[Cause Exploration Prizes] Dynamic democracy to guard against authoritarian lock-in

This anonymous essay was submitted to Open Philanthropy’s Cause Exploration Prizes contest and published with the author’s permission.

If you’re seeing this in summer 2022, we’ll be posting many submissions in a short period. If you want to stop seeing them so often, apply a filter for the appropriate tag!

This document is intended to argue that the Open Philanthropy Foundation should start funding research and (highly specific) activism to further real-politik democracy, i.e. democratic institutions, practices and culture as a pragmatic tool to solve economic coordination systems and safely increase the power our society wields. We start with Caveats, disclosure, epistemic status, then give a TL;DR on top and then discuss Importance, Neglectedness and Tractability with specific funding examples and directions as a subsection. We close with Final words and a personal note by the authors.

Caveats, disclosure, epistemic status

  1. Certainty/​quality self assessment: We learned about this call last week, so this should be taken as an honest attempt on inducing further, more serious investigation of this cause area, not a strong argument in favor of it.

  2. Due to the sparsity of data and difficulty in making quantified statements involved, most of our arguments will be qualitative or highly prior informed, although we do try to source quantified data whenever possible and add back of the envelope quantifications. We find it highly dubious to try and estimate something like DALYs and effect interventions when we talk about politics without better attribution studies and exploitation of methods like those pioneered by Card, Angrist and Imbens to estimate causal relationships

  3. More detailed and less rushed discussions on this subject have been had on the EA forum,

    1. Democracy Promotion as an EA cause area is basically a better sourced version of this document, albeit with less focus on dynamic democracy and avoiding lock-in

    2. Towards a longtermist framework for evaluating democracy-related interventions discuss the properties a democratic state should have from a longtermist perspective and links it to other considerations, but does not attempt to do an ITN analysis

    3. A bundle of related projects have been proposed or announced, including better democracy indices that focus on longtermist concerns , a (now apparently defunct?) Democracy Defense Fund was announced, as well as a longtermist-democracy focused research institute (still extant)

    4. Is Democracy a Fad? makes an argument that the economic conditions which enable democracy are rare and dictatorships are the default mode of societal organization

  4. 80k-hours, Rethink priorities and a few other EA/​longtermist organizations have furthering democracy as “possibly important, but not yet evaluated” in their cause profiles

Author disclosures:

  1. This is an argument for an inherently political cause area, coming from a political person. While we don’t like putting labels but in the interest of upfront disclosure, our politics can be thought of as somewhere in the anarcho-communist/​syndicalist corner, with sympathies for Bookchin, but materialist/​realist (until we find out working, efficient and *scalable* systems of organizing under anarchist principles, social democratic states are perfectly acceptable structures) and not revolutionary (extremely pessimistic on the feasibility or actual good outcome).

  2. Authors background: ML/​game theory PhD students who’ve been thinking about politics, economics and ways to achieve some scalable, feasible approximation to anarchocommunism for about 8 years now. Experience in organizing (mainly through associations), small scale political actions (think tank and input into parliamentary calls) as well as founding startups and financial markets (trading).

  3. Authors are also German with Polish ancestry and so might be emotionally biased to take authoritarian lock-in as a higher threat than others due to history

TL;DR on top

Democracy is a social structure which, if it can be implemented efficiently, achieves multiple desirable outcomes:

1. It solves a coordination problem, unlocking value which would be lost to the “Price of Anarchy” [11] if participants were forced to rely on the Nash equilibrium outcome;

2. To degrees varying between simple majority rule to consensus democracies, it lends a degree of moral legitimacy to these decisions;

3. Pragmatically, it allows for dynamic stability by ensuring regime change and adaptability can happen at low systemic cost;

4. The deliberative process allows information transmission and aggregation which can make the whole system perform better.

For sources on these claims, we refer to [7], Democracy Promotion as an EA cause area and Towards a longtermist framework for evaluating democracy-related interventions.

Despite these apparent advantages, global democracy as measured by [2] has been in decline for 16 years. While Freedomhouse, the authors of [2] are not without criticism, Is Democracy a Fad? and [19] give potential arguments for why this might be the case.

Because of the potential value unlocked by well functioning democratic systems and the suffering and stagnation that can come with authoritarian regimes, we argue that OPF should start funding the following:

1. Initiatives and policy work to measure, fortify and expand trust in democratic practices in the developed nations, as well as to shift power towards democratic political entities in particular (e.g. cooperatives, public-but-not-state entities like the German statutory health insurance [20]) . Particular focus can be paid towards these with a longtermist perspective.

2. Stabilizing and supporting democratic movements and institutions and self-governance in developing and emerging nations.

3. Developing a theory and practice of pro-democratic regime change, i.e. working towards a playbook for different actors (individuals, entrepreneurs, NGOs, parties, institutional actors) that helps them take effective pro-democratic action in different regimes (strengthening existing democracy, moving towards full democracy in partial democracies, choosing your effectiveness-risk trade-off in illiberal democracies and autocracies).

We start by arguing for importance, then discuss neglectedness and conclude by sketching some concrete exemplary directions for this cause area which illustrate tractability.


The role and state of democracy

Avoiding authoritarian lock-in has been already identified as a longtermist cause area in the Precipice [1] but given comparatively little attention in terms of page count and discussion (5.5 pages out of 280 main body). In the meantime, according to Freedomhouse, the world has experienced 16 years of democratic backsliding as of 2021 [2]. Writing this proposal in August 2022, the Russian invasion of Ukraine has provided cutting evidence that this is not simply the world splitting into an affluent, democratic, safe “western” and an economically rising or poor, but in any case authoritarian “non-western” block which do not interact. Instead, predictions of a shift in terms of the power balance between democratic and non-democratic nations as those raised in 2019 by [3] might have been more prescient and urgent than expected.

Similarly, the Uighur genocide [4] has made clear that economic liberalization does not always cause political liberalization, decades after this has been pointed out by scholars [5], partially because of the Chinese state deliberately adopting a liberalizing posture to aid the growth of both it’s economical and political might [6]. This gels with research by [7] which have presented evidence that:

1. It might be democratic institutions and political liberalization which drive economic growth and liberalization and not vice versa.

2. Increasing a country’s economical liberalization and wealth does not lead to political liberalization, or necessarily an increase in the standard of living of the broad population.

Chinas Uyghurs constitute 12.8 million people, Russia’s war in Ukraine affects over 42 million people and, if one only counts the population of China, North Korea, and Russia, it already amounts to over 1.5 billion people directly living under authoritarian regimes. As a very crude, back of the envelope estimation of what this might translate, the analysis done in https://​​ourworldindata.org/​​democracy-health indicates that moving from 0.5 to 0.7 on the liberal democracy index moved the 2019 life expectancy from no more than 60 to at least 70, i.e. 10 years/​​0.2. Extremely simplifying, if we assume at least 1 year per 0.2 points increases would hold across all levels below 0.7, moving China and India alone (1.402 and 1.38 billion with 2020 numbers=2.782 billion people) by 0.1 points could bring 1.3 billion years of additional life. Of course, China already has an extremely high life expectancy despite it’s low score, while Indian obstacles to higher life expectancy might not be due to public goods alone, but since we are also neglecting other benefits of more efficient and equitable political systems it still seems to be a number worth keeping in mind.

The potential “market” or these comparatively low-democracy index nations is large. Using Freedomhouse’s Free, Partially Free and Not Free categories[1], only about 80 countries are Free, about 60 qualify as Partially Free and the remainder is Not Free. But even this underestimates the true impact, since the free countries include countries like Tunisia, which is still struggling with the genesis of its democracy during the Arabic Spring; Malta, in which the murder of investigative journalist Daphne Caruana Galizias mars the nations liberal credentials [8]; and the world’s still-hegemon itself, the US, which experienced an attack on its nation’s capital in the process of peaceful transition of power in 2021 [9] and continues suffer intense political polarization and tension similar to large chunks of the western world.

While we might want to be careful with ascribing a crisis of democracy[21], one should not forget that democratic institutions as westerners think of them – full franchise of all ethnic groups, of women and of people of all social classes – are between 100 and 200 years old. Generously discounting the full emancipation of gay, trans and racialised people in countries like the US, Germany and others, we arrive at at most 50 or 70 years of true democracy. We should similarly remind ourselves that the rise of Hitler, Mussolini, Stalin, etc. happened less than a century ago, fascist regimes persisted in Western Europe till the 70s (Spain), and there are still stable ultra-authoritarian regimes such as North Korea, Saudi Arabia and the various warlords of central and southern Africa extant, all of which (together with states like Russia and China) have a vested interest in undermining democratic processes and culture. One could go on discussing India (struggling with its Hindu-nationalist party firmly locked in power) or Brazil’s Jair Bolsonaro, who at time of writing follows the anti-democratic playbook of casting doubt on election systems in the face of unpopularity [10], but all of this is to say that, depending on how strictly one counts, the affected population of democratic backsliding, illiberal democracies and lack of political participation is between hundreds of millions at minimum and every single human on this planet.


One might argue that democracy cannot possibly be neglected with multiple large and powerful nation states having a democratic identity, including the still-hegemon of the USA. On top of this, countless NGOs like Freedomhouse and foundations like the Open Society Foundation are already working on this.

Our counterarguments to these are centered around two key concepts: necessary partisanship and the stability trap.

Inherent and necessary partisanship

By necessity, working towards or researching democratic realities is a political endeavor, which inherently means that everyone involved (including the authors) are partisans. This is, in fact, the point of the democratic framework: that, by involving all stakeholders in a process which they can trust to give them a certain minimum of power and ideally find a all-sides-win solution, we can solve the coordination problem and avoid paying the price of anarchy [11].

But this also means that, without a dedicated longtermist player at the table that advocates for democracy in the sense of avoiding democratic lock in, there will be no voice, no representation of the perspective that no matter what you think is “the good politics”, you might want to make sure that any viewpoint compatible with democracy should still remain powerful enough to influence things, i.e. avoiding a lock-in of any specific ideology that risks becoming authoritarian.

This is an extremely hard needle to thread when there are also literal neonazis and anti-democratic populists in the political system trying to exploit the openness of the system, which maybe should be excluded and deplatformed. However, this risks giving up that exact openness that one swore to protect.

Walking this balance and making judicious decisions, heck, properly researching this is difficult even if the available funding and/​or political power base is exclusively interested in increasing political participation and democratic franchise. When, on top of that, the funding comes from political sources, corporations which might or might not be attempting regulatory capture, or “good billionaires” trying to put their stamp (and thus their personal politics) on the landscape, we argue that it becomes nigh impossible. The democratic institutions themselves are likewise constituted of flawed human beings and might suffer from the Iron Law of Bureaucracy, i.e. be more interested in their own persistence and political entrenchment than actually preserving a dynamic, changing democracy which might do away with them.

Therefore, because of the inherent and even necessary partisanship, we argue that there is as of right now few to no actors which can be categorized as longtermist-democratic, with the specific goal of avoiding authoritarian lock-in.

The stability trap

The second reason for neglectedness is a special case of the above, which we term the stability trap. It is, in short, the fact that humans and other stable systems naturally favor stability, which can lead a democratic system – whose whole purpose is to facilitate appropriate adaptation to the needs of even the smallest minority, peaceful transfer of power and safe change– to become non-adaptive, crustified and locked in by choice, as well as those who are striving to protect democracy to focus on stability instead of embracing necessary –potentially radical, but safe if done prudently– changes.

This trap is instantiated in examples like the aforementioned Iron Law of Bureaucracy, the lack of a timely response to the climate crisis due to multiple factors, the slow creep of unchecked police and executive power in the name of security [12], increases in regulatory burden to the point of decentralized regulatory capture, and many other examples where the need for immediate safety and stability slows down radical changes or prohibits risk taking that would lead to dynamic adaptation towards a real solution. This is, in a sense, also the need that the aforementioned populists exploit by promising radical change to go back to past greatness– stated differently, break down democratic cruft to take the radical actions to ensure that… nothing changes.

Strengthening democracy, in the longtermist sense, to avoid authoritarian lock-in will mean to strengthen democratic dynamism instead, to promote productive instability and to push for actions which undermine the power structure of institutions like themselves. Instantiations might be pushing for devolution into confederacies while maintaining top level cohesion for top-level concerns, mandates to reduce and simplify regulations periodically– adopting a term from programming, refactoring laws to keep them nimble and maybe even exploring different instantiations of democracy like the privatized regulation explored in [13].

Taken together, we argue that while there might be broad support of the general vibe of democracy across institutions, the specific goal of democracy to avoid authoritarian lock-in is severely neglected in relationship to its importance, in the same vein that two existing OPF focus areas, i.e., macro-economic stabilization and scientific research. We also note that the secondary effects of democratic governance are immense and touch not only on these two but basically all OPF cause areas.


How to strengthen dynamic democracy

Tractability in this area has a sharp divide between the observation and study of dynamic democracy and strengthening democracy in self-proclaimed democratic states both of which are highly tractable and in fact might be funded by OPF (e.g. as part of the macroeconomic stabilization package) and in taking steps towards increasing the dynamism of democracy outside of the societally standard norm, e.g. advocating for more dynamic voting systems in first-past-the-post nations, any democratic activism in authoritarian countries, or establishing strong democratic institutions in unstable or poor countries.

In the former, a relatively small amount of funding and operative support can go a long way, since these are generally non-monetizeable efforts. A possible funding opportunity would be historic and economic research in past and present highly democratic systems and institutions, e.g. aggregating and critically evaluating the various histories of anarchist libraries and trying to detangle the “usable” democratic dynamics from historical and ideological cruft. Another avenue in this road would be collecting a “democratic meme seed bank” from these sources to study ideas and systems which did not win the historic-evolutionary competition, but might still hold important concepts and inspiration for future instantiations of democracy, as well as simply the voices and perspectives of marginalized peoples, which can serve as a warning and reality check to the “winners of history” - the present peoples.

By studying alternative systems of organization like the Mondragon Corporation (scalable cooperative), the Haudenosaunee (ancient tribal democratic system) and getting a fact-based view of the AANES and Zapatista social experiments, one might be able to distill general practices into sections of a “playbook” for moving towards effective democracies. But there are also other candidates not centered on citizenship engagement and coming from one might call a “leftist” worldview. The ideas of moving towards a more privatized and liberalized law-as-technology view from [13] and the liberal-capitalist ideas of [14] would also qualify as entries into the “democratic meme seed bank”, since they have at their core the goal of political liberalization and making more efficient (and hence increasing the survival chances) of a democratic state.

Less conceptual and exotic and more immediately applicable funding opportunities might be expanding the OurWorldInData, FreedomHouse and other NGOs attempts in quantifying democracy to more fine-grained, dynamically updated data that can be used to enable coordination and activation, as well as providing (digital) infrastructure for pro-democratic movements aligned with the longtermist notion of dynamic democracy. Candidates might be platforms like transparency platforms https://​​www.nosdeputes.fr/​​ and https://​​fragdenstaat.de/​​, where marketing, international coordination, and support in lobbying for strong transparency laws can have long-lasting dividends, funding work on and pushing for the adoption and federation of tools like https://​​decidim.org/​​ and https://​​en.wikipedia.org/​​wiki/​​Liquid_democracy as well as the studies tracking the efficiency and real dynamics surrounding them.

Another branch, after collecting the “democratic meme seed bank”, would be research into formalizing game theoretic definitions of democracy extending the groundwork of [7] into algorithmic game theoretic notions. In particular, using analyses like [14], there is fruitful ground analyzing dynamic systems and in formalizing and developing notions of “good dynamics”. These are possible formal exit paths from the stability trap, allowing for instability that still attains a high degree of welfare for all participants of the system.

From this one foundational “sure wins” one could then branch into more difficult, ambitious and less assured goals like developing an applicable theory and practice of pro-democratic regime change. Here we must admit that we have no clear idea that is not rooted in our own political biases; however, we think the compilation of a “playbook” of strategies and tactics as is common amongst political radicals [16], [17] ought to also be compiled, publicized and importantly kept up to date and refined by those wanting to caution against authoritarian lock-in. This might include political, economic, technological, but also military/​insurgent knowledge as argued for by e.g. [18], and would have to take into account difficult trade-offs like maintaining an “arsenal of democracy” and a capable industry while avoiding regulatory capture by a military-industrial complex.

Final words and a personal note

The work is cut out for anyone attempting to tackle the area of dynamics democracy. However, we believe the tractability and impact profile of this field to be multimodel, with modes similar to that of the aforementioned macro-economic stabilization focus area, the “land reform” focus area, and the “future risk of AI” focus area and, frankly, dwarfing and subsuming all of these in importance and urgency. Because if we don’t manage to build robust systems against lock-in, there is no reason to expect that the fragile flower of democracy will vanish again into a stagnant stability of suffering for everyone but the elite, as had been the historical norm for millennia [19].


[1] T. Ord, The Precipice: Existential Risk and the Future of Humanity. Hachette UK, 2020.

[2] “The Global Expansion of Authoritarian Rule,” Freedom House. https://​​freedomhouse.org/​​report/​​freedom-world/​​2022/​​global-expansion-authoritarian-rule (accessed Aug. 10, 2022).

[3] V. K. Fouskas and B. Gökay, The Disintegration of Euro-Atlanticism and New Authoritarianism: Global Power-Shift. Cham: Springer International Publishing, 2019. doi: 10.1007/​978-3-319-96818-6.

[4] J. Smith Finley, “Why Scholars and Activists Increasingly Fear a Uyghur Genocide in Xinjiang,” J. Genocide Res., vol. 23, no. 3, pp. 348–370, Jul. 2021, doi: 10.1080/​14623528.2020.1848109.

[5] M. E. Gallagher, “‘Reform and Openness’: Why China’s Economic Reforms Have Delayed Democracy,” World Polit., vol. 54, no. 3, pp. 338–372, Apr. 2002, doi: 10.1353/​wp.2002.0009.

[6] Why China Sucks at Soft Power — China’s Reckoning (Part 4), (May 13, 2021). Accessed: Aug. 10, 2022. [Online Video]. Available: https://​​www.youtube.com/​​watch?v=y87R3Lp0jd0

[7] “The Logic of Political Survival,” MIT Press. https://​​mitpress.mit.edu/​​9780262524407/​​the-logic-of-political-survival/​​ (accessed Aug. 10, 2022).

[8] B. Taub, “Murder in Malta,” The New Yorker, Dec. 14, 2020. Accessed: Aug. 10, 2022. [Online]. Available: https://​​www.newyorker.com/​​magazine/​​2020/​​12/​​21/​​murder-in-malta

[9] “2021 United States Capitol attack,” Wikipedia. Aug. 10, 2022. Accessed: Aug. 10, 2022. [Online]. Available: https://​​en.wikipedia.org/​​w/​​index.php?title=2021_United_States_Capitol_attack&oldid=1103550712

[10] “Brésil: Jair Bolsonaro remet en cause le système électoral,” Le Temps, Jul. 19, 2022. Accessed: Aug. 10, 2022. [Online]. Available: https://​​www.letemps.ch/​​monde/​​bresil-jair-bolsonaro-remet-cause-systeme-electoral

[11] N. Nisan, Ed., Algorithmic game theory. Cambridge ; New York: Cambridge University Press, 2007.

[12] M. D. Bonner, “What democratic policing is … and is not,” Polic. Soc., vol. 30, no. 9, pp. 1044–1060, Oct. 2020, doi: 10.1080/​10439463.2019.1649405.

[13] G. K. Hadfield, Rules for a Flat World: Why Humans Invented Law and how to Reinvent it for a Complex Global Economy. Oxford University Press, 2017.

[14] Radical Markets. 2018. Accessed: Aug. 10, 2022. [Online]. Available: https://​​press.princeton.edu/​​books/​​hardcover/​​9780691177502/​​radical-markets

[15] M. Benaïm and M. Faure, “Stochastic approximation, cooperative dynamics and supermodular games,” Ann. Appl. Probab., vol. 22, no. 5, Oct. 2012, doi: 10.1214/​11-AAP816.

[16] The Alt-Right Playbook, (Oct. 11, 2017).

[17] “Guerrilla Warfare (book),” Wikipedia. Jul. 01, 2022. Accessed: Aug. 10, 2022. [Online]. Available: https://​​en.wikipedia.org/​​w/​​index.php?title=Guerrilla_Warfare_(book)&oldid=1095939783

[18] A. Industries, “Rebooting the Arsenal of Democracy: Anduril Mission Document,” Medium, Jun. 06, 2022. https://​​blog.anduril.com/​​rebooting-the-arsenal-of-democracy-anduril-mission-document-67fdbf442799 (accessed Aug. 10, 2022).

[19] “Capital in the Twenty-First Century — Thomas Piketty.” https://​​www.hup.harvard.edu/​​catalog.php?isbn=9780674430006 (accessed Aug. 10, 2022).

[20] Busse R, Blümel M, Knieps F, Bärnighausen T. Statutory health insurance in Germany: a health system shaped by 135 years of solidarity, self-governance, and competition. Lancet. 2017 Aug 26;390(10097):882-897. doi: 10.1016/​S0140-6736(17)31280-1. Epub 2017 Jul 3. PMID: 28684025.

[21] Merkel, Wolfgang. “Challenge or Crisis of Democracy.” In Democracy and Crisis: Challenges in Turbulent Times, edited by Wolfgang Merkel and Sascha Kneip, 1–28. Cham: Springer International Publishing, 2018. https://​​doi.org/​​10.1007/​​978-3-319-72559-8_1.

  1. ^

    Not ignoring the criticisms levied at FH, but using them due to lack of easily available other indices