The phrase “most people” is horribly abused, especially in arguments. Usually it’s a sort of implicit social proof—an argument from mass appeal. Just because most people believe something doesn’t make that thing true, however.
If we take “most people” to mean “the majority of individuals alive anywhere in the world today”, then most people are male, either Christian or Muslim, live somewhere in Asia, don’t speak one of the three most popular languages (Mandarin Chinese, Spanish, English), and are either under 14 or between 25 and 54 years old.1
There are all kinds of ways to mess with this data, often by segmenting it badly. For example, most people are older than 25, despite the fact that most people are either 25-54 or under 14. It’s also strange that although most people are from Asia, most people don’t speak the most popular Asian languages. Although most people are either Christian or Muslim, those groups have some pretty significant disagreements and typically shouldn’t be seen as a unified bloc.
This is all very different from arguing about the median person, or the average person. A common debate tactic is to try to construct some “average” individual and see how your arguments apply to their specific situation. In the real world, unfortunately, although everyone is average on average, nobody is fully average. “Most people” gives us a different kind of person to construct—based on the above data this person is both Muslim and Christian, either speaks zero or hundreds of languages, etc. This is not a realistic case on which to base an argument.
There’s also just outright misapplication. In casual arguments with friends, how often does “most people” mean “most people who are very much like me”? In typical conversations I’m often guilty of this. I catch myself saying things like “most people enjoy Python’s readability” without considering that globally most people have never even seen a single line of Python code. I’m using the phrase differently from its most literal meaning, but without explicit restrictions on the domain of who my statements apply to, I’m opening my arguments up to misunderstandings.
Debaters often use “most people” when they should instead use “many people”. We can know that many people believe something without being sure whether the number is over 50%. In these cases, “most people” is just acting as a sign that you’re with the majority, and your opponents should feel bad because the social weight of the majority’s approval is against them. There are cases where what “most people” do is trivially obvious to both sides, but what we care about is some tiny sliver of the population and the phrase is a shortcut to changing the underlying demographics.
It’s tough to restrict your arguments to just phrases that actually make total sense, unreliant on a web of implicit background assumptions. In many cases, it’s legitimately impossible to ensure you’ve conveyed your ideas fully with every nuance captured exactly as you understand it and transplanted into your fellow conversationalists’ brains. Here’s yet another argument for the principle of charity in argumentation: if we struggle to make ourselves clear, how much of what people say to us is being misconstrued by our own preconceptions?
Most people could improve their arguments by not using “most people”.