RALEIGH — For weeks after the ballots were counted, supporters of the defeated presidential candidate insisted the election had been stolen. Some alleged a shadowy conspiracy to rig vote-counting machines, throwing out just enough legal votes and manufacturing just enough illegal ones to decide the outcome.

I heard these conspiracy theories many times, and not just from folks on the political fringe. I heard them from prominent North Carolinians who supported the defeated candidate. They were Democrats. It was 2004.

No, voting machines from the company then known as Diebold were not rigged to deny John Kerry his rightful victory. If you are one of those Democrats who repeatedly flogged that story to me back then, or who still believe it to be true today, then you should know I’m just going to tune out your outrage at Donald Trump supporters for espousing a similar conspiracy theory.

The public discourse about elections and election laws has become thoroughly suffused with hyperbole, paranoia, and misleading claims. Republicans discount Democratic allegations as fanciful and then make their own fanciful allegations. Democrats act similarly.

In reality, our election system, while exhibiting some correctable flaws, works fairly well. Generally speaking, it has never been easier to vote than it is right now. And, generally speaking, vote counts have never been harder to fake or manipulate than they are now.

The rampant voter suppression about which some Democrats complain so vociferously is simply not evident in election statistics. Neither is the rampant voter fraud about which some Republicans complain so vociferously.

Indeed, the same studies that disprove one tend to disprove the other at the same time. Consider the example of voter-ID requirements. Most scholarly research has found that requiring identification to vote has either a tiny effect or no discernible effect on the number of ballots cast. The vast majority of citizens, in other words, either possess an ID already, make easy use of state programs to get one, or have no interest in voting, anyway. Keep in mind, though, that a lack of a significant relationship between ID laws and vote totals also suggests impersonation fraud is very rare. Otherwise we’d see vote totals dip after enactment.

That’s not an argument against voter ID, by the way. It strengthens public confidence at a low cost. Nudged by the requirement, some folks without photo IDs obtain them to vote and then enjoy the ancillary benefits of having ID cards. Moreover, in North Carolina, a photo-ID requirement is constitutionally mandated.

My point is simply that at least the kind of voter fraud an ID requirement might deter does not occur at a scale large enough to tip the vast majority of elections. Nor do most other election irregularities or mistakes.

It is, however, prudent to take reasonable, low-cost precautions against the rare exception. That means requiring identification to cast in-person or absentee ballots. It means forbidding third parties from harvesting and turning in ballots — a potentially abusive practice that Democrats properly criticized in North Carolina’s 9th District race in 2018 but that would be expanded, not curtailed, by the “For the People” bill Democrats are advancing in Congress.

Plenty of Republican activists, and even a few Republican officials, indulge in conspiratorial thinking about elections, too. When I answer their calls or emails, I remind them that extraordinary claims require extraordinary evidence.

Do left-of-center columnists and political analysts do the same? I don’t see it. I see them lionizing Stacey Abrams, the former Democratic lawmaker who lost Georgia’s 2018 gubernatorial contest by a much larger margin than Trump lost the state two years later. Yet she insisted that she was cheated out of her rightful victory by a GOP-led conspiracy. As far as I know, Abrams refuses even today to concede the 2018 race.

Over-the-top claims about voter suppression and electoral conspiracies are mostly political theater, not serious analysis about the (usually modest) effects of actual policy changes on actual voting behavior. I know that. I don’t have to like it.