It’s inherently an argument that democracy does not work.
Clearly, voters are not casting votes based on objective measurements of the things that some candidates believe are important to them (e.g. household income, life expectancy, health care quality etc).
But that means either that they are voting based on other issues that they consider important, or they are not voting based on likely outcomes of a candidate's policy preferences at all.
It's not trivial to differentiate these two (and of course, there may even be a mixture of all 2, or even all 3, reasons to vote).
And sure, people may vote for a candidate (implicitly, for a policy) that benefits society as a whole even if it negatively impacts them. It does stretch credibility, however, to try to make the case this is what is happening when people earning median incomes or below vote for candidates who cut taxes on the wealthiest in a society, as well as reducing the share of GDP going to labor, and claiming "well, those folks just think this candidate is doing a good job on <cultural issue>". I'm not suggesting it is impossible that this happens sometimes, but across the entirety of working class Republican voters (for example) ... I find it hard to believe.
The suggestion wasn't overt, it was kind of implicity - telling people that they don't know their own self interest, even when they manifestly don't, is not very ahh "politic" :)
The GP did not call voters "selfish". It said
> People routinely vote for candidates that will worsen their lives [...]
Now, I would personally reword that as "People routinely vote for candidates despite evidence that these candidates policies will worsen one or more aspects of their lives ...".
But nowhere is there the suggestion that "you didn't vote for my preferred candidate and therefore you are selfish".