What are your ideas to run a national, non-binding referendum on the issues?

Politics & Elections
1

Suppose a candidate, party, or government wanted to take the temperature of their constituents with a national, non-binding “referendum” on the issues. They don’t want to use traditional polling for a variety of reasons, particularly because they want to increase voter engagement. The survey results would not compel any action, but they would be made public.

The goals, in no order:

  • Minimize Friction: Maximize the number of responses by minimizing the effort needed to participate. However, it is expected that some people will be unreachable or choose not to participate.
  • Constrain Responses: One response per person from the set of constituents.
  • Minimize Abuse: Eliminate widespread fraud and abuse such that the published results are meaningful. They are non-binding and do not have to be perfect, but need to be good enough to be useful.
  • Minimize Cost: This task is akin to the US Census, but it’s not practical to spend $10B or more on this survey.

I’ve defined the goals subjectively since it is premature to define objective metrics. For example, I don’t want to define an acceptable cost. Instead it is enough that Solution X might cost $Y and that may or may not be too much.

One obvious solution is a simple internet poll. It is cheap and easy, but completely unreliable; there are too many ways to “game” the results.

Using an internet poll as the baseline, what are some potential better solutions?

Consider your answers from your locale; for example: some countries might have infrastructure to support this already.

2

Are you sure that non-binding referenda are a good idea in the first place? My expectation would be that if the sane side wins the poll, it’ll be ignored, but if the stupid side wins, it’d be treated as though it were binding. That’s what happened with Brexit.

3

The obvious solution would be to have the questions placed on the ballot for the actual election, with the statements that these are explicitly there to determine what the people want rather than to be enacted as law.

4

Good point. I wasn’t clear exactly what I was asking. I am curious about a non-state sanctioned survey.

Adding questions to a ballot, both binding and non-binding, increases the authenticity of the results, but also increases voter friction. It also comes with a lot of approval baggage.

There could be value in a candidate or party holding a survey, publishing the results, and stating “this is what the people actually want and I will use this to guide my platform”.

There’s also an advantage that the question text doesn’t need to be negotiated into legalese that sounds the opposite of what it means.

It could be used to defuse wedge issues and engage with younger voters and generate excitement.

5

The fact that it’s non-binding would probably deter the majority of centrist or lukewarm folks from bothering to go out to vote (“why bother?..it’s just symbolic…”), and only the fanatic zealots would go to cast a ballot, thus making the outcome look much more far-right or far-left than public mainstream opinion actually is.

6

That’s why minimize friction is a goal – to maximum participation.

If a candidate for governor could survey people from the state (voters and non-voters, Dems and Pubs) such that people felt “listened to”, then that would be a powerful tool.

7

I can’t think of a national options that would beat an internet poll considering cost-effective options, even if prone to abuse.

The scale means that nearly anything is going to be pricey.

Speaking for myself, I’ve responded to things like Neilson requests because they offered $1-2 cash up front, and managed to generate enough interest and implied responsibility that I took the time and got another $5 cash after finishing.

That sort of thing might be effective (though how many just take the cash and don’t bother is a metric I don’t have access to!), but would probably be far too expensive compared to the baseline internet option.

The other common option I see these days, that is probably cost effective is setting up a mass-text message poll. For me though, 99% of the time I just click “message is spam” and move on. But the costs are probably low, and since our society is increasingly smart-device dependent, would get your information out to a very wide range of people without as much website based self-segregation.

8

Good point. I wasn’t really considering how to get out word about the poll and how to keep it from getting lumped in with all of the noise.

I was thinking if a presidential candidate did this, it would be enough to give it wings. Maybe the first time, but the once the novelty wore off…

For accuracy sake, would it be sufficient to simply collect name and address (like a petiton does)? Is it possible and cost-effective for a private organization to verify those?

9

I personally have no idea, and the more private identifying information you have to give increases the chance someone nopes out.

I think the text message polls are reasonably fair in that relatively few people have multiple active lines of service given the monthly fees. Sure there are a number who have both work AND personal phones, but I’d guess it’s a pretty small percentage.

You could do a simple whitepages lookup in moments (I just did on my number) and get information, but it could very well be out of date (mine was). I bet with the various paid reverse-look up options, you might get more accurate information, and it seems it would still be pretty inexpensive, though the sheer number means it wouldn’t be negligible of course.

10

Crazy idea:

I wonder if you could use a system where your survey is worth a fraction based on the information you choose to share?

11

If it’s set up by a candidate for a position, then it’ll never work, because no matter how robustly you set it up, the other side will always just claim that it was rigged.

12

For some cases, sure.

But for other cases, what would be a better survey solution than an open Internet poll?

13

Put it on the same ballot as the one for president.

14

Which issues, and how many? And if it’s not binding, why not run an opinion poll instead?

15

To engage with voters, to be an active listener, to get a wider view of your constituents, to try a more modern approach to campaigning, to build excitement among your base and comfort with the middle.

16

Just a comment, but if the questions are on the same ballot with a presidential general election, the results will be less progressive than if at a non-presidential congressional election, and much less progressive than if at an odd-numbered-year election.

Also, the results are going to vary depending on which party gets more of their preferred question wording.

Ideally, the wording would come from non-partisan polling outfits like YouGov or Pew. But in practice, that ideal is unlikely.

17

Among other problems, “national” elections aren’t national. They’re run by the states. What if a Republican administration wants to put a national referendum on the ballot and Democratic secretaries of state simply refuse to allow it? What if a Democratic governor wants to put a non-binding referendum on a state ballot and the Republican legislature passes a law that says “no non-binding referendums on the ballot” accidentally foiling the Republican President from holding a national referendum?

Not necessarily. Off-year elections and primaries tend to attract more activists on both ends of the politcal spectrum. Progressive or reactionary, the results will be skewed from the broad middle that (hopefully) votes in Presidential elections.

18

A forced yes/no answer is the opposite of active listening.

19

It’s an optional survey; no one is forced to respond. And survey questions don’t have to be yes/no.

IRL and on these boards, I hear a lot of frustration that their voices aren’t being heard. That primaries are controlled by the parties. That candidates are dominated by donors. I hear frustration that wedge issues are distracting from more important issues.

The idea for this survey is counter some of that frustration.

20

The first thing I’d suggest would be to separate the HOW from the WHAT.

WHAT n

What exactly do you want to know, and what are you going to do with the information?
I think this would make a big difference to the how, and also the when.
if we are looking at, for example, relatively straight forward issues like abortion or gay marriage, that are largely values (as opposed to science or reasoned analysis) based it would be one thing.

Should we allow gay marriage, is mostly a yes/no question

Should abortion be legal - likewise, while there will be a continumn of when or if, it’s pretty easy to gather accurate opinions. they might range from always, through “only in the first trimester” to “never”, there may be exceptions for mothers health and rape, but these are all (relatively) discrete.

For these two - I’d be mostly happy collecting public opinion

For something like Transgender participation in sport, I’d suggest the issue is far more nuanced., and the design of a survey question is going to be harder. Adding to that - I’d be wanting to know more about the which (general, not specific) groups of respondents said what.

Then what about issues that are contentious but firmly grounded in science - like Measles vaccine should be compulsory. With undeniable benefits, would it be wise to open that up to even a non-binding referendum?

As to the what - does it need to be a discrete thing only happening at election time?
What if it could be done continuously?

Setting aside all of the cultural baggage of the US -

In both NZ and Singapore, there is the ability to login to government services via a national ID number (like your social security number) with a password and 2FA.
It would be pretty easy to tag on a survey module to this - and with the advent of AI, collecting and analysing data is becoming easier and more accurate all of the time.
What if, overtime, something like this could be done?
The actual data collection via any of many govt based locations, from libraries, to polling stations, to DMV to at home?
It would allow for the type of demographics that I would like to see tagged to some questions, it would allow for a method to ensure that data was representative (or at least, let us see the flaws )
It would allow for ongoing data collection. It should allow for far greater nuance.
Over time, if we could build that appropriate level of confidence in the data collection, could something like this work?