we have to take into consideration what factors are most important in their electoral decision-making. This research says that symbolic threats are the key, and for the US data, specifically the threats to national security, the economy, and one’s sense of national ethnic identity. If someone believes that their national ethnic identity is seriously under threat, then I argue that Democrats and progressives can’t speak to and alleviate that anxiety without disavowing core values of equality and inclusion. So let’s not do that. Perhaps this is a moment in political history when we need to:
· stand strong on issues of racial and economic inequality and inclusive policies,
· let go of the votes we will lose,
· and focus on other anxieties and pressure points that impact voters’ attitudes which also resonate with our core values as progressives and Democrats.
So what can we do? We can focus on the other symbolic threats, while also digging in our heels on standing for equality and inclusion. If that means we lose some people, so be it, but if we attempt to alleviate someone’s anxiety from threats to their national ethnic identify in the name of winning votes then we still probably won’t win them over because Republican messaging accomplishes this much more effectively (and frankly, we wouldn’t recognize ourselves and our party afterwards). Democrats have already dedicated considerable effort to articulating a message and implementing policies on the national economy, and to be fair, they’ve been successful on improving the economy over the last eight years. And so I recommend that in the up-coming months, Democrats focus on messages and policies to address national security. This is an area that plays a considerable role in American voters’ electoral decision-making. Messages and policies on this issue area can easily correspond with our core progressive and Democratic values.
Notes on understanding this research in the context of elections and social science
Can I say that my friend/acquaintance found the one thing that explains Brexit and Trump? Sure, that would be cool, but not necessarily accurate. The thing with social science is that it takes a bunch of factors into account and tries to see which factors are most closely (or significantly) associated with the occurrence of another factor or an event. Of course, each study can’t take into account every single factor that contributes to real life. I would say one of the merits of this research is that numerous demographic factors[6] are analysed along with a two cognitive factors,[7] and so it’s combing a lot of things that impact how someone votes. And yet, my model didn’t include the weather in the location of where each survey participant was located when they took my survey – or on the day of the election for that matter. It didn’t take their up-bringing or child-rearing attitudes into account, and it doesn’t include measures of social trust in institutions.[8]
Here’s how I would describe and contextualize this research: It very clearly identifies a factor that lead to both of these surprising elections. Is it the only factor? No, of course not, but it is one factor that must be considered. And that’s the challenge of social scientists, political prognosticators, and informed citizens: we need to synthesize the insane wealth of research and analysis that’s being produced by the dumpster fire of 2016. Each factor that’s highlighted by this study here or that research there needs to be viewed as one lense that will help bring a very fuzzy picture into focus. Honestly, it’s not yet realistic to imagine that we’ll create some super model that combines all of the factors that have been shown to significantly diagnosis the political landscape of the US or UK (maybe someday soon with the help of some nifty artificial intelligence and Bayesian models), but at this point we need to use our organic intelligence to combine all of the scientifically significant results, including this study, and not the unsubstantiated, untested punditry and hot air from opinion mongers. Just sayin’.
A note on the sample size for my polling aficionado friends: When you’re looking at the correlations between various factors, you need a sample size that sufficiently allows for a p-value below 0.05 (according to general academic social science standards). A randomly collected sample above 30 participants can sufficiently confirm or nullify the hypothesis of a correlation between factors at a statistically significant level (if you ascribe to Bayesian statistical theory, and I do). If you’re interested in seeing whether these correlations hold at the level of a representative sample (and have the resources to fund a survey of that size…) let’s talk!
Want more data and information? Of course, you do! And I’d be happy to provide it. This document aimed to articulate the top-line results of this research, but I’ve also prepared some data tables and statistical analysis that underlie these findings. You might also be curious about my definitions of certain terms and how I constructed the measures of these attitudes and constructs in my survey. If you want to hear more about any of this, just get in touch, and I’ll be happy to walk you through it.
Endnotes
[1] The UK survey was conducted on June 18, 2016 before the EU referendum on June 23, 2016. The US survey was conducted on November 4, 2016 before the US Presidential election on November 8, 2016.
[2] If that sample size makes you raise your eyebrows, read the note on the sample size for my polling aficionado friends at the end of this document, but this sample size is actually a bit large than similar psychological studies that have been published in academic journals. This sample size meets the standards of academically rigorous research.
[3] The main exception in the US data is income, but it doesn’t go in the direction that you think. Technical note: This data shows that when controlling for all of the variables put into my model, the only variable with a larger standardized coefficient than nationalistic nostalgia is income. It shows that as income increases so does your likelihood of voting for Trump, again controlling for all of the other variables.
[4] Yes, I’m inferring correlation as causation, because of previous literature that links up threats and cognitive responses in this way.
[5] Which Walter Stephan has proposed and a lot of people agree with. His theory is called Integrated Threat Theory.
[6] Such as age, gender, political party, ethnicity, income, education, and religion
[7] Nationalistic nostalgia and conspiratorial thinking
[8] All of which other research has demonstrated has considerable correlations with political attitudes and voting patterns.