Problem 2 - Hill 2 - voter/legislative feedback

[davidnixon]

** 2nd Hill is Voters can view and understand the impact of their vote on every relevant piece of legislation and impact to their community:**

Real time poling of certified voters:
Should we create an app that would show realtime updates to legislatures and feedback to voters about issues. As an example consider H.R. 7120: George Floyd Justice in Policing Act of 2020 it was passed by the house but will it be passed by the Senate?
If you are a voter you might be interested in how your representatives voted on this. If you are a Senator you should be interested in how your constituents feel about this legislation. But the trouble is in connecting the two. I believe Senators are interested to a certain degree on polling but a much more powerful polling solution would be much more interesting.

• Allow voters to participate in polling directly through the app
• Elected official would get de-identified data on exactly who has participated in the poll
  • Raw number of "yes" and "no" opinions
  • Demographic information - including how many elections in which the participants have voted. i.e:
    • 50% of "yes" have voted in at least 3 of the past 5 elections
    • 10% of "yes" voted in the most recent democratic primary
    • etc.
• Voter would be able to see this same information and be able to decide if their elected official are effective

Uniqueness:

• I do not know of a solution like this. Ordinary polling relies on a random sampling and some guessing about who and who is not a likely voter
• Internet polls are usually completely anonymous and easy to spoof. You can vote for your opinion multiple times.
• Internet opinion polls cannot tell elected official if the people participating in the polls are even in their state/district. This solution should do that.

I believe most of the data to implement this kind of solution is available through state data. I have myself looked at absentee voting data in GA and was surprised to find my own name and address and when I voted is freely available for download.

Impact if implemented:
I believe if elected official understand what "likely voters" actually think about legislation they are empowered.
I believe that if voters understand that their voice is louder the more they vote, they will vote more.

Challenges:

• Getting data
• Certifying who is using the app. I think this is the most critical challenge. The app is useless unless we know exactly who is using it. The de-identified data would depend on this.
• ....

[Ladielindsay]

Hi David, I wonder if we can pair this with something folks already use, like twitter or FB, perhaps even state id (for non drivers) - true you can "fake" or have multiple accounts on twitter or FB platforms but maybe we give someone a unique code, or track their IP addresses. (Note. I am not a "technical" person). I like the idea though of having more direct information - in near real time about how local elections are moving AND how local officials are voting.

[davidnixon]

I like that. I think at least reusing your fb/twitter/google/ibm id/linkedin id to authenticate into an app would be great. Have you seen those "login with your google account" apps? I think that would certainly work but I do feel very strongly that the data is only useful if we have verified users. So maybe we mail them something to their voter address?

BUT once they are verified ... how about you can participate in a poll with a tweet? Maybe #AtSomePoll yes ?

[daveshack-ibm]

Hmm, public input like tweets/FB might discourage people from voting their conscience. We have anonymous (but identified) elections for a reason. Plus think of the conspiracy theorists going nuts over it. And peer pressure pushing people into voting in a poll on an issue they don't really care about.

I don't know how practical it would be, but the ability to write in more than just a yes/no answer would be extremely powerful both for voters to give more useful feedback and for elected officials to get the same. I know when we do internal surveys the percentages are almost always predictable but the comments often have hidden gems.

If write in comments are possible, then machine learning / analytics to boil that ocean down to a paragraph or a page would be a great bolt on to this.

[Ladielindsay]

To be clear, the voting would not be public. I agree and would like people to remain their anonymity. I was suggesting using validated accounts in a manner to confirm that the person is "real" as a manner of avoiding spoofing or invalidating our data.

[davidnixon]

@daveshack-ibm I think the public input issue a valid concern with FB/tweets. I don't envision that being the only way to participate but just a way to participate. I think there would still be a separate app for participation.

The peer pressure thing is also real. But given that there is a separate app I think that is fixable. You could (A) not enable your twitter account with the app or (B) if it is enabled you could "pause" it whenever you want so that none of your tweets/posts get counted until its un-paused.

I love the machine learning angle to open the input to more than just "yes" or "no". Watson understands sentiment so we could easily use that to sort through comments. Now I am imagining a UI somewhere where at least elected officials could see a list of "positive", "negative", and "neutral" comments. Maybe that would look kinda like what you would see in the reviews section of amazon? Could you mark comments as "helpful" so that the more "helpful" comments float to the top of the different sections?

@Ladielindsay I do think participation can be public but it would be up to the user to decide that. I personally would not tweet to participate but my adult kids certainly would.

Also, on vocabulary, I am proposing a "polling" app and not a "voting" app. I hope it would become "super, turbo charged, relevant empowering polling" but still ultimately "polling". So I do not think people "vote" with this app I think they "participate". I hope "participation" leads to empowered IRL "voting".