[logfromblammo]

Studies on statistical aggregations often show that a crowd can make a better guess collectively than any individual member.

If you ask a crowd to guess the number of jellybeans in a jar, the people who are the worst overestimaters and underestimaters tend to cancel one another out, and the mean and median of all responses will be shockingly close to the actual numeric value. Therefore, my hypothesis is that reaching 100% voter turnout will be more beneficial per unit cost than any attempt to "inform" the citizenry already most likely to vote.

Here is a thought experiment. 20% of voters are knowledgeable about a subject, and vote accordingly. 80% vote based on a coin flip. How do the random voters harm the outcome of the vote? They add noise to the result, certainly. But if the signal from knowledgeable voters is unable to overcome the random noise, how certain can you really be that those people are correct? Is it at all important to know what percentage of all voters cared enough about the subject to self-inform, rather than just trust their lucky voting coin?

Outside a hypothetical, people are very rarely entirely ignorant of a subject. Even if they know only one true thing about it, when they vote based on that thing, it is incorporated into the statistical aggregate, and therefore influences the final result in some small way. If you restrict the vote to knowing certain things, only those things end up influencing the final result, and you can therefore bias the result by changing the test criteria.

[notahacker]

As you point out, being "uninformed" seldom means total ignorance and a random vote, so throwing relatively uninformed non voters into the mix is unlikely to be neutral

So the question is what factors people who are largely indifferent to and ignorant of politics will incorporating into the statistical aggregate to a greater extent than voters that do care.

I'd suggest that the little pieces that people who don't know or care very much about politics tend to be aware of are [i] the recognised status quo (incumbent, major parties) [ii] the status quo ante (incumbents actually picking up votes based on old campaign promises they didn't deliver! and "the party of Lincoln") and [iii] the most simplistic elements of campaign advertising and media coverage.

Are these likely to be signals of who will be the more competent and popular government which the more motivated and generally more informed voters have tended to unfairly overlook, or just noise?

[mason240]

Ask 1000 6th graders the value of Euler's constant, or what the best fiscal policy is, and see how good the results.

[michaelkeenan]

I used to think this way. The thing that changed my mind was learning how much people think the USA spends on foreign aid[1]. On average, Americans think 28 percent of the federal budget is spent on foreign aid, but it is about 1 percent. People base policy preference on their mistaken impression. When informed of the correct amount, the number who think America spends too much on foreign aid changes from 61% to 30%.

Foreign aid is one persistently misunderstood issue that I know of, but I worry that there might be many similar issues.

[1] https://www.washingtonpost.com/news/wonk/wp/2013/11/07/the-b...

[logfromblammo]

To be more precise[0]:

  2013 US Budget                 ~= $3 803 300 000 000
  +- International Affairs        = $   52 018 676 000 ( 1.37%)
     +- State Operations          = $   17 702 825 000 ( 0.47%)
        +- Int'l Orgs             = $    3 386 331 000 ( 0.09%)
     +- Foreign Operations        = $   33 810 927 000 ( 0.89%)
        +- Bilateral Assistance   = $   21 134 577 000 ( 0.56%)
        +- Int'l Security         = $    8 791 500 000 ( 0.23%)
        +- Multilat. Int'l Orgs   = $    2 875 204 000 ( 0.08%)
        +- Foreign Banks/Funds    = $    2 548 553 000 ( 0.07%)
        +- Direct Food Aid        = $    1 533 859 000 ( 0.04%) [1]
        +- US AID                 = $    1 450 806 000 ( 0.04%)
        +- Independent Agencies   = $    1 258 585 000 ( 0.03%)
  +- US Dept. of Defense         ~= $  672 900 000 000 (17.7 %)
  Total Military SA+FR+UK+DE+JP  ~= $  283 500 000 000
  2013 AAPL total expenditures   ~= $  136 000 000 000

What people think of as "foreign aid" may vary. The budget covers everything from bed nets to bullets. If you count only spending on operations that most directly assist poor foreigners, such as US AID, Peace Corps, and UNICEF, rather than just writing checks to foreign politicians, militaries, and bankers, it amounts to about $7-$10 billion, or 0.2%-0.3% of the budget.

And nothing in the US budget takes up more than 25% of it. The top 4 items are, in fact, Medicare/Medicaid, Social Security, military, and debt service (6.5%!). Even if you count all military spending as some sinister form of foreign aid, you can't get to 28%. The problem there is not just being uninformed. Someone must be actively spreading misinformation--lying to the public. That's a much bigger problem than simple ignorance, and trying to restrict turnout to only "informed" voters is not going to help when people believe they are informed after hearing enough lies.

[0] http://www.state.gov/documents/organization/224071.pdf [1] paid to US Dept. of Agriculture

[hackuser]

Nobody was permitted to see the Emperor of China, and the question was, What is the length of the Emperor of China's nose? To find out, you go all over the country asking people what they think the length of the Emperor of China's nose is, and you average it. And that would be very "accurate" because you averaged so many people.

http://www.textbookleague.org/103feyn.htm

[logfromblammo]

The averaged length is likely to be close to the length of the median Chinese person's nose.

It would not in any way be accurate in terms of judging the fact that is the length of the Emperor's nose, but it would be the best possible guess China could possibly make based on the information then available to it.

You could as easily ask America how many grams of cocaine George W. Bush used while avoiding overseas military operations. And if you can ask that, you can also ask whether Donald Trump or Hillary Clinton would make a better U.S. President.

You don't vote on facts. You vote on opinions. And if you want to know "America's opinion", more responses will yield more accurate results.

[hnamazon123]

The comparison is fundamentally flawed because the amount of jellybeans in a jar is an objective fact not subject to interpretation by anyone's values.