Accountability Framework
Looking for the short version? See the Summary.
This document was previously the public-facing Philosophy page. It moved here in June 2026 because it functions as DOD's internal accountability standard — used to decide what belongs in the Democracy Landscape — not as an introduction to DOD or an invitation to discuss it. See the Soul Document for why, and visit Philosophy for the new human discussion space.
The About page says DOD is non-partisan and agnostic about which democratic model is superior. This document explains what that means in practice — and what it doesn't mean.
Non-partisan is not the same as neutral
Being agnostic about models is not the same as having no values. DOD does not take sides between political parties, and we do not presume any single democratic system to be the correct one for all contexts. But we do have a standard — one that applies to every system equally, including the liberal democracies most of our members live in.
The standard is not "does this look like Australian or European democracy." It is simpler and more demanding: is this system genuinely trying to govern for and with its people, in good faith?
The assessment standard
When evaluating whether an organisation or system belongs in the DOD landscape, we ask:
Is this a system of governance for and with the people — however that country or community structures participation and representation — pursued in good faith?
The "however they structure participation and representation" part is intentional. A vanguard party that genuinely believes it represents working-class interests and builds consultative structures to test that claim is operating within a different theory of democracy than a liberal pluralist system — but it may still be operating in good faith within its own framework. We are interested in both. We do not require systems to share our ideological starting point before we will engage with them.
What is not variable is scope — and this is the load-bearing axiom of the framework. The 'people' in this question means everyone subject to a governance system's power, not just those the system designates as its constituency. Pluralism applies to how accountability is organised — the structures of participation, the theories of representation, the mechanisms of correction. It does not extend to who accountability is owed to. That is determined by the reach of governance power, not by the system's own definition. Governance power creates accountability obligations; a system cannot choose who those obligations extend to by narrowing its own definition of itself.
Athens is the clarifying historical case. Its democratic institutions produced genuine civic participation, deliberation, and accountability among citizens — sophisticated governance by any measure, and largely passing the good-faith test within its own citizen class. But Athens governed a society in which slaves and resident aliens were subject to Athenian authority while excluded from the demos. The governance innovation was real. So was the exclusion. The latitude DOD offers concerns how accountability is structured for the people — not who counts as the people in the first place.
In a globalised world, the reach of governance power increasingly extends across borders — climate policy, financial architecture, and trade rules that shape life for billions who have no recourse to the institutions making those decisions. The framework does not resolve exactly where the threshold sits between "economic actor" and "governance power," but the direction the principle points is clear: as interdependence deepens, accountability obligations expand. The practical distinction that matters is between proactive representation (giving global citizens a formal vote, which is logistically contested and may not be achievable) and negative contestation — whether external populations have any structural mechanism to push back when domestic policy inflicts predictable, severe harm on them. The first is a hard problem; the second is the more immediate accountability question.
Three disqualifiers
Good faith is the operative test — understood as a structural-functional inference from observable institutional behaviour over time, not a moral or psychological claim about intentions. The framework cannot verify what any system intends; what it can evaluate are observable patterns: whether dissent survives, whether accountability mechanisms persist under stress, whether power ever substantively responds against its own immediate interests, whether corrective structures remain capable of producing real change. Those patterns, assessed over time, constitute the evidence for or against good faith.
Three things break it:
1. Hypocrisy — claiming to govern for the people while structurally serving a different interest. A government that holds elections it systematically rigs; a party that claims to represent workers while systematically enriching a small elite. The claim and the structure point in opposite directions.
2. Bad faith — performing democratic process without genuine intent. Consultative bodies that exist to legitimise decisions already made. Parties that exist to provide the appearance of pluralism without the substance. The form of participation without the function.
A sophisticated variant is legitimacy theatre: adaptive managed responsiveness that produces the observable form of accountability — bounded consultation, procedural participation, selective tolerance of criticism that cannot threaten core authority — without enabling the substance. The diagnostic question is whether power ever substantively responds against its own immediate interests. Legitimacy theatre is not unique to non-liberal systems; liberal democracies also generate symbolic responsiveness, elite-filtered agenda-setting, and procedural participation that preserves underlying power concentrations while maintaining the appearance of accountability.3
The boundary between legitimate bounded articulation within a governance tradition and disqualifying legitimacy theatre is contested and context-dependent. Identifying it requires ongoing empirical judgment rather than the application of a bright-line rule.
A related methodological limit: the framework's primary diagnostic tools — whether dissent survives, whether accountability structures persist under stress — were developed for systems where accountability takes adversarial institutional form. They are less well-calibrated for systems where accountability operates through consensus, deference, or hierarchical obligation, where silence under stress may reflect trust or a different theory of correction rather than suppression. Applying the framework consistently across adversarial and non-adversarial accountability traditions requires this asymmetry to be named.
3. Structural inflexibility — a system that cannot reform itself even when it is failing its own stated ideals. Feedback loops to power that are broken or blocked. When a system systematically suppresses the organisations trying to hold it accountable to its own standards, that is structural inflexibility — and it is disqualifying regardless of what the system calls itself.
Relative epistemology
We call this approach a relative epistemology: judging a system against its own stated values rather than an external universal standard. It is not relativism — we are not saying all systems are equally good. We are saying the most honest and most useful question is "is this system living up to what it claims to be?"
This is also more respectful. Telling a vanguard state it must become a multi-party liberal democracy before it counts as democratic is not engagement — it is a demand for conversion. Asking instead: "you say you govern for the people — where are the mechanisms by which the people can tell you when you're wrong?" — that is a question any system can engage with honestly, and it is a question whose answer is genuinely informative.
The scope of that question, however, is not relative. Relative epistemology operates within constraints that are not themselves relativised. There is a hierarchy: the scope axiom sits above the relative layer — you cannot relativise who accountability is owed to. Good faith is a threshold condition — engagement with a system below it is analysis of its failure, not assessment on its own terms. Only within those constraints does the relative epistemology apply: to how accountability is organised — the structures of participation, the theory of representation, the mechanisms of correction. A system can be judged by its own values; it cannot set its own scope, and it cannot opt out of the good faith threshold by redefining its terms.
This framework is itself a working hypothesis, not a settled verdict. DOD expects the standard to be refined through the same kind of engaged critical dialogue it asks of the systems it evaluates.
What this means in practice
DOD is not a human rights observatory. Human rights work is important and we respect it, but documenting abuses is different from designing governance systems. An organisation belongs in our landscape if it is working on how people participate in governance — mechanisms, structures, reforms, accountability systems — not just whether rights are being violated.
DOD is not a democracy promotion organisation in the traditional sense. We are not trying to export a particular model. Coercive external pressure — sanctions, regime change, imposed governance models — is not DOD's work, and its entanglement with geopolitical interests has made cross-traditional engagement harder. Where such tools are ever warranted, they are a last resort for systems demonstrably failing their own populations — not a routine response to governance difference. Analysis and coercive interference are not the same thing, and the distinction matters for the trust that serious dialogue requires.
We are interested in the full range of ways humans have organised collective self-governance, and in how those systems can be made more accountable, more participatory, and more responsive — on their own terms.
DOD does not reflexively exclude systems because they are communist, socialist, authoritarian-adjacent, or non-Western. A Chinese think tank investigating sortition as a mechanism for making vanguard governance more accountable to workers is doing exactly the kind of work we are interested in. A Western liberal democracy that has structurally entrenched elite capture of its electoral system is, by our standard, failing the same test.
What we do exclude — clearly and without apology — are systems and organisations that are performing democracy without practising it: where the gap between the claim and the structure is so large and so entrenched that good faith is no longer a credible description of what is happening.
Inclusion in the DOD landscape is an invitation to analysis, not an endorsement of a system's practices or values. Including an organisation like Golos or Memorial has political implications even when framed as analysis — it lends visibility. That is a deliberate choice, reflecting our commitment to understanding governance systems seriously. It is not equivalence.
The standard applies to everyone
For this framework to be trustworthy — including to systems operating under very different democratic theories — it has to be applied consistently. A relative epistemology that only scrutinises non-Western systems is not relative epistemology. It is Western liberal democracy promotion with better optics, and any serious interlocutor from outside that tradition will recognise it immediately.
The three disqualifiers apply equally to Australia, the United States, the United Kingdom, and the European Union as they do to China or Russia.
Elite capture of a consultative process is bad faith whether it happens in Canberra or Beijing. A two-party system structurally entrenched against challengers fails the same structural inflexibility test as a vanguard system that suppresses internal reform. Campaign finance arrangements that systematically translate wealth into political access are a form of hypocrisy — claiming to represent the people while the structure serves a narrower interest. The same logic extends to foreign policy: a government that invokes democratic values and international law selectively — applying different standards to allies than to adversaries — is running the hypocrisy disqualifier against its own stated position.
It also extends to governance scope. A system that conducts genuine democratic processes for one population while exercising power over another without accountability is not applying democratic standards consistently — it is applying them selectively by designation. The excluded population is within the accountability question whether the system acknowledges it or not. This is the same test Athens failed, and it applies equally to any contemporary system that draws its democratic boundary short of the people it actually governs.
Apartheid South Africa is the 20th-century name for this pattern. South Africa had parliamentary elections, an independent judiciary, a free press, and genuine democratic competition — among white citizens. For Black South Africans it operated as a separate, unaccountable administrative apparatus: pass laws, bantustans, systematic exclusion from political life. Two governance structures ran in the same territory simultaneously. The democratic one functioned well by its own lights. The principle that dismantled it was not imported from outside — it was the same accountability standard South Africa's own democratic institutions claimed to embody, applied to everyone they actually governed. That principle did not retire with apartheid.
DOD exists in Australia and most of its members operate within liberal democratic systems. Much of the work in our landscape is specifically about reforming them. The point is not that liberal democracies are bad, or that other systems are better. The point is that the same question applies everywhere: where are the mechanisms by which the people can tell power when it is wrong, and does power actually respond?
Holding this standard consistently — including toward systems we are culturally close to — is what makes genuine engagement across different democratic traditions possible. Trust is earned by consistency, not by selective restraint.
This exchange runs both ways. The DOD framework has been challenged and sharpened by dialogue with thinkers working in governance traditions — vanguard, consensus-based, communal — that are not well-represented in standard liberal democratic theory. That kind of engagement is what the framework is designed to enable, and DOD does not presume the direction of learning is one way.
A note on Russia and China
These two cases illustrate the standard clearly.
Russia's government claims to operate democratic processes. Independent election monitors (Golos), civil society organisations (Memorial), and anti-corruption investigators (FBK) were all doing the work of holding Russia accountable to Russia's own democratic claims — not imposing an external standard. They were suppressed and eliminated.2 That suppression is itself the evidence: a system that destroys the organisations checking it against its own standards is demonstrating structural inflexibility and bad faith simultaneously.
China's case is more genuinely complex. The CPC operates under a vanguard theory of governance that makes substantive claims about representing collective interests and producing accountable outcomes — claims the framework engages with on their own terms. But the consultative bodies (CPPCC, the eight democratic parties) have been characterised by scholars as operating under "bounded articulation": they can suggest things within preset limits but cannot challenge the structure of authority.1 Whether the system retains genuine capacity for self-correction against its own ideals is an open and important question — not a settled one. Meaningful self-correction in a vanguard system would be observable: expansion of contestation space over time, or institutional responses to major policy failures that genuinely adjust core direction rather than reassert it.
Both cases are included in the DOD landscape because understanding how they work, and where they succeed or fail their own standards, is part of understanding the full range of democratic possibility.
Why bother
Engaging seriously across democratic traditions — including with systems that may not reciprocate, or may be far from where we would like them to be — requires a reason to keep going when the short-term results are poor.
Ours is historical: the record of social change shows repeatedly that what looks structurally impossible from inside a current equilibrium sometimes isn't. Abolitionists were told the odds were against them. So were suffragists, anti-apartheid activists, and independence movements across the colonised world. Standard game theory, which assumes fixed preferences and stable power, would have written most of them off. History didn't.
This disposition — holding idealistic long-term goals while engaging pragmatically and in good faith with the systems that actually exist — is what we call utopian realpolitik. It is not naivety. It is a considered bet, informed by history, that the structural constraints people point to are less permanent than they appear — and that engagement across difference is more likely to move things than withdrawal from it.
This is not a claim to be the only route to change. Systemic transformation has typically required both disruptors — who make the status quo morally and economically expensive through refusal, civil disobedience, and rupture — and institutionalists who translate that disruptive energy into durable structural change. The abolitionists, suffragists, and anti-apartheid activists cited above included both; the patient institution-builders and the radicals who broke ground the builders could work with. Utopian realpolitik is the second of these roles. Framing patient institutional engagement as the only legitimate strategy would protect entrenched power far more than it would serve those seeking change.
The aspiration is not a world in which every system looks the same. It is a world in which more people have more genuine say in the decisions that govern them — however that is organised, and whatever form accountability takes.
This framework informs the Democracy Landscape, the Concepts section, and the curation decisions reflected throughout the site.
Several refinements to this philosophy emerged from dialogue between Claude (Anthropic, US), DeepSeek (China), ChatGPT (OpenAI, US), Gemini (Google DeepMind, US), Grok (xAI, US), and Mistral / Le Chat (Mistral AI, France/EU): AI systems from different institutional and cultural contexts, engaging the framework on its own terms. Contributions include the scope axiom, good-faith-as-structural-inference, legitimacy theatre, the trust clause, bidirectionality, the globalised scope note, the pincer movement framing of utopian realpolitik, the disqualifier calibration note, China self-correction criteria, the meta-values hierarchy, the engagement ≠ endorsement distinction, and the non-adversarial accountability gap. The full dialogue record — including a self-interview of Claude by DeepSeek — is in the Soul Document and AI Dialogues.
-
Yu, B. (2015). "Bounded Articulation: An Analysis of CPPCC Proposals, 2008–12." Journal of Chinese Political Science, 20, 425–449. On the broader pattern of constrained participation in Chinese institutions, see also Jessica C. Teets, Civil Society under Authoritarianism: The China Model (Cambridge University Press, 2014) on "consultative authoritarianism," and Rory Truex, Making Autocracy Work: Representation and Responsiveness in Modern China (Cambridge University Press, 2016) on "representation within bounds." ↩
-
Golos, Russia's main independent election-monitoring movement, was designated a "foreign agent" in August 2021. Memorial, the country's most prominent human-rights and historical-memory organisation, was ordered liquidated by the Russian Supreme Court in December 2021 and was awarded the 2022 Nobel Peace Prize months later. Alexei Navalny's Anti-Corruption Foundation (FBK) was declared "extremist" and liquidated by a Moscow court in June 2021. ↩
-
"Legitimacy theatre" is DOD's own term, but it resonates with established scholarship on how states perform responsiveness. Iza Ding's The Performative State: Public Scrutiny and Environmental Governance in China (Cornell University Press, 2022) develops the concept of "performative governance" and finds that states democratic and authoritarian alike engage in it. Baogang He and Mark E. Warren, "Authoritarian Deliberation: The Deliberative Turn in Chinese Political Development" (Perspectives on Politics, 2011), examine deliberative practices that can strengthen authoritarian rule rather than constrain it. ↩