Every May, an analytics company, Gallup, releases a poll of where Americans all across the country stand on abortion. This is part of their Values and Beliefs series, which is in turn part of their Gallup Poll Social Series. Along with ages, education levels, political affiliation, gender, and religious service attendance, region is also considered in the polling data. When I compared the region data from 2018 against that of 2019, I found something about my region, the Midwest, that initially troubled me. Upon further reflection, however, my mind began to relax.
In 2018, out of the 213 people interviewed in the Midwest, 42% described themselves as pro-choice and 55% as pro-life while 3% said they had no opinion. In 2019, 212 Midwesterners were interviewed, 50% described themselves as pro-choice and 48% as pro-life while 2% claimed to have no opinion. The difference between these two sets of numbers is obvious, the percentage of people in the Midwest who support abortion appears to have risen (+8%) while the percentage who support life has apparently dropped (-7%). At face value this is quite troubling as it could very well be indicative of a morality shift in my region. However, all polling data is to be taken with a grain of salt…sometimes perhaps even a spoonful of it.
My home state of Minnesota alone has a population of about 5.5 million people. Claiming that the survey results of just 213 Minnesotans is absolutely indicative of the entire Minnesota population is ridiculous enough. Yet, this poll isn’t just out of Minnesotans, but the entire Midwest, which has a population of about 67.7 million people. To think that this Gallup poll is definitively accurate is absurd. Not only is 213 out of 67.7 million a miniscule sample size, but it could very well be an insignificant one. More information about the interviewee would need to be known. Are they urban or rural dwellers? Are they Catholic, Methodist, Mormon, Muslim, Lutheran, etc. The different denominations within a religion, to say nothing of different religions, can show extremely different information. Methodists tend to support abortion, whereas Catholics tend to be pro-life. Therefore, if 32 of the 213 Midwesterners surveyed were Methodist and only 2 were Catholic, the results of the poll aren’t accurately representing the region. In order to even feign definitive accuracy the poll would need to include a proportionate amount of people surveyed from each possible group and subgroup. This simply wouldn’t happen because the logistics and cost would be unmanageable and the list of groups and subgroups is never ending. This is why all polls, whether they support your cause or not, should not be relied upon alone.
In my opinion, if you wish to ascertain where Americans stand on specific issues, then you need to look at how Americans are voting within their states and nationally. I know it is tempting to latch onto a poll that features results that line up with our own positions, but it is a dangerous practice as many polls simply do not include a large enough, or diverse enough, sample size. Not only that, but we typically have no way of knowing if the sample size they do use is even proportionately representative of the area’s groups and sub groups. Polls can help point us in the right direction or lend possible evidence of a trend, but they cannot be relied upon alone.