Is voting obsolete?
Matthew is on to something here:
OK, he is just (tounge-in-cheek) talking about primaries, but if you think about it, this could equally apply to national elections as well. Bear with me for a moment...
There are several major problems with voting :
1. Fraud increases the margin of error, which means in a highly polarized electorate, may become the same order of magnitude as the margin of victory, in which case the outcome of the election is essentially random to begin with. Technology is one solution, but is susceptible to funding, human error, and fraud of a different nature.
2. Victory requires a majority of the electorate alone, not the population, which means in a low-turnout scenario the winner can not claim a true mandate of consent to be governed. There is selection bias in the electorate because they are the ones who have the will/free time to go stand in line for hours to excercise their franchise.
3. The geography-centric structure of the electoral college means that some states are swing states whic garner disproprtionately more attention from candidates than others, which results in skewed policies such as the Farm Bill that do not benefit the nation as a whole.
A truly random sampling of voters, with N chosen high enough to give a margin of error arbitrarily low, would make it essentially impossible to perpetrate fraud while still delivering higher fidelity of results than is possible with the present system. Further, the selection bias is avoided because the vote can be done by mail or phone to a truly random cohort. Finally, the geographic considerations are mooted as candidates will have to campaign nationally on issues of truly national interest since they can no longer assume that their plurality will hail from a specific subset of regions.
What's the flaw?
NOTE: the idea is not new. I believe it was Asimov who wrote a short story where society needed only a single voter to determine the electoral outcome, assisted by a powerful computer that took the voter data point into account along with other social variables.
Also, I am deliberately avoiding including actual examples such as the role of exit polling in detecting Ukraine fraud, the well-documented racial voter suppression in Cuyahoga County, and the drama of the WA governor's race. Lets discuss the thought experiment in the absence of partisan bias.
Indeed, the whole voting concept may be outdated. You could just get together a list of every registered Democrat in the country, then take a statistically valid random sample of 1,000 or so of these people fly them all to a big hotel in Dayton (shades of Balkan diplomacy), and tell them they're not leaving until some candidate has the support of 600 people. Important elected officials, interest group leaders, and consultant types could make themselves available to hang around the premises and offer their off-the-record opinions.
OK, he is just (tounge-in-cheek) talking about primaries, but if you think about it, this could equally apply to national elections as well. Bear with me for a moment...
There are several major problems with voting :
1. Fraud increases the margin of error, which means in a highly polarized electorate, may become the same order of magnitude as the margin of victory, in which case the outcome of the election is essentially random to begin with. Technology is one solution, but is susceptible to funding, human error, and fraud of a different nature.
2. Victory requires a majority of the electorate alone, not the population, which means in a low-turnout scenario the winner can not claim a true mandate of consent to be governed. There is selection bias in the electorate because they are the ones who have the will/free time to go stand in line for hours to excercise their franchise.
3. The geography-centric structure of the electoral college means that some states are swing states whic garner disproprtionately more attention from candidates than others, which results in skewed policies such as the Farm Bill that do not benefit the nation as a whole.
A truly random sampling of voters, with N chosen high enough to give a margin of error arbitrarily low, would make it essentially impossible to perpetrate fraud while still delivering higher fidelity of results than is possible with the present system. Further, the selection bias is avoided because the vote can be done by mail or phone to a truly random cohort. Finally, the geographic considerations are mooted as candidates will have to campaign nationally on issues of truly national interest since they can no longer assume that their plurality will hail from a specific subset of regions.
What's the flaw?
NOTE: the idea is not new. I believe it was Asimov who wrote a short story where society needed only a single voter to determine the electoral outcome, assisted by a powerful computer that took the voter data point into account along with other social variables.
Also, I am deliberately avoiding including actual examples such as the role of exit polling in detecting Ukraine fraud, the well-documented racial voter suppression in Cuyahoga County, and the drama of the WA governor's race. Lets discuss the thought experiment in the absence of partisan bias.
Comments
You can find more information about it at: http://www.la.utexas.edu/research/delpol/