Some people like to eat octopus. Liberals, mostly.
Russell Baker
Check out how your food preferences dictate your ideology (or is it the other way around?) in this interesting piece of pedagogy:
How Food Preferences Vary by Political Ideology
November 10, 2009
Kelly Ford
Content lead, Hunch Inc.
Abstract
- This report examines the differences in selected food-related preferences and choices made by self-described conservatives and liberals.
- The report draws on aggregated data collected between April 2009 and November 2009 from Hunch, a website which aids in decision making.
Key Findings
- While there is significant common ground in the food choices made between the two groups, there are also important differences. Liberals show a consistent tendency to enjoy more international and exotic cuisines, with conservatives often leaning more towards mainstream, comfort food staples.
- Significant differences also surface between the two groups in consumption preferences for meat, vegetables, fruit, and “healthy alternatives”, with conservatives generally choosing the less healthy options.
Introduction
Hunch is a website which uses multiple-choice questions to help people make decisions, whether it’s what kind of camera to buy, where to go on vacation, or more than 5,000 other topics.
To arrive at a customized suggestion for each user, Hunch asks both “topic-specific” questions (Are you an experienced photographer?, Will you be traveling with kids?) and also more general/personality-based questions (Do you like snow cones?, Are you superstitious?, What types of books do you like to read?) which are called Teach Hunch About You (THAY) questions. There are more than 1,500 THAY questions in Hunch’s question pool, and more than 25 million of them have been answered since Hunch’s launch. THAYs help Hunch get to know each user so that decision outcomes are customized to each individual.
As a by-product of all the questions being answered on Hunch to make decisions, the response data is available for analysis in an aggregated and anonymous form. Hunch frequently blogs about interesting correlations using this data.
Methodology and baseline
The data in this report relies on both “THAY to THAY” correlations (which might entail exploring how political views correspond with age, for example), and also “THAY to RESULT” correlations (which might entail examining how preference for brie cheese corresponds with someone’s age, for example).
More than 64,000 Hunch users have answer the THAY question “Do you tend to support liberal or conservative politicians?“ Hunch users responded as follows:
15% answered “conservative”
54% answered “liberal”
31% answered “middle of the road”
All secondary questions were examined within the context of this initial, baseline question, with the focus on difference between conservatives and liberals. Note that users responding “Middle of the Road” tended to have food preferences squarely in between the preferences of the self-labeled liberals and conservatives. “Middle of the Road” responses are thus not discussed in this report’s primary data tables, but those responses are always included in supporting graphs.
How Hunch works
Findings
Category |
Conservatives trend towards/are strongly correlated with… |
Liberals trend towards/are strongly correlated with… |
---|---|---|
54% more likely to choose very modern style kitchen |
||
|
|
|
Chain restaurants (1) |
|
Pret a Manger, Cosi, In-N-Out Burger |
Cheese varieties (1) |
||
Bacon double cheeseburger: delicious or disgusting? |
||
Cocktails (1) |
||
Pizza crust: thick, thin, or normal? |
30% more likely to prefer thick/deep dish |
|
|
|
|
|
|
|
|
More than twice as likely to prefer arugula |
|
26% more likely to prefer “bigger portion, plainly arranged” |
40% more likely to prefer “smaller portion, artfully presented and garnished” |
|
|
- Items designated with this footnote are based on a lower sample size. They involve comparing the THAY question about political stance (“are you liberal, conservative, or middle of the road?”) with the individual ‘like or dislike’ responses which users can provide for Hunch decision results within one of Hunch’s 5,000+ decision topics. For example, a Hunch user who tries the “Cheese varieties” topic would answer a few qualifying questions and then be presented with a cheese variety, together with a “Do you like this?” question. The stats above involve correlating a ‘yes’ answer for each type of cheese the user liked with the user’s answer to the political THAY question.
Can’t we find some common ground here for the family dinner table?
Despite there being many dimensions of food for which conservatives and liberals skew in different directions, there is also plenty of common ground. The following are some of the cases where conservatives and liberals shared the same #1 answer (even if there is still some skewing towards one group or the other in many cases):
- Romaine is everyone’s favorite type of lettuce. (top choice of 42% of conservatives/ 37% of liberals)
- Both groups prefer salt on their margarita glass. (44% of conservatives, 49% of liberals)
- There’s practically no difference in how both groups describe their cooking skill. Most people (41%) say “There are a few things I can cook well”, the 3rd lowest skill set listed on a 4-point scale.
- Forced to choose “delicious or disgusting”, most people say the bacon double cheeseburger is indeed delicious (66% of conservatives, 54% of liberals)
- Hold the sprinkles on our ice cream, please (say 73% of conservatives and 68% of liberals).
- Both groups tip about the same in restaurants. 46% of both groups consider themselves “average tippers” and about 44% consider themselves “generous tippers”.
- More than 70% of both groups prefer their sandwich cut diagonally rather than vertically. Still, conservatives are 18% more likely than liberals to prefer the vertical approach.
- About 68% of both groups prefers “traditional corks” in their wine. But liberals are still more likely to prefer plastic corks or screw-on caps.
- About 61% of both groups say they care for their kitchen tools by “just washing them well”, although 13% of conservatives (vs. 10% of liberals) say they “care for them like a child”.
- Who doesn’t like a good hot dog? 74% of liberals and 83% of conservatives enjoy them.
- About 43% of both liberals and conservatives “grow something indoors or outdoors that they also eat”.
Summary and conclusions
The data in this report shows a consistent pattern for conservatives to trend towards “homey”, familiar, comfort foods and meat-heavy options. They are more likely than liberals to indulge in fast food and enjoy splurges like cheeseburgers, hot dogs, deep dish pizza and sugar soda. Their idea of international food is a “mainstream” option such as Italian.
Liberals are more likely to be adventuresome eaters, choosing international options such as Japanese or Thai. They eat fast food less frequently than conservatives, and when they do splurge on fast food they have a tendency to favor specialty, regional chains. Liberals are more likely to be vegetarians and to choose healthier options such as whole grain bread, darker greens of lettuce, and more frequent servings of fruit.
The food preferences expressed are no doubt heavily influenced by the regional tastes of areas which are relatively more conserative or liberal. For example, fried chicken is a staple of the conservative American South. “In-N-Out” burger, favored by liberals, is a popular chain in the liberal state of California.
Methodology and disclaimers
Hunch was launched in June 2009 and since then has had millions of people use the site. More than 25 million THAY (“Teach Hunch About You”) questions have been answered since that time. The motivation for people to answer Hunch questions is simply to receive better and more customized recommendations. For example, for the decision “Which magazine would I like?”, Hunch would suggest a different answer to a “suburban working mom” than to a “college student in an urban area.”
Hunch’s THAY questions are designed to be engaging, innocuous and fun, and they are interspersed throughout the site so that they can be answered at a user’s leisure. Questions are not grouped by any sort of theme; a common sequence of 3 might include: “How old are you?”, “Can you juggle?”, and “What’s your opinion of stem cell research?” As a result, the Hunch team has found that users enjoy answering the questions and do so honestly and consistently. There is little reason to answer dishonestly or attempt to “game” the system, since answers are solely designed to help the user get better recommendations. This research data is a by-product of Hunch’s core business and mission, which is to provide smart recommendations to users. (a similar analogy might be the way search engines release data on popular searches. That data is a by-product of their core mission to help people find what they are looking for online).
Summary findings in this report are noted when there is a statistically significant difference in the answers of the two subsets being compared. Links are always provided to graph data, which also includes the sample size used for the question and error ranges based on a 95% statistical confidence level.
Hunch strives to uncover insights which are interesting, provocative, and which will stimulate discussion and debate. Hunch does not endorse or favor particular ideological points of view, and the contents of Hunch reports do not necessarily reflect the opinions of Hunch as an entity or its staff or directors.
It should be noted that Hunch is not a professional research organization and this data was not collected in a perfectly controlled way. For example, Hunch data is based on 1) a group of individuals who are by definition all users of the Hunch website; these users are likely more internet-savvy than the general population, more open to new technologies, and more interested in social media. The group, while large at millions of users over the last 4 months, has not been weighted to reflect actual demographic segmentation, and is not necessarily representative of general populations. 2) Questions on Hunch are voluntarily answered by users who choose to answer them, and any question can be selectively skipped. Therefore it’s possible that those users who skipped a question might have different views than those who chose to answer a question. This said, generally only about 3-5% of users choose to skip a question when it is presented to them, and the percentage of users who skipped the question is always included in graph data.
Why are you doing this to us? I just wasted over an hour of my time there answering questions just to see what they’ll suggest for dinner! ;-)
In your case, I recommend the sausage.
This finally proves, beyond a shadow of a doubt, that liberals are snobs, but conservatives are honest.