On Saturday mornings in the summer, I head for the North Asheville Farmers' Market. The diversity and quality of produce at the market is impressive, but good veggies don’t come cheap. Tomatoes are running $4 a pound this year as compared to $1.48 a pound for chicken breasts at my local supermarket and $2.28 a pound for pork loin. But last week, despite the cost, I did load up on the tomatoes.
That evening while my wife and I were enjoying a terrific tomato nicoise salad (here’s the recipe), I asked Mary Jean whether she thought vegetarians tended to be better or worse off financially than the average American. “Are you kidding?” she said. “Most of the vegetarians I know seem pretty well off. I think they make more money than the typical American.” I told her she was wrong, that I had just read on the website Faunalytics about a national survey of American food habits. It found that vegetarianism and veganism were more common among lower income groups than among people making more money. She said she didn’t believe it.
Faunalytics (formerly called the Humane Research Council) is one of my go-to sources for information on animal issues. The blog post I read described a recent survey by the Vegetarian Resource Group. Over the last 20 years, the VRG has regularly commissioned large scale surveys of American diets (here). The 2015 VRG survey was conducted by the widely respected Harris Poll and was based on the responses of 2,017 American adults. (Details of the 2015 poll and the results are here.)
Current Trends in Vegetarian/Veganism
The 2015 VRG poll found that 3.4% of Americans are presently vegetarians; this incudes respondents who indicated that they never eat red meat, poultry, or fish. About 15% of the vegetarians were also vegans who, in addition to avoiding meat, did not consume dairy products. As shown in this graph, the percent of people who eat no meat at all is down slightly from the 2011 and 2012 polls, but I suspect this is due to random fluctuations in sampling rather than a meaningful trend.
I was not surprised that fewer than 1 in 20 Americans is vegetarian/vegan. Other national polls have obtained similar results. I was, however, very surprised by one of the Harris Poll findings: people with below average household incomes were much more likely to be vegetarians than people in higher income brackets. The VRG poll found that 7% of people in households with under $50,000 total income were vegetarians. (The median household income in the US is about $54,000.) By comparison, only 2% of people in the $50,000 to $75,000 bracket were vegetarian or vegan, 1% in the $75,000 to $100,000 bracket, and 2% in the over $100,000 household income bracket.
The Links between Vegetarianism and Income?
I confess that my intuition was the same as my wife’s. I thought vegetarianism would be more prevalent among the affluent. How can we explain the fact that the VRG poll found just the opposite? A couple of my friends had innovative explanations. For example, my colleague Mickey Randolph (a long-time vegetarian) lamented the exorbitant prices she pays for organic non-GMO veggies, fruits and nuts. Half-joking, she told me vegetarianism makes people poor. But I don’t buy Mickey’s theory. Here are a couple of more plausible explanations for the income disparity between vegetarians and meat-eaters.
Age, Marital Status, and Lifestyle. Demographic and lifestyle differences may account for the income differences between meat eaters and vegetarians. For example, younger people in the survey were more likely to be vegetarians/vegans than the older adults. For example, 6% of survey respondents between 18 and 34 were vegetarians compared to only 2% who were over 55. Young people are also more likely to make less money than older adults as more of them are students or are starting their careers. In addition, young people are less likely to be married. Being single is associated with a lower household income, and this Gallup poll found that single people were nearly twice as likely to be vegetarian as people who are married. It is also possible that vegetarians and vegans make less money because they are more inclined to have alternative lifestyles.
Race. Another factor that could explain the income differences between vegetarians/vegans in household income is race. In the VRG poll, 7% of the Black respondents indicated that they did not eat any meat, as did 3% of Hispanic respondents. The VRG report does not mention the percentage of White respondents who were vegetarians. However, based on the overall rate of vegetarianism among the 2,000 respondents, it is likely that about 3% of White respondents did not eat meat. Note that another internet survey also reported higher rates of vegetarianism among minority groups. In the United States, the disparity between Black and White households is stunning. In 2013, the median income for White, Non-Hispanic households was $58,270 compared to $34,598 for Black households (here).
So, because of demographic and lifestyle differences it is certainly possible that vegetarians do tend to be less well off. But there is another factor that might explain the VRG study results even better – dumb luck.
Sampling “Error” and Public Opinion Polls
It is impossible to assess the attitudes and behaviors of Americans by interviewing every adult in the country. Hence pollsters use samples which are (in theory) randomly drawn and then statistically corrected to closely approximate the demographic profile of the entire population of American adults. The political polls that proliferate in the US during presidential election cycles are usually based on samples of about 1,000 respondents, and they have a margin of error of plus or minus 4%. This means that if 36% of respondents in a poll say they support a presidential candidate, there is a 95% chance that between 32% and 40% of people in the country as a whole actually favor the candidate. The differences between the true level of support in the population and the value obtained when you ask 1,000 respondents is called sampling error. Note that this is not really an error, just an inevitable consequence of using a sample to represent a much larger population.
Could sampling error be responsible for the Harris Poll finding that vegetarians and vegans come from poorer households than meat eaters? Yes. The reason is that when it comes to subgroups of a sample, the accuracy of surveys decline as the groups get smaller. The size of the total sample in the Harris Poll (2,016) was substantially larger than the typical national survey. But only 3.4% of the people questioned were vegetarians. As a result, if I did the math right, there were only 69 vegetarians and vegans in the sample.
The Vegetarian Resources Group’s polls represent some of the best data on general trends in the percentage of Americans that are forsaking meat. And, it may well be that, as a group, vegetarians make less money than meat-eaters associated with differences in age, race, marital status, and life style. But it is equally likely that the disparity in household income found by the pollsters were the result of the random throw of the sample selection dice.
After all, can the diets of 69 people really represent the 8 million vegetarians and vegans in the United States?
* * * * *
Other Animals and Us posts on vegarianism and veganism.
Hal Herzog is professor of psychology at Western Carolina University and the author of Some We Love, Some We Hate, Some We Eat: Why It's So Hard To Think Straight About Animals.