Either way, even if it's true that some significant percentage of voters have below-average IQ, how strong is the evidence that they'll make a "worse" collective choice of government than those with above-average IQs? That's basically assuming that the properties of candidates that will ensure they can act as part of an effective government are somehow "too hard" to accurately judge unless you have above average IQ. It seems just as possible the opposite is true - i.e. those with above average IQ aren't good at judging the sorts of properties candidates need to be govern effectively - at least partly because that implies being able to govern in way that works well for those in the below-average-IQ group (i.e. those in the above-average-IQ group will probably do reasonably well for themselves regardless of the quality of government).
Google provides many hits of people using average to mean within one standard deviation, or 85-115, but in a few minutes of searching I couldn't figure out a proper authority for that.
The average person has an IQ of 100. Given that average IQ is defined as being one standard deviation of average, the vast majority of voters are between low intelligence and average intelligence. About 84% of voters thus have low-to-average intelligence (68% fits in one standard deviation of average; 68/2 + 50 = 84%).
Regardless of your political leaning, I think it’s worthwhile accepting that most governments are elected by people who aren’t as smart as the typical HN contributor.