> the goal of elections isn't to get bulletproof results by mathematical standards. The goal is to create agreeable consent among those who voted. A good election system is one where even sworn enemies can begrudgingly agree on the result.
First you must explain to them why the former is not an example of the latter.
Just imagine you have to explain a child in kindergarden how the collective choice is made. Raising hands works. Putting different pieces of paper into a jars works. Magical machine says the result was X does not work unless they trust it, regardless of how correct the magical machine was under the hood, because the majority lacks the skill of intuitively understanding this themselves. Sure, they could trust an expert or an figure of authority, but that is a fleeting thing. A fleeting thing that may be enough for inconsequencial decisions, but not enough to steer countries.
Even I as someone who would have the skillset to understand why it has to be correct would have an easier time verifying a paper ballot process than ensuring that network connected complexity behemoth was running the program I checked for weeks correctly in any moment during an election. And even if you had a way to guarantee that, who tells me this was the case in the whole country or thst evidence wasn't faked a millisecond before I checked?
Meanwhile with paper and poll watchers from each party it is very easy to find actual irregularities and potential tampering — trust is a gradual thing with paper while it is much more binary with digital. If there is a sign for the digital machine being untrustworthy you can throw the whole result into the bin.
That's easy to explain. We live in a world where A&W's 1/3rd pound burger failed to compete with the McDonald's 1/4th pound burger because 4 is bigger than 3 so people thought the McDondald's product was bigger. There's zero hope that this public will understand fancy encryption.
GP already said.
> eVoting cannot be understood and audited by normal citizens, not even by nerdy ones.
I suggest you explain the verifiability of evoting systems to your grandma or your friend with an art degree. Then ask them to explain the same to their peer while you just listen. Then repeat the exercise with paper voting. You will see the difference.