Can we design a voting system that could withstand the attacks on modern democracy
Democracy is dying. The governance system first established in 507 BC in Athens is slowly decaying. Manipulated by the wealthy and powerful, it does no longer serve the people.
Evidence is everywhere and has been there for a long time. It is not solely exemplified by the 2016 US election skewed by targeted social network campaigns, or the proliferation of fake news. Democracy has been hacked a long time ago by big money. Money that buys lobbyists, lobbyists who buy politicians and media, and media that creates (a false) reality.
Fundamentals of Democracy
Any functional democracy is based on these two assumptions:
- People are rationally self-serving
- People are well informed
However human decision making has proved to be irrational, emotional, heuristic and suffers from a number of cognitive biases that can (and are) being used by those running for office. Which leaves us with: well informed.
Are we really?
To answer that question we need to take a quick look at voters’ behavioral patterns.
Why Do We Vote (as we do)?
An interesting article in Psychology Today states a number of non-obvious factors affecting voting patterns.
In the 2016 US presidential election, roughly one hundred million eligible voters didn’t vote. Voter suppression and discontent are oft-cited reasons for these dismal numbers. But studies show that there are other less obvious factors that can significantly influence not only voter turnout but also an individual’s decision-making process in the voting booth. They reveal that even the most earnest voters can be swayed.
Weather. Can hot temperatures influence voter turnout? According to the findings of one study, the answer is yes. When the temperature heats up, it increases physiological arousal. In turn, heightened arousal can increase both antisocial and prosocial behavior — like voting. This phenomenon is in accordance with what’s technically known as excitation transfer theory, which maintains that the excitation from one stimulus can be transferred to influence an individual’s response to another stimulus.
In order to investigate whether hot temperatures influence voter turnout, researchers examined the relationship between temperature and voting, using data from presidential elections from 1960 to 2016 in each state in the United States. What did they find? Changes in temperature were positively related to voter turnout. What’s more, when temperatures were higher, voters were more likely to show leniency with the party in power and support the incumbent party.
Weather, again. According to multiple lines of research, bad weather on an election day drives down voter turnout — possibly swinging the results of an election. But the weather could affect more than just voter turnout. Studies show that weather also affects mood, and in turn, could influence decision making in the voting booth.
Consider research (a working paper) that looked at how weather (measured in terms of precipitation, sunlight exposure, and subjective evaluation) influenced constituents who turned out to vote when candidates were viewed as being more or less risky. Even after the researchers controlled for policy preferences, partisanship, and other pertinent variables, bad weather was found to depress individual mood and risk tolerance, such that voters tended to vote for the candidate who they perceived as less risky. In other words, bad weather made voters in this study more risk-averse.
Competence. Studies show that voters tend to support candidates whose voices and faces they find competent. Yet, what is the simultaneous effect of these two characteristics? Take a two-part study that explored this question. In the first part of the study, participants rated the faces of Members of the U.S. House of Representatives for competence. Then, the faces that were rated the most and least competent were paired with recordings of competent (i.e., lower pitched) and incompetent (i.e., higher pitched) voices to produce “simulated candidates.” In the second part of the study, a separate group of participants voted between these randomly generated pairs of simulated candidates. The results were fascinating. Candidates with competent faces or competent voices garnered more votes — but the effect of facial competence was nearly three times that of vocal competence. In other words, candidates fared better if they had a competent looking face than a competent sounding voice.
Mental overload. Voters use psychological shortcuts (i.e., heuristics) to help them make decisions when they lack information about candidates for office. The way a candidate looks is one particularly potent shortcut voters use to form a judgment about a candidate. However, consider research that points out that forming a judgment on the basis of physical appearance can drive down support for candidates of color because of insidiousness racial stereotypes. One particularly revealing study argued that racial prejudices would be more likely to influence voters’ decision making when voting environments demanded more of their cognitive resources — such as having to choose multiple candidates at once. In experiments involving simple and complex cognitive tasks, the results revealed that black candidates won less support from voters who were cognitively overloaded than from voters who had the “cognitive space” to consciously keep their prejudices in check when voting. And remarkably, this pattern was especially salient among politically liberal voters. Thus, the findings demonstrate that participants who express politically liberal views support black candidates more often than white candidates when the cognitive task was simple — but were less likely to do so when they were mentally overloaded.
Do you still think democracy is the best possible political system?
How Well-Informed Are We?
It depends on your definition of “well”.
Surely we are highly informed, possibly the most highly informed society in human history. But are the massive amounts of data making us better informed?
The simple answer is no, the more gruesome one is: we are highly misinformed.
With social networks becoming our main source of news, quickly followed up by mainstream media, our news feed has become a targeted echo chamber controlled by those who are willing and able to pay for our data and manage massive online campaigns to sway votes utilizing our built-in cognitive biases. We are becoming puppets in a reality show whose rules are governed by the unscrupulously well-funded.
In Ancient Greece, only citizens could vote. Children and slaves were not considered citizens, so they could not vote. Women, though considered citizens had no political rights, and were not allowed to vote either.
As slavery was abolished, and women have had voting rights as early as 1881 (in the Isle of Man) gradually going global throughout the early 20th century, it would seem that the notion of equal vote is a natural part of modern western democracy. A birthright.
But people are not equal. Some are more intelligent than others, some physically better built, some born rich, some uneducated, others are prone to bad decision-making, yet, when it comes to determining who would control every aspect of our lives, everyone with a pulse has an equal say.
One could argue that populations are naturally distributed such that the one-person-one-vote system correctly reflects the normal distribution of political agendas within any given population. Such an assertion is easily refuted due to the inherent bias built into the current democratic process. When you can use mass media and social networks to create false realities swaying and suppressing voters, there is no normal distribution anymore.
But let us go deeper.
Should all votes be considered equal?
If you are asking your bank for a loan, they examine your credit history in order to decide whether to supply you with the requested cash. If you want to drive a car you are obliged to be properly pre-trained and pass a mandatory state test. Why would voting be any different? Why should we allow people who have a history of bad decision-making to take part in the democratic process with a single, equally distributed vote. When you think of it this way it borders on irresponsibility.
At this point, the reader may be inclined to put down this article at what she takes to be a blatant attack on the very essence of democracy as the partnership of equals, coexisting, jointly determining their shared faith. However, the current system is none of the above, and submitting to this romantic view of democracy would most probably be its demise.
There is no partnership of equals, as big money has much more power shaping reality than the minions targeted by data companies led to believe they are making decisions based on truthful data.
There is no partnership of equals as many people vote based on emotion, lack of data, biased gut feeling or never vote at all.
There is no partnership of equals since many voters do not have the capacity to handle even the simplest forms of logical deduction required in order to see through campaign fallacies, spins, and manipulation of facts.
So. What should we do about it?
First, we must do away with the one-person-one-vote axiom. Then, devise a system that truly works for the people, taking into account cognitive biases, mass social media adversaries, and decision-making abilities’ diversity.
Quadratic voting (QV) is a collective decision-making procedure, where participants express how strongly they feel about an issue rather than just whether they are in favor of it or opposed to it.
According to its authors, Quadratic voting is claimed to achieve the greatest possible good for the greatest number of group members (although other proponents of Quadratic Voting admit that is at best an approximation). It addresses issues of voting paradox and majority-rule.
Based on market principles, each voter is endowed with a budget of “voice credits” that they may spend influencing the outcome of a range of decisions. If a participant has a strong preference for or against a particular decision, additional votes can be allocated. A vote pricing rule determines the cost of additional votes, whereby each vote increasingly becomes more expensive.
The quadratic nature of the voting means that a voter can use his or her votes more efficiently spread across many issues. For example, a voter with a budget of 16 vote credits can apply 1 vote credit to each of 16 issues. But if they feel strongly about a single issue and apply 4 votes at the cost of 16 credits to a single issue. This will use their entire budget. This also means there is a large incentive to buy and sell votes, although using a strictly secret ballot gives some protection against vote-buying as the purchase cannot be verified.
As radical as Weyl’s method may seem, it only partially addresses mass social media bias, and completely ignores the fact that people greatly differ in their decision-making skills.
To overcome the latter we need to radicalize voting even further. We must convert democracy into a meritocracy.
What if we keep a record of decisions taken by each adult citizen, much like banks and nation states maintain credit history, ranking each decision according to its goodness. I.e. how well did it benefit the individual making it(bar illegal ones of course). After all, we are already tracked and targeted, so we may want to put our data to good use.
Once we have collected it, we could properly assign a weight for each individual voter based on his overall decision-making score (DMS).
When the general elections are held, each vote would be weighted according to its voter DMS or, better so, determine the number of QV voting tokens allocated to him.
Is this practical?
The first thing that needs to be considered is implementability. After all, each of us makes hundreds of decisions every single day, some of which are life changing while others are minuscule. How would a credit-history-like system differentiate between the above let alone properly score our decisions?
To overcome this seemingly tough hurdle we need to create an ontology of decision classes. Such an ontology may contain economical, medical, career and marital decisions. Each class may be further hierarchically detailed until we get an agreed-upon construct that best reflects individual decision-making process outcomes.
If you think this is sci-fi, take a look at the online ad industry intent and topical ontologies, or Facebook’s targeting parameters (careful, it may come as a shock).
Once we get the ontology right, it becomes easier to imagine how decision-making data could be individually tracked. Banking apps, job listings, career moves, health status, are already digitally stored. Your Apple Watch, your LinkedIn account, your bank statements, as well as your insurance profile and much more, can provide a detailed and accurate ontology-based profile of your past and present decision-making abilities.
Opponents of the above may suggest that such a system can put too much power in the hands of the well-educated, creating a meritocracy of the wiser, replacing that of big money. That is not the case.
As we measure individual decision making rather than IQ (or some other similar metric) we allow street-wise as well as PhDs to rank high regardless of their college education.
Another argument against weighted voting may be that by creating such databases we infringe on the right to privacy. Sadly, privacy has been long gone for a number of years. We are already tracked and targeted on a daily basis with ever-growing accuracy (and it is not going away anytime soon).
This massive data collection has allowed adversaries to manipulate us, to manipulate democracy, to set us each against the other for their own vile purposes. It is about time we the people use our data in a way that benefits the greater good. Our good.