The Economist magazine has a sister company, The Economist Intelligence Unit (EIU), which, among other things, maintains what it calls a Democracy Index. You can find the details of how the numbers are calculated on the web site, so I won’t go into them here. According to the index, democracy is in retreat, globally. According to the magazine “The global score of 5.44 out of ten is the lowest recorded since the index began in 2006. Just 22 countries, home to 430m people, were deemed “full democracies” by the EIU. More than a third of the world’s population, meanwhile, still live under authoritarian rule.”
In one of my previous blogs this spring, talking about the climate crisis, I noted that the very design of social media and the internet’s information seeking mechanisms are a poor way of spreading useful or correct new information and countering old or harmful information. This design flaw is not only problematic for solving the climate crisis, I think it has worsened the practice of democracy itself.
I wrote then that “The internet and social media excel at making huge volumes of information instantly accessible to everyone. This has led to a kind of lowest-common-denominator democratization of the perceived value of information on the internet where every piece of information is as good as another. There are no good mechanisms in either search engines or social media for determining what is valid and what is not, unlike traditional sources of information on which decisions were based.”
Also, I wrote that representative democratic forms of government “are structurally adversarial (i.e. based on some variation of the Westminster form of parliament, where the government and opposition benches are placed literally two sword lengths apart) rather than collaborative. While there can be plenty of dissatisfaction with the status quo, this model of government is not designed to seek shared meaning or to agree on a common vision of what is possible.”
This is not to say that there are not other threats to democratic processes, such as things like rampant gerrymandering, voter suppression, foreign interference, active disinformation by Trumpists and other populist/autocrats, but in this blog I want to focus on the role of the internet.
Democracy, which depends on an informed and engaged electorate, is difficult enough at the best of times. We all suffer from cognitive biases, which Wikipedia defines as “systematic patterns of deviation from norm and/or rationality in judgment”. For a graphically presented list of a dizzying array of such biases, see this graphic: https://www.visualcapitalist.com/wp-content/uploads/2017/09/cognitive-bias-infographic.html It makes for a fascinating rabbit hole to chase if you can spare the time. What this means is that what we see is biased by our own preferences, but that these biases and fringe beliefs and positions are exaggerated and amplified by social media, which is main way that most people experience the internet now.
This is not just conjecture. The Economist ran a test of how the Twitter recommendations algorithm works. They concluded that: “the recommendation engine appears to reward inflammatory language and outlandish claims.” and “Its business is serving ads to 330m users, even if that means grabbing their attention by showing them exactly what they want to believe”. Such beliefs which are frequently not based on fact or truth, sows division and discord.
A dire situation. But what if the internet could also be used to engage people constructively to inform government decisions and actions?
This is the premise behind Polis (https://pol.is/home) an open source “real-time system for gathering, analyzing and understanding what large groups of people think in their own words, enabled by advanced statistics and machine learning.”
The way Pol.is works, as I understand it, is that on any given topic, anyone can come in and comment. What’s different is that you cannot reply to comments at all. The only thing you can do is agree, disagree or pass on the comment. Apparently this does away with escalating flame wars. What it also does is generate a matrix, a large database of how a lot of people feel about any given comment. Machine learning then analyses those replies and visually presents them in aggregate, which very quickly begins to show where there is consensus and where there is not.
Sound good in principle but what about the real world? The system, launched and run by civilian “civic hackers”, has been used in Taiwan to inform government decisions on how to best regulate ride-sharing and how to enact regulations for the platform economy, according to Audrey Tang, the Digital Minister of Taiwan: (https://www.economist.com/open-future/2019/03/12/inside-taiwans-new-digital-democracy ).
To check whether this rather optimistic view of the platform and its use in Taiwan is shared by those not quite so invested in viewing it as a success, I wrote to a friend of mine who lives in Taiwan. He said that he had not been personally been involved in participating in any of the initiatives launched by the platform but that he thought it could be used positively but also had the potential for government to use it to confirm its previously decided direction. As he has not been involved, that is necessarily an outsiders view.
At a much more modest scale, Pol.is was used to run a virtual town hall in 2018 in Bowling Green, Kentucky, around the question “What do you believe should change in Bowling Green/Warren County in order to make it a better place to live, work and spend time?”. https://www.bgdailynews.com/civicassembly/
The results (https://www.bgdailynews.com/civicassembly/2018_civic_assembly_results/ ) showed that there was high consensus on 5 major issues (>76% agreement and very little disagreement) and 5 issues that were quite contentious (disagreement, agreement and pass percentages fairly evenly split). These results were then used to program a face-to-face town hall and after that a working group. The most significant thing to me is that I don’t know of any other process that could so quickly uncover consensus in a large group of people in such a transparent way.
The pol.is web site has links to many other case studies around the world, as well.
Whether Pol.is is that last word on how the internet can be used to show people what they do and don’t agree on in a non-emotionally charged way, I can’t say. Also, there is not guarantee that the results of such engagement will be used by governments in a way the people intended, but that is no more problematic than expecting politicians to live up to campaign promises now. At least with this approach politicians can’t say they don’t know what people say they want.
I see this system as a positive development and an encouraging sign that the internet, in this age of socially-mediated, always-on, 24-hour surveillance capitalism, can also be a tool for progress.