Consumer behaviour is influenced by personal politics

Blog post by Prof. Stahl (Chair of Quantitative Marketing & Consumer Analytics).

A quick Google search will reveal no shortage of newspaper articles claiming the US is a divided country, or prevaricating on the so-called “culture wars”. Such articles suggest an invisible chasm is widening in the heart of the country, the hearts of its people. Indeed, there is some evidence to suggest growing political polarization in the US population since the 2016 presidential election, and that this trend has been stronger among liberals than conservatives.

With colleagues from ESADE and Columbia Business School, I analysed data on the political leanings of all Twitter users who followed one of 307 major brand accounts between February 2016 and December 2018, to see if this growing polarization was reflected in consumer behaviour.

We also harvested data recorded over the same time period from the YouGov BrandIndex – a survey in which respondents state their political affiliation, brand preferences, and purchasing intentions over time – and Nielsen Retail Scanner Data – a record of weekly sales for 27,043 brands.

From our findings, it is clear that a link exists between people’s political identities and their consumer behaviour. In the wake of the 2016 election results, a large number of people changed their brand preferences and shopping behaviour, preferring to give their money to brands which they perceived shared their political views.

Both liberals and conservatives engaged in this polarization but it was more pronounced among liberals. This lines up with news reports and research that shows a rise in liberal activism since Donald Trump took office. It’s also consistent with the notion of compensatory consumption.

What do we mean by this?

Previous research shows that consumers use purchasing decisions to fight against perceived threats to their identity. These threats could be individual (a direct threat to one’s person), interpersonal (feeling less powerful than others), or group-level (such as a threat to your religious beliefs). In each case, a coping strategy is to buy and display products that signal commitment to the threatened identity. Essentially, it is putting your money behind your beliefs, reaffirming your loyalty to them in doing so. This is what is meant by compensatory consumption.

In the case of the 2016 US presidential election, the Republican candidate won. Furthermore, a strong characteristic of Trump’s election campaign, and later his time in office, was his vocal and often controversial presence on social media, especially Twitter. This created a situation in which liberals might have perceived a threat to their political identity, so it’s unsurprising they might have chosen to reinforce it through their consumer behaviour more than conservatives.

This polarization can be observed in the selection of brands (called a “brand basket”) Twitter users choose to follow on social media. After the 2016 election, the basket of brands followed by Democrat supporters leaned strongly in favour of firms which they saw as aligned with their political views. Similarly, during the Obama administration, both Democrat and Republican supporters followed more Republican brands, indicating a greater polarization among conservatives than liberals.

From the perspective of high-profile brands, many have witnessed their consumer bases become more politically homogenous. Brands with a mostly liberal following have seen this majority increase; in this same way, conservative firms have become more conservative.

This is especially interesting with regard to media outlets. Publications such as the New Yorker and the Atlantic had increasingly Democratic followings after the election, while conservative platforms like Fox News and Fox Business either became more Republican in their support base, or stayed the same. For example, we found that the percentage of the Atlantic’s readership that are Democrat supporters rose by around 4 percent between February 2016 and December 2018. In that same time, Fox Business’s audience became roughly 5 percent less Democratic.

This suggests that people, particularly liberals, increasingly preferred consuming media from outlets that held similar political stances to their own. Of course, there are implications from this that we did not delve into, such as the creation of “echo chambers”.

Naturally, the reasons why US consumers have become increasingly polarized in their purchasing choices are many, and this trend cannot just be explained by the theory of compensatory consumption. It’s also likely that when brands take a political stance, this greatly affects who chooses to buy their products or use their services.

For example, in late 2018 Nike embarked on a media campaign with NFL star Colin Kaepernick, who made headlines for kneeling during the national anthem before games in 2016. We found that this coincided with a substantial shift in Nike’s consumer base, from being around 51 percent Democratic to 62 percent. So, we must consider that other factors besides the election might have contributed to the increasing polarization since 2016.

The media points to an American society that is riddled with stereotypes on both sides of the political aisle. A casual onlooker might observe consumption choices occasionally factor into these stereotypes. Republicans style Democrats as godless, Atlantic-reading, latte-sipping Nordstrom-lovers. On the other hand, Democrats conceptualise the average Republican as rifle-toting, Budweiser-guzzling, fanatical Fox News viewers.

While it is interesting to note that some of these stereotypes contain a grain of truth, in that the Atlantic has grown more Democrat while Fox News has become more Republican, it is a matter for future research to analyse to what, if any, extent polarization in shopping behaviour influences the creation of stereotypes.

For now, we believe our research will be useful in helping firms understand how getting involved in politics, through issuing statements, corporate activism, or endorsing political candidates, will affect their consumer bases in the short-term future.

Author: Professor Dr. Florian Stahl

Florian Stahl is a Professor of Quantiative Marketing & Consumer Analytics at the University of Mannheim, Business School.