Public: Politics
Open source distributed political big data project   (+1, -1)  [vote for, against]
Scientific basis to politics and governance

Most people agree that governance is necessary; most people acquiesce in the use of politics to form governments. So far, only private and ethically dubious political movements seem to have used large-scale data analysis in their political campaigns.

I imagine that most scientists think that public policy and governance should have a rational and scientific basis.

Many political and governance decisions do have an element of moral decision-making, which is perhaps where voting has its role.

Suggested is that scientific bodies and instiitutions, on a voluntary basis, donate staff time and computer resources to create a large-scale global distributed project or system to gather data, analyse it, create social and economic models, and disseminate the results. Data could be gathered through mining social media and other methods that are coming to be used. Results would be open source and open access to allow anyone to use them. But part of the project would be filtering results and presenting them in as simple a way possible (consistent with them being true), always linked back to the source data.

This may allow political debate to become more solidly grounded in empirical research, and allow political choices to become limited to the moral assessment of the data and situation.

For example, different Brexit deals or global-warming-mitigation-strategies or political party manifestos could be analysed by real-time economic modelling, the likely benefits and disadvantages to groups and individuals could be presented. Paid adverts could be shown to target individuals flagging up the actual likely outcomes of different courses of action. The moral judgements leading to different situations could be highlighted, and then people could be encouraged to make their political choice on explicitly moral grounds.
-- pocmloc, Oct 26 2019

oh for fuck's sake.

According to McGann, there are currently 1.777 think tanks in the USA, 58 % of which were founded during the past 25 years. In terms of the number of think tanks, the USA ranks first worldwide, followed by Great Britain and Germany.

Are we starting to post "why can't we just get along ideas" now?
-- theircompetitor, Oct 26 2019

What is a think tank?
-- pocmloc, Oct 26 2019

//1.777// So, almost two?

//What is a think tank?// It's a sort of armoured, tracked carrier for philosophers.
-- MaxwellBuchanan, Oct 26 2019

So a research tank that mediates unquestionable facts, on current situations, where everyone is still free to make the stupid decision? Sounds like what we already have.
-- wjt, Oct 27 2019

I know this idea is incoherently formed and badly expressed but I don't think it is a think tank because it is not about thinking and proposing policy. I don't think it is what we already have either. The novel suggestion is the large scale data mining and aggregation, to feed into modelling of current and potential scenarious. Think of climate science, where data gathering feeds into model making. The interpretation of the climate modelling is left to emotionally driven self-interested stakeholders, but the science is trying to be as large-scale and open and rigourous as possible.

At the moment political data gathering and modelling is done in a closed way, by partisan players. I think from a scientific point of view this hugely reduces the value of the research because the conscious and unconscious biases of the researchers will colour the exercise, and also because the lack of openness means there cannot be the scientific cross-checking of the data gathering and modelling methodologies.
-- pocmloc, Oct 27 2019

// The novel suggestion is the large scale data mining and aggregation//

You think governments and other research bodies are not sufficiently mining data? This is 6 years after Snowden, when the Chinese have already implemented an "asshole flag" for travel, insurance companies offer discounts for speed monitors and wearing the apple watch, and the UN issues dire climate report warnings based on C02 concentrations variations of one one hundredth of a percent in the planetary atmosphere and Facebook has been used to swing an election to an improbable winner? When A/B tests are run on a continuous basis on every commerce website you've ever used? And finally, when every politician (not counting Trump) poll tests every sentence that comes out of their month?

Data mining might make the self-driving car drive better, or discover a hidden risk or benefit to a lifestyle, but it cannot uncover some larger "political" truth that then magically would convince everyone to agree.
-- theircompetitor, Oct 27 2019

All those examples are closed, proprietary, and presumably biased. I'm thinking of a neutral open version which is communally held to the highest scientific standards.

Am I stupid, naïve, or all three?
-- pocmloc, Oct 27 2019

//Am I stupid, naïve, or all three?//


Only if you include numerically illiterate, not sure if that was deliberate or a typo? :)
-- Skewed, Oct 27 2019

//All those examples are closed, proprietary, and presumably biased. I'm thinking of a neutral open version which is communally held to the highest scientific standards.//

It used to be that you were only entitled to your own opinion, not to your own facts. But this has morphed into "you are entitled to use facts to attempt to change my opinion. good luck with that".

this is still a "why can't we all just get along" idea.
-- theircompetitor, Oct 27 2019

Well, I disagree, because I think you're wrong. I don't think we can all just get along.
-- pocmloc, Oct 27 2019

No-one could disagree with a single data point. It is what it is, there and then with the equipment used. The blurriness comes with the attempt to say 'factually' the scientific lines between data points, hence statistics, probabilities and error rates.

Having a certified organisation come out with a current top 100 highest probable causations might not be a bad idea. But like a leader board it's going to change with data points with a new aspect and the associated analysis.
-- wjt, Oct 27 2019

random, halfbakery