LSE - Small Logo
LSE - Small Logo

Lorenz Graf-Vlachy

March 19th, 2021

Does the partisan divide extend to online reviews?

0 comments | 29 shares

Estimated reading time: 3 minutes

Lorenz Graf-Vlachy

March 19th, 2021

Does the partisan divide extend to online reviews?

0 comments | 29 shares

Estimated reading time: 3 minutes

Liberals and conservatives differ starkly in their political convictions, and their assessment of social and economic issues. Surprisingly, they also write very different online reviews when they evaluate online purchases. Researchers analysed the web browsing behaviour of a sample of US consumers and found that liberal customers write online reviews that are longer and contain more arguments. Lorenz Graf-Vlachy explains the findings and examines the factors behind these behavioural differences.


 

In these hyper-partisan times, there seems to be little that unites us. One thing we all do, however, is rely on online reviews. We do this when shopping on Amazon, when browsing for restaurants (or delivery services) on Yelp, and when making post-pandemic plans on TripAdvisor. But what influences how these reviews are written? Does politics play a role? In our study, we show that political leanings have an impact beyond people’s voting behaviour. Specifically, we show that they actually affect how people write online reviews.

Overall, we find that more liberal customers write online reviews that are longer and contain more arguments. In addition, they provide more diverse arguments in that they highlight both positive and negative aspects of the product or service they cover. Finally, more liberal reviewers tend to use more positive language. All these review characteristics are important because they impact how helpful other customers will find the reviews. Typically, longer reviews, reviews that are more diverse, as well as negative reviews, are thought to be particularly helpful.

There are different likely explanations for each of the observed differences between the reviews of liberals and conservatives. First, a liberal political ideology is generally associated with greater altruism, i.e., a tendency to have a selfless concern for the well-being of others. This greater altruism might lead more liberal reviewers to commit more time and effort to writing reviews, which, in turn, can lead to reviews that are longer and contain more arguments.

Second, more liberal people also typically exhibit lower need for closure and greater comfort with ambiguity. That is to say that liberals typically can live with contradictions and are more at ease with conflicting information. This may explain why their reviews contain more diverse arguments, i.e., they mention both favourable and unfavourable aspects of the reviewed product or service.

Finally, liberals and conservatives differ in what researchers have labelled “negativity bias.” Generally speaking, more conservative individuals are more sensitive to negative stimuli than more liberal individuals. For example, experiments have shown that angry faces grab conservatives’ attention much more than liberals’. This likely explains why more liberal reviewers strike an overall more positive tone in their online reviews.

One particular challenge we faced in this study was that it is hard to reliably study political ideology. Much research relies on self-reports of political leanings. But such data is often problematic, especially when data collection is not anonymous, because people do not always respond truthfully. Just remember how wrong many pollsters were in recent US presidential elections.

To study the link between political ideology and online reviews, we therefore analysed the web browsing behaviour of a sample of US consumers. We measured consumers’ political ideology by examining which online news outlets they frequented. Conservative users, for example, are much more likely to visit the website of Fox News than liberal users. Liberals, in turn, are more likely to read the New York Times online. We could thus observe users’ actual web browsing behaviour and used it to infer their position on the political spectrum. We then matched this information with online reviews the same users wrote on Amazon, Yelp, and TripAdvisor, and analysed the reviews using manual and computer-aided text analysis techniques.

Of course, it would be easy to imagine a variety of factors that might explain the findings. That is why we carefully ruled out distortions by influential factors like socioeconomic status, i.e., people’s household income. Even more importantly, the effects we observed were not caused by how highly the reviewers rated the product or service they reviewed. We found the differences between reviews from liberals and conservatives to remain meaningful even after taking into account how many ‘stars’ they awarded the reviewed product or service. So, it is not just that customers’ ideology colours how much they like a given product or service. Their ideology directly influences how they write their review.

Next time you check out an online review, think about how you might now have a clue about who the writer voted for in the last election!

♣♣♣

Notes:

 

About the author

Lorenz Graf-Vlachy

Lorenz Graf-Vlachy is a professor for corporate management at TU Dortmund University and a senior research fellow at ESCP Business School

Posted In: Technology

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.