Flawed Method, Bad Results
Why are we using unscientific social research methods to govern ourselves?
With controlled sampling, it is possible that a group of 2,500 people will accurately reflect the opinions of a population exceeding 300,000,000. The best method to ensure a survey’s accuracy, no matter the population size, is to oversample. That is, to collect multiple sample groups from the same population. Depending on the complexity of what social researchers call the instrument (“questionnaire”), these samples are then adjusted to produce what is believed to be an accurate snapshot of public opinion.
Republicanism is founded on the principle that a sample may accurately reflect the interest of the governed such that they will continue to consent to be governed. Democracy generates consent via direct participation of the governed. How these respective forms of government generate consent is an important distinction, since they are seeking the same ends, through very different methods.
A census is a survey conducted without sampling—the constitution mandates we conduct a decennial population census. This total-count is then used to apportion Congressional representation by population-geography (the geography of the Senate is itself set by state boundaries), to which seats are assigned. The census is then used to un-scientifically determine population samples (Congressional districts) that will each be represented by one member of Congress. That member will be chosen from an unscientific, self-selected sample of candidates, every two years.
Republicans seek to run society through the careful selection of those who will govern, however this will always produce non-random samples whose opinions (and material interests) will not match the governed. Republicanism does not seek to build consent through majority appeal.
Democrats (once upon a time) believed the governed should run society, and sought the largest selection of candidates and voters. Democrats still want large voter turnout when they know it gives them an advantage over Republicans, but not when it means more votes for Greens and Socialists.
There are multiple sources for error in survey research. Researchers are responsible for error produced through survey design, execution, and analysis. Respondents also produce error, sometimes unintentionally. They may not be qualified to answer, but do so anyway; or they might offer an erroneous response due to a lack of knowledge (opinions cannot be incorrect, but the “facts” they are based on can be). Finally, when given close-ended questions, respondents may choose an answer that is not their actual preference, if they don’t see an “Other” or “None of the Above” response.
2024 Presidential Election tallies.
Eligible, Did Not Vote 87,900,651
Trump 77,302,580
Harris 75,017,613
Stein 862,049
Kennedy 756,393
Oliver 650,126
Write-In (All Names) 210,381
The straight-up calculation of the Margin of Error in the 2024 presidential election is very small (est. +/- 1.5%). This is because survey methodologies do not test for error outside of the range of questions. As for the accuracy of popular vote results, we can be assured that the outcome would reflect the opinions of eligible voters, when provided the six options listed above on Election Day, 2024. Other options not included remain unmeasured, as well as motivations for refusal to participate.
Elections are not scientific surveys. If the democratic ideal is rule by the governed, we will not get accurate representation via any election as they are run now. Ironically, we have a much better chance at ruling in the way the governed wish to be ruled through the Republican approach. However, that means first having to destroy bias in the sampling process, which no elected Republican (or Democrat) wants.
While the Republican form of governance is more practical, the means of establishing government (Forming districts, Choosing a candidate pool, Selecting a winner) is so biased as to be unreliable and scientifically invalid. Both corporate parties had agreed to focus on manufacturing consent to the process—it’s why Republicans blamed Ross Perot for Bush the Elder losing in 1992, why Al Gore conceded to the Supreme Court in 2000, and why Democrats blamed Jill Stein for #HerTurn’s loss in 2016.
Trump, of course, threw that all away in 2020, refusing himself to consent to the Presidential election outcome and now going so far as to absolve and pardon the “victims” of January 6. He does not plan to leave office this time, either. What makes anyone think he would?
We need to fix our means of selecting those who will govern, since it is being scrap-heaped anyway.
