The need for a citizen-centered perspective
Citizens increasingly resolve disputes online, often without ever speaking to a person: from filing a complaint about a defective product on a platform to receiving an algorithmic decision on a parking fine. Online dispute resolution (ODR) mechanisms have grown to such a scale that they are no longer an alternative to courts, but the default route for many everyday disputes. This raises a fundamental question: how do citizens relate to the institutions providing justice in the digital era?
The human-centered justice movement insists that justice services are designed for, and with, those they serve. Even though empirical research on citizens’ experiences with digital justice and algorithmic decision-making is growing, the debate still relies heavily on assumptions rather than data. When such assumptions go untested, the risk is not only a flawed procedural design but also an erosion of public trust in the justice system. Procedural justice research has long shown that perceived fairness depends on meaningful citizen involvement. Yet, voice – the opportunity to be heard and to have a say in the process, one of the central notions of procedural justice – is particularly at risk of being sidelined in algorithmic systems.
Therefore, we adopt a citizen-centered perspective by treating access to justice as a lived and differentiated experience that takes different shapes for different people. This requires working at the intersection of law and social psychology, as well as using a mixed-methods approach: in a representative survey combined with focus groups drawn from the same group of survey participants, we asked Dutch citizens themselves how they think about digital and judicial dispute resolution.
In this blog, we share three important findings. First, citizens turn out to be more articulate about the trade-offs of digital justice than is often assumed. They appreciate the speed and simplicity of digital procedures, but they also expect their case to be taken seriously, not just processed quickly. Second, perceived procedural fairness (having a voice, a neutral decision-maker, careful consideration of one’s case) is consistently rated lower for digital dispute resolution than for traditional court proceedings, but that does not mean it matters less in digital settings. Third, citizens are not a homogeneous group: we identify three recurring archetypes, each with their own balance of trust, skepticism and pragmatism. Across all three, however, the human element remains central.
Cheaper, faster and more effective?
Digital mechanisms are typically promoted for lowering costs, increasing speed and improving scalability, but our findings paint a more nuanced picture. Participants made clear that effective access to justice is also shaped by trust, perceptions of fairness, and the value of human involvement.
Among legal scholars, regulators and policymakers there is a tendency to talk about citizens rather than with them, as if they were insufficiently informed to assess what serves their access to justice. Our data show the opposite. In the focus groups, participants were strikingly well-equipped to articulate the very concerns that dominate the academic and policy debate. They spoke about autonomy, wanting a say in the proceedings, the importance of being able to tell their own story, and the need for a context-sensitive approach. Many of them volunteered the idea that the appropriate route, digital or judicial, should depend on the nature and complexity of the case. When asked about AI in the work of judges, participants raised, in their own words, much of the checklist that international guidelines now use, such as the need for independence and impartiality; awareness of bias; privacy and security; and the importance of transparency and human oversight.
Digital dispute resolution v. judicial decision-making
In our survey, perceived procedural fairness is consistently rated higher for traditional judicial decision-making than for digital dispute resolution, but it is expected there too. The focus groups confirmed this: participants saw the value of a quick online procedure, but speed never outweighed the need for a careful, diligent assessment of their case. To bring this trade-off to the surface, we deliberately presented participants with two ends of the spectrum: fully digital, out-of-court dispute resolution and traditional judicial decision-making. This was not because we believe these are the only options, but because the contrast functions as a methodological lens.
Our findings challenge the assumption that citizens will readily set aside procedural safeguards in exchange for ease, and it resonates with experimental evidence of a “human–AI fairness gap”: people systematically perceive AI decisions as less fair than identical human ones. What our participants seem to be telling us, in essence, is something about what it means to be human in the automated state: justice that feels just depends on the recognition of citizens as people with their own stories, not as inputs to be resolved.
Three citizen archetypes
Moreover, citizens do not form a homogeneous group. They cannot all be reduced to the same experience, which is precisely why a citizen-centered perspective matters. Across the focus groups, three recurring archetypes emerged, each with a distinctive way of weighing the trade-offs:
The skeptic is critical of digital dispute resolution and strongly values human judgment and oversight. A typical skeptic worries about being squeezed into pre-set boxes: when a chatbot or web form leaves no room for the specifics of a case, the citizen has to make their problem fit categories that someone else has chosen. They miss the cues you pick up in a face-to-face exchange – body language, tone, whether the other side actually understands what you are saying – and they feel a person should be involved whenever a dispute is, in their words, about people, not just things.
The optimist sees digital dispute resolution as efficient, accessible and trustworthy, and frames it as the natural fit for our era. A typical optimist points to a positive online experience (for instance, settling a complaint about a faulty product quickly through a portal), and emphasizes user-friendliness, speed and lower cost. They also see digitalization as part of an evolution that suits modern life, especially for routine matters such as small consumer disputes or low-stakes administrative decisions.
The cautious participant emphasizes the importance of a human element above all in cases that are complex or have significant consequences. They accept that simple matters – a parking fine, a defective appliance, a small consumer complaint – can be resolved digitally, but draw a clear line for disputes that affect people’s lives more deeply: a divorce, the custody of children, a serious criminal case. In those situations, they say, the assessment is genuinely human work. Important to them is that even within digital procedures there should always be a route to a real, qualified person if you get stuck.
What unites these three archetypes is a wish for systems that combine the advantages of automation with meaningful human involvement where it counts.
Access to justice in an algorithmic society
Our findings open up a broader conversation about what access to justice can mean in a society where digital and algorithmic procedures are becoming the norm. We highlight three themes based on our data: the digital divide, the role of voice, and the difficulty of defining digital justice from the bottom up.
First, the digital divide which, across all three archetypes, our participants were acutely aware of. Concerns about unequal access, digital literacy and exclusion came up consistently in the focus groups. They worried about people who cannot navigate complaint portals, or who lack the language, connectivity or confidence to follow online procedures. Digitalization, despite its promise of accessibility, can deepen exclusion. It risks “engineering disadvantage”: embedding existing inequalities into system design and creating new vulnerabilities for those it purports to serve. What our findings add is that ordinary citizens, including those who personally benefit from digitalization, recognize this risk: they do not see digital justice as a self-evident answer to access barriers, and they are wary of solving such barriers for some while creating new ones for others.
Second, core insight from the procedural justice tradition is that people value having a voice in processes that affect them. Not only because it can influence the outcome, but because being heard signals respect and recognition of their agency. Our findings align with this insight: when citizens feel they have no real voice in the process, when they have no opportunity to be heard or insight into how a decision is reached, their trust in the system erodes. Traditional judicial decision-making can already be a black box, and digitalization risks deepening this opacity. Digital procedures may therefore need different procedural safeguards than offline ones. This is even more true for algorithmic decision-making, where participation, transparency and human contact all take on a different shape.
Third, the question of definition. hat does (access to) digital justice actually mean? Legal scholarship tends to approach it through predetermined categories and institutional routes, with all the path dependencies that come with established frameworks. The risk is that this starts from the legal system rather than from the people who navigate it. Our participants showed us they do not think in legal categories but in terms of what they need, what they trust, and what feels fair. These vary considerably with the type of problem, its complexity and what is at stake.
Access to digital justice cannot be defined from the top down or reduced to a question of efficiency It is a differentiated experience that demands attention to what makes justice both just and human. Defining it remains our commitment, but only with citizens as co-authors of that definition.
This blog is based on a manuscript in preparation for publication.
Elke Olthuis
