Data Science for Improved Democracy

The current (November 2014) United States election reminds us that sophisticated data science techniques are employed on the public in attempt to influence opinion and persuade votes. The slick television advertising, debate prevarications, and policy position distortions and exaggerations have soured many citizens on the current state of modern democracy. Indeed, most feel we are not getting the straight scoop - the real positions and policy agendas - and our pool of public candidates and leaders is mediocre at best. The best and brightest avoid public office and it shows.

When Winston Churchill said "The best argument against democracy is a five-minute conversation with the average voter" he likely was referring to low-information voters as well as bad and biased information provided to voters to help them select political leaders and policy goals.

Michael Malak provides great insights in his blog post "Data Science for Reverse Democracy" about the use and misuse of data science by politicians to obtain more votes to win elections.

Brother Malak makes the case that data science today is being misused to invert traditional consent-of-the-governed democratic theory. The way democracy should work is the candidate or political leader states publicly their agendas and policy goals and hope voters ratify those goals in an election. Or an elected official does what he or she wants, then let voters approve or reject. This creates political accountability - vital to the health of democracy.

Yet strong evidence suggests the modern political class has subverted accountability. The way it really works is politicians cloak agendas in some mystery - use data science to tailor different specific messages to different segmented citizens (telling each segment what they want to hear) - then does what they desire anyway, even if a majority of voters reject the agenda disclosed. What this produces is a confused citizenry who have little to no idea what the candidate or political leader really believes and what in fact are the policy goals. As a result the polity finds it difficult, if not impossible, to distinguish political signal from noise.

I suggest this has produced both a cynical electorate and poor quality political leadership from all ideological persuasions. A lethal combination for healthy democracy in need of urgent repair.

What modern democracies need is an ideologically and politically neutral third-party organization that uses data science to distinguish between what politicians disclose and what evidence shows are the real agendas. In other words, help the public separate political signal from noise.

To exercise it's duty as citizens and to keep politicians honest, the public needs an easy to understand data visualization of real agendas and likely predicted coarse(s) of action - so citizens can make an informed public leadership selection - a vote based on reality (signal) and not hype (noise).

Data science has the potential to even the playing field and provide citizens with evidence-based information in the form of data visualizations - in contrast to slick advertisements and marketing - that are simple and informative, allowing voters to understand true agendas and policy goals - as well as achievements and failures based on facts and evidence to hold public leaders accountable.

Furthermore, ideologically and politically neutral data science can distinguish between successful and failed public policies based on facts and evidence and provide findings to both leaders to improve policy and citizens so they can hold leaders accountable and make better policy goal selections.

Modern democracies demand better leaders and higher quality information to make improved political choices for a better future.

See: http://bit.ly/1ow9EVF

Tags: Accountability, Agendas, Data, Democracy, Elections., Policy, Politics, Public, Science, Voters

1 member likes this

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Sione Palu on November 15, 2014 at 9:15am: I concur with Richard's comment that "However the readers are data Illiterate", I would further add that it basically means that voters are simply illiterate to political issues. Put being data illiterate aside, but one can check out the massive US debts live on the net which is 17 trillion (+) and still on the rise, yet voters voted for a second term for the current administration which has doubled the US national debt since taking office. I bet that majority of voters don't care or even aware that this is a problem that may lead to economic collapse of the US. History has told us that no economic super power can't avoid this problem of over debt. The final outcome is always a collapse. Japan is on a tight rope too, with its massive debt and they've been struggling over the last 20 years to come out of it with various monetary policies implemented over that period, but it is no better today than 20 years ago.

I think that the population/voters need to be educated in political philosophy/literacy first then may be follow with data literacy. If the population is data literate but being political philosophy illiterate then the historical problems of the past will keep recurring. Greece is an example of a such population. EU demanded a political reform before the loan is approved, Greece's population marched down in their capital & protested that their government shouldn't capitulate to the EU's demand, and wages/salaries shouldn't fall (pay cut for public servants) if reform goes ahead.

Comment by Richard Ordowich on November 10, 2014 at 8:21am: Great article but there is an overarching oxymoron that helps contribute to the perceptions about data and democracy and and the frequent unbridled acceptance by the media, the public and politicians of data.

Firstly, data science is a myth. Data is no more scientific than journalism or story telling. In the political arena, data is used to tell a story. The story can be fiction or non-fiction and the reader must be critical to discern the differences.

However the readers are data Illiterate and have not been educated in Data Literacy except for trite clichés such as "correlation is not necessarily causation". I draw the analogy how a literate population has a better opportunity to understand and participate in democracy. In today's world the population has to be Data Literate to participate effectively in democracy.

All data should carry the warning "all data lies". We should educate the public in data Literacy and stop referring to the practice of telling stories with data as data science. Then and only then will the citizenry be able to effectively participate in democracy in our data world.