Could you rig all the political opinion polls in a country?
For one of my digressive footnotes, I have been thinking about this question recently.
It would be beyond ordinary people, but if you had all the resources of the Chinese or Russian state at your disposal? It would not be easy. But there are some things that make it less hard.
For one, behind the scenes there is a lot of shared infrastructure in the polling industry. Different polling firms use the same telephone call centre operations. Different polling firms also use the same software. Different polling data is stored in the same cloud services.
There is no one call centre, one piece of software and one cloud to corrupt. You would need to corrupt multiple things, but not nearly as many as the number of polling firms may make it seem.
Could the Chinese bury a hidden back door in a shared piece of code used in multiple polling firms that provides hackers access which is then used to pollute final pre-election polls?
Of course, you would not need to hack every poll. Just enough to make the others look like weird outliers. Weird outliers which could also be discredited by a campaign of online disinformation, such as faked leaked documents over who really owns certain polling firms.
It is certainly plausible enough to make a Hollywood movie out of the idea. It is also one of the more niche reasons why being ‘99% certain’ or similar about what the polls says is unwise. There are an awful lot of very unlikely events – including sudden illness or war – that you have to pack into that remaining 1%. And 1% is super, massively unlikely. With a general election every four years typically, saying there is a 1% chance of something is pegging it at a 1 in 400 years chance. Four. Hundred. Years.
So don’t worry too much about the hacking risk. But do worry about those who tout 99% type statistics.
This thought did seriously cross my mind in 2017, when we had increasingly bonkers data coming in day after day.
But in reality it would be a very very difficult thing to do.
— Chris Curtis (@chriscurtis94) October 19, 2020
Not clear what the outcome would be to justify the investment required, which would mainly be embarrassing pollsters on twitter. Plus most Russian exploits tend to be in technical terms surprisingly straightforward (in this area, anyway)…
— James Ball (@jamesrbuk) October 19, 2020
why is it always the Russians and Chinese in the cross-hairs.? The US and western based multi-national powers have equally as much access and ability. But here is one good reason why the paper and stubby pencil voting system, followed by open counting is the best way to do it.. We just have to persuade the punters to ignore the biased newspapers, the celebrity party leaders and the on-line chat, and simply concentrate on the choice of candidates in front of them.. ignore everything, the only question is which of these individuals will best represent you?
If the country in question is yours, yes, easy. You call all the heads of the polling organisations in and say, “Right: you are going to show 99% support for the President; and you, you’re going to show 97.8. Got that?”
As for unexpected events, yes indeed. I remember the Coventry being sunk two days before local election polling. We still won our ward, but I reckoned 150-200 Tory tactical voters for us went and voted Tory.