By Prof. Dirk Helbing and Stefan Klauser, ETH Zurich
Recently, we have heard many complaints about how democracy works these days – or maybe rather why it doesn’t work. In a recent Huffington post article, Dhruva Jaishankar, a Fellow at the Brookings Institution in India, claimed that digital democracy is the evil that makes our world ungovernable.[1] We argue that Iaishankar defines digital democracy in a flawed and misleading way. This could cause serious misunderstandings of what the problems are and what are the possible solutions. In the following we will show that digital democracy – if properly understood[2] – is the most promising way to build prosperous societies in the digital age.
Brexit, Trump, AfD - is the Internet creating protest voters?
Most commenters agree that we see a polarization of society in recent times. This has to do with the way modern mass media and social media work. They tend to create ‘filter bubbles’ reinforcing the own opinion, while reducing the ability to handle different points of view. They become increasingly personalized, manipulative, and deceptive, spreading oversimplified messages or misinformation.
It is in fact true that the Median Voter Theorem, stating that in a two party system parties should move towards the center,[3] has come to its limits. The political positions of parties in a majority-based two party system do not always develop like this. Sometimes they actually get more polarized. The presidential elections in the US, where candidates tend to stress ever more extreme positions, provide clear evidence of that. In Europe though, we find a different pattern. The centralized bureaucracy in Brussels increasingly supersedes the federal (region-based) organization and, thereby, violates the well-established subsidiarity principle. As the so-called Bryce’s Law[4] predicts for young institutions, there is evidence for an increasing centralization of the EU.[5] However, considering the growing diversity in the European Union (with marked differences between rural and urban areas, and between groups of different age or cultural background), it becomes clear that an over-standardized “one size fits all approach” decided by the majority or most powerful does not work anymore. This incumbent-dominated governance engenders unpredictable reactions from diverse citizens, and the evidence shows that the neglected citizens react by voting for protest candidates and/or supporting radical solutions to complex challenges (see Brexit).[6]
Politics is not one-dimensional
It is certainly true that the societal and political complexity has dramatically increased with the multitude of interdependencies in our highly networked, globalized world. Therefore, the question is how to deal with the compounded complexity and unpredictable outcomes - even for experts and for expert systems (based on big data and artificial intelligence). Some commenters suggest that voters cannot handle this complexity - they would be easily manipulable and lean towards populist and inadequate solutions. And therefore, voters should be disregarded overall in favor of a data-driven society, they say. Then it would be Google’s “omniscient algorithm” or IBM’s cognitive computer, called Watson, deciding about what had to be done. Depending on the implementation, this experiment may very well end in fascism 2.0 (a big brother and brave new world society), communism 2.0 (distributing rights and resources based on a “benevolent dictator” approach), or feudalism 2.0 (based on a few monopolies and a new kind of caste system). However, the data-driven variants of governance models that have failed in the past will not suddenly become more acceptable. The question is rather how to use the digital opportunities of today and upgrade democracy, “the worst form of government, except for all the others”, as Churchill framed it.
Democracy 2.0 – how to harness collective intelligence by digital means
The long-term consequences of centralized top-down control could be devastating due to a loss of socio-economic diversity and resilience, a decline in the innovation rate and socio-economic progress, political instability and war or revolution. Centralized top-down optimization may be a proper paradigm for companies or supply chains, but complex societies need pluralism and combinatorial innovation to thrive. The success principles of the past - globalization, optimization, and administration - have more or less hit their limit. To reach the next level of society, an economy dominated by networks must build on the principles of co-creation, co-evolution, and collective intelligence.
Overall, to achieve culturally fitting, sustainable and legitimate results that leverage the benefits of complexity and diversity, it is crucial to move from a government paradigm based on power to a paradigm based on empowerment. Combining smart technologies with smart citizens is the recipe to create smarter societies. This can be reached by creating Massive Open Online Deliberation Platforms (MOODs), which allow all interest groups to put their arguments on a particular subject on a virtual table, where they can be structured into different points of view. Or in the words of Landa and Meirowitz: “In revealing correct, fuller, or simply better organized information, deliberation provides an opportunity for participants to arrive at more considered judgments themselves and to affect collective decision making by influencing the judgments of others.”[7]
In a second step, it is important to work out innovative solutions that integrate several perspectives and, thereby, benefit several interest groups well, not just the incumbent or 51% majority. This is the essence of “digital democracy”. It is based on “collective intelligence” - on bringing the knowledge and ideas of many minds (and artificially intelligent systems) together. It is the combination of ideas and interaction of humans that have shown to deliver the best results in most challenges.
While ensuring collective innovation, an updated democratic process should be able to reach equally distributed opportunities and satisfaction, as much as this can be done. While this cannot always be achieved in each single decision, we could certainly get much better in satisfying diverse interest groups than today. Putting it differently, digital democracy is about creating the digital tools to make deliberative democracy as described by a Habermas or Fishkin work efficiently.
So, instead of trying to revive governance principles of the past, which have failed to embrace the complexity and diversity of modern societies, we should engage in digitally upgrading democracy. After all, being the result of many wars and revolutions – democracy is a highly advanced governance system that has taken on board the wisdom of some of the smartest and most respected people in human history. Rather than accepting data-driven governance to control and abate societal diversity and complexity, we propose a way to leverage complexity for our benefit, through a platform of participation and decentralization.
Overcome the dictatorship of the majority
Some scholars state that, especially in polarized societies and societies with significant minority groups, simple majority voting is not adequate, because it can lead to a dictatorship of the majority over the minorities. Additionally, extremely close votes (see Brexit vote in the UK and the vote on mass immigration in Switzerland) often evoke protests about the fairness of this process. The loosing side, representing a significant share of the population, fears suffering under the new rule and has a strong interest to fight for an agreeable implementation of the initiative.
The fact that large minorities are often being ignored presents a serious issue, which is not easy to solve. However, with the means of MOODs, one can find solutions that consider various views on certain aspects of a topic. Today, one of the main problems is that people can only cast a “yes” or a “no”, i.e. to either agree on a proposed solution or disagree. The topic is often extremely complex and has many facets. So, letting people decide about “yes” or “no” is simply not enough. We suggest that citizens should be able to continuously engage in a specific type of online deliberation processes, where they can feed in their ideas and voice their preferences on different aspects of a topic. Brexit, for example, has many implications for Britain. Most voters saw some advantages and some disadvantages of it. Had one known how the electorate ranked the importance of the different facets of the political issue (for example, regarding immigration on the one hand and economic interdependence on the other hand), policymakers could have tried to disentangle some of the aspects and prevent severe dissatisfaction of the electorate. One could argue that politics is not a self-service wonderland, so voters would ultimately have to choose between different packages. Nevertheless, it is pretty clear that understanding voters’ preferences could improve political processes and lead to tailor-made solutions, especially when a reasonable level of federalism is embraced.[8]
A refined, more inclusive process has several advantages. It enables people to learn about the different aspects of a complex political topic. At the same time they can contribute to the solution from the beginning, which is believed to lead to a higher satisfaction.[9] Analogously, it should diminish the chances that protest movements and extreme solutions will find good breeding grounds.
Even if the results of the deliberation process would not be binding for policymakers, they MOODs would give them ample guidance when drafting new laws. It would also be possible to take regional, ethnic and religious differences into account, which could lead to culturally fitting law-making and easily show whether it makes more sense for a specific law to be adopted on a federal or on a regional level. There could still be a majority vote at the end of a deliberation process. But at this point, the solution would already include a substantial amount of the ideas and wishes of the citizens and it is likely that we would not see extremely polarized situations anymore.[10] Regardless of whether a proposed new law engendered a 50:50 polarization of society, deliberation processes after the vote could substantially lower the dissatisfaction of the minority, especially again, when mixed with a high level of regional autonomy in the way the vote / law is being implemented.
Counterbalance misinformation / spread of extreme views through (Social) Media
Missing media competence, scarce time, filter bubbles and financial means of potent influencers all lead to the spread of inaccurate, misleading or even wrong information. For citizens it is increasingly hard to judge, which information can be trusted and why. Governments, companies and rich individuals today can buy armies of bloggers and social media experts to run profiles and chat bots, which are flooding social media channels with the information they want to spread. Most of today’s largest social media platforms have currently no means to moderate these discussions. This information asymmetry contradicts the notion of an authentic deliberation process as e.g. defined by John S. Dryzek.[11] We thus need to create new platforms allowing for an informed, balanced, conscientious, substantive and comprehensive deliberation processes.
Deliberation and influence exercised through ranking, voting and discussions
There are some important features that the MOODs need to have: 1) They should be transparent in the way they work and decentralized, to reduce manipulation and censorship. 2) They should be moderated (by community-elected moderators) to ensure that discussion are constructive and fair. 3) Artificial Intelligence should be used to detect abnormal activity and reveal chat bots as well as ghostwriters. AI could also be used to organize the arguments made and support a multi-faceted discussion. 4) Reputation systems should incentivize responsible behavior and high-quality contributions. For example, one could give a greater visibility to contributors with a higher reputation. It is also important to detect manipulative rating activities. 5) Finally, a transparent and fair qualification mechanism could determine the roles that individuals can play in the deliberation process, and what additional data and platform functionality would be at their disposal.
To define Digital Democracy merely as democratic processes in a media-dominated and digitalized world falls short of what a reasonably advanced idea of Digital Democracy encompasses. A sophisticated model of Digital Democracy must be based on the concepts of co-creation, co-evolution and collective intelligence, enabled through the use of modern digital means.[12]
It certainly takes a substantial amount of work to build these platforms and to upgrade democratic processes to be fit for the digital age. The task we have to accomplish has technical, legal and motivational aspects. Especially the question of how to engage enough people in the deliberation process will be crucial. One has to secure easy access to the platform and experiment with incentives and gamification to reach sufficiently broad participation. However, there are no obstacles that could not be overcome. The potential benefits of a suitably refined (direct) democratic process clearly outweigh the costs of turning history back and neglecting “we the people”. Digital democracy supports society’s historical achievements: self-determination and freedom, the division of power and fairness, social inclusion and participation as well as diversity and resilience.
Dirk Helbing is Professor of Computational Social Science at ETH Zurich, Switzerland, and member of the German Academy of Sciences “Leopoldina”. He leads the nervousnet.info initiative, which wants to open up the opportunities of the Internet of Things for everyone.
Stefan Klauser is a Political Scientist and FinTech Expert and leads the subject “Digital Society” in Professor Helbing’s team. While working for the Swiss Department of Foreign Affairs, he was responsible for creating an exhibition on Modern Direct Democracy. Later, when working for the Swiss State Secretariat for Education, Research and Innovation, he was Project Leader Innovation Initiatives and Swiss Representative to the OECD Global Science Forum.
[2] D. Helbing and E. Pournaras, Build digital democracy, Nature 527, 33-34 (2015): http://bit.ly/1WCSzi4
[3] see also: Hotelling, Harold (1929). “Stability in Competition”. The Economic Journal 39: 41–57.
[4] The Bryce’s Law describes the tendency of a federal state to become more centralized over time. See: H. Badinger / V. Nitsch (Ed.) (2016): Routledge Handbook of the Economics of European Integration.
[5] L. Olai, L. Lehmkuhl (2012): Centralizing the EU? An analysis of the European Court of Justice’s tendency to rule in favor of centralization of the European Union.
[6] Dixit & Weibull (2007) present in their book “Political polarization” a model highlighting that different priors are necessasry conditions for polarization. In the context of this article here, filter bubbles and selective media engender heterogeneous priors amongst the population. Then, when a population is presented with the same information, people arrive at dichotomized conclusions, and politlical opinions shift toward extremes - and not toward the center.
[7] D. Landa / A. Meirowitz (2009): „Game Theory, Information, and Deliberative Democracy“, in American Journal of Political Science, Volume 53, Issue 2, pages 427–444.
[8] The Partido de la Red movement in Argentina makes digital democracy to the central element in its program. See https://www.ted.com/talks/pia_mancini_how_to_upgrade_democracy_for_the_internet_era?language=en
[9] Compare E. Ostrom (1990): Governing the Commons.
[10] Compare: James S. Fishkin/ Robert C. Luskin (1999): Bringing Deliberation To The Democratic Dialogue.
[11] Dryzek (1990): Discursive Democracy: Politics, Policy, and Political Science.
[12] D. Helbing, Society 4.0: Upgrading society, but how? https://www.researchgate.net/publication/304352735; D. Helbing, Why we need democracy 2.0 and capitalism 2.0 to survive, Jusletter IT (May 25, 2016), see http://bit.ly/1O5axWZ