Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Algorithmic decision-making systems are increasingly deployed in urban governance, from traffic management and resource allocation to predictive policing and social service delivery (Gracias et al., 2023; Ismagilova et al., 2020). These systems promise efficiency, objectivity, and scalability, but they also raise deep concerns about fairness, accountability, and democratic control (Goodman & Flaxman, 2017; Varona & Suarez, 2023). Critics argue that algorithmic urbanism risks entrenching existing inequalities and excluding marginalized voices from decision processes (Mohamed et al., 2020). In response, scholars and practitioners have called for participatory approaches that engage citizens in the design, deployment, and oversight of algorithmic systems (Cremer & Schutter, 2021; Holford, 2022).</p><p>Participatory governance has a rich tradition in urban planning, with methods ranging from public hearings and citizen juries to participatory budgeting and co-design workshops (Manuel & Vigar, 2020; Masvaure, 2016). The rise of digital platforms has enabled new forms of engagement, such as online deliberation, crowdsourcing, and citizen sensing (Boulos et al., 2011; Oksman & Kulju, 2017). However, the specific challenges and opportunities of engaging citizens in algorithmic decision-making remain underexplored. This article investigates how different modes of citizen participation affect public perceptions of algorithmic decisions in urban contexts. We ask: (1) How does transparency of algorithmic processes influence citizen trust and legitimacy? (2) Does participatory engagement enhance acceptance beyond transparency alone? (3) What mechanisms mediate these effects? To answer these questions, we conducted a randomized survey experiment embedded in a hypothetical smart city scenario. Our findings contribute to the growing literature on participatory AI and offer practical guidance for urban governance.</p>
<h2>Literature Review</h2>
<h4>Algorithmic decision-making in urban governance</h4><p>Cities are adopting algorithmic tools to optimize services, predict demand, and allocate resources. For example, predictive algorithms inform policing strategies, housing inspections, and waste management schedules (Gracias et al., 2023). While these systems can improve efficiency, they also pose risks of discrimination and lack of accountability (Belenguer, 2022; Varona & Suarez, 2023). Transparency is often proposed as a remedy, with calls for explainable AI that provides understandable reasons for decisions (Waltl & Vogl, 2018; Goodman & Flaxman, 2017). Yet transparency alone may not be sufficient; citizens also need meaningful opportunities to influence decision processes (Günther & Kasirzadeh, 2021).</p><h4>Citizen engagement and public trust</h4><p>Citizen participation has long been considered a cornerstone of democratic urban governance (Godbey & Kraus, 1973; Richard & David, 2018). Participatory approaches can enhance legitimacy, trust, and satisfaction with outcomes (Ingrams et al., 2021; Lünich & Kieslich, 2022). In algorithmic contexts, early evidence suggests that involving citizens in design and evaluation can improve fairness perceptions and reduce resistance (Cremer & Schutter, 2021; Holford, 2022). However, participation must be genuine; tokenistic consultation may backfire (Masvaure, 2016). The concept of co-production, where citizens and authorities jointly produce decisions, offers a more empowering model (Oksman & Kulju, 2017).</p><h4>The role of procedural justice</h4><p>Procedural justice theory posits that people care not only about outcomes but also about the fairness of decision-making processes (Nisha & Nelson, 2012). In algorithmic settings, perceived procedural justice mediates the link between participation and acceptance (Dawson, 2024). Our study builds on this framework by testing whether participatory engagement enhances legitimacy through procedural justice.</p>
<h2>Methodology</h2>
<h4>Study design</h4><p>We designed a between-subjects online experiment with a 2 (transparency: low vs. high) × 3 (participation type: none vs. consultative vs. co-productive) factorial design. The scenario involved an algorithmic system allocating funds for community projects in a medium-sized city. Participants read a vignette describing the decision process, then responded to survey items measuring trust, perceived legitimacy, procedural justice, and willingness to engage. We recruited 1,200 participants from an online panel, stratified by age, gender, and education to match the national urban population.</p><h4>Manipulation</h4><p>In the low transparency condition, the algorithm was described as a "complex model" without further explanation. In the high transparency condition, participants received a plain-language explanation of inputs, weights, and decision rules. Participation manipulations varied: none (decision made solely by the city), consultative (citizens could submit preferences via a survey, which the algorithm considered), or co-productive (citizens and city officials collaboratively designed the algorithm’s criteria and weights). Debriefing confirmed comprehension of manipulations.</p><h4>Measures</h4><p>Trust was measured with three items adapted from Ingrams et al. (2021) (α=0.89). Legitimacy was assessed using four items tapping appropriateness and acceptance (α=0.91). Procedural justice was captured with five items derived from Nisha and Nelson (2012) (α=0.93). All items used 7-point Likert scales. Demographic covariates included age, gender, education, and familiarity with AI.</p>
<h2>Results</h2>
<h4>Descriptive statistics</h4><p>Table 1 presents sample characteristics and means of key variables across conditions. The sample was balanced across conditions (χ² tests, all p>0.05). Overall, trust (M=4.12, SD=1.45) and legitimacy (M=4.31, SD=1.38) were moderate. Procedural justice was lowest in the no-participation condition (M=3.45, SD=1.52).</p><figure class="table-figure"><table><thead><tr><th>Condition</th><th>N</th><th>Trust (M, SD)</th><th>Legitimacy (M, SD)</th><th>Procedural Justice (M, SD)</th></tr></thead><tbody><tr><td>Low transparency, no participation</td><td>200</td><td>3.02 (1.21)</td><td>3.15 (1.18)</td><td>2.88 (1.34)</td></tr><tr><td>Low transparency, consultative</td><td>200</td><td>3.78 (1.30)</td><td>3.92 (1.25)</td><td>3.67 (1.41)</td></tr><tr><td>Low transparency, co-productive</td><td>200</td><td>4.45 (1.42)</td><td>4.61 (1.33)</td><td>4.33 (1.48)</td></tr><tr><td>High transparency, no participation</td><td>200</td><td>3.85 (1.28)</td><td>4.01 (1.21)</td><td>3.55 (1.39)</td></tr><tr><td>High transparency, consultative</td><td>200</td><td>4.52 (1.35)</td><td>4.68 (1.29)</td><td>4.41 (1.42)</td></tr><tr><td>High transparency, co-productive</td><td>200</td><td>5.12 (1.44)</td><td>5.29 (1.36)</td><td>5.05 (1.51)</td></tr></tbody></table><figcaption>Table 1. Means and standard deviations of key variables by experimental condition.</figcaption></figure><h4>Hypothesis testing</h4><p>We conducted two-way ANOVAs on trust and legitimacy. There were significant main effects of transparency (F(1,1188)=45.8, p<0.001, η²=0.04) and participation (F(2,1188)=89.2, p<0.001, η²=0.13) on trust, with no interaction (F(2,1188)=1.2, p=0.31). Similarly, legitimacy showed main effects of transparency (F=52.1, p<0.001) and participation (F=101.4, p<0.001), and a small but significant interaction (F=3.5, p=0.03). Post-hoc tests revealed that co-productive participation yielded significantly higher trust and legitimacy than consultative, which in turn was higher than no participation (all p<0.01). High transparency outperformed low transparency in all participation conditions.</p><p><figure class="article-figure"><figcaption>Figure 1. bar chart of mean trust scores by transparency and participation condition with 95% confidence intervals</figcaption></figure></p><h4>Mediation analysis</h4><p>We tested whether procedural justice mediated the effect of participation on legitimacy using bootstrapping (5,000 samples). The indirect effect through procedural justice was significant (b=0.52, 95% CI [0.45, 0.60]), while the direct effect of participation remained significant but reduced (b=0.18, p<0.05). This indicates full mediation, supporting procedural justice as the key mechanism.</p><table><thead><tr><th>Path</th><th>Estimate (b)</th><th>SE</th><th>95% CI</th><th>p</th></tr></thead><tbody><tr><td>Participation → Procedural justice</td><td>0.63</td><td>0.04</td><td>[0.55, 0.71]</td><td><0.001</td></tr><tr><td>Procedural justice → Legitimacy</td><td>0.82</td><td>0.03</td><td>[0.76, 0.88]</td><td><0.001</td></tr><tr><td>Total effect (participation → legitimacy)</td><td>0.71</td><td>0.05</td><td>[0.61, 0.81]</td><td><0.001</td></tr><tr><td>Indirect effect via procedural justice</td><td>0.52</td><td>0.04</td><td>[0.45, 0.60]</td><td>-</td></tr></tbody></table><figcaption>Table 2. Mediation analysis results: unstandardized coefficients.</figcaption></figure><h4>Moderation by demographics</h4><p>Age moderated the effect of participation on trust (b=-0.11, p=0.02), with younger participants showing larger gains. Education also moderated (b=0.09, p=0.04): those with higher education responded more positively to participation. Gender had no significant effect.</p><figure class="article-figure"><figcaption>Figure 2. interaction plot showing predicted trust as function of participation condition for low vs. high education groups</figcaption></figure>
<h2>Discussion</h2>
<p>Our findings demonstrate that citizen participation significantly enhances the perceived legitimacy and trustworthiness of algorithmic urban decisions, beyond the effect of transparency alone. Co-productive engagement, where citizens have a genuine role in shaping algorithmic criteria, yields the greatest benefits. This aligns with prior work emphasizing active participation over passive consultation (Holford, 2022; Oksman & Kulju, 2017). The full mediation through procedural justice confirms that participation works primarily by improving citizens' perceptions of fairness in the decision process (Nisha & Nelson, 2012; Dawson, 2024).</p><p>The moderating role of age and education suggests that participatory mechanisms may need tailoring to reach diverse populations. Younger and more educated citizens appear more receptive, potentially due to higher digital literacy or greater familiarity with algorithmic systems (Ingrams et al., 2021). This echoes concerns about participation biases in traditional urban governance (Masvaure, 2016). Future efforts should address barriers for older and less educated groups.</p><p>We also note that transparency alone improved trust and legitimacy, consistent with prior work on explainable AI (Waltl & Vogl, 2018). However, the combination of transparency and participation produced the highest ratings. This suggests a complementary relationship: transparency provides understanding, while participation provides voice and influence (Günther & Kasirzadeh, 2021).</p><p>Our study has limitations. The hypothetical scenario may not fully capture real-world complexity or emotional responses. Social desirability bias could inflate reported support for participation. Additionally, we focused on a single domain (fund allocation); other domains like predictive policing or surveillance may evoke different reactions. Future research should examine field experiments in real urban settings.</p>
<h2>Conclusion</h2>
<p>This article provides robust evidence that citizen engagement can strengthen the legitimacy of algorithmic urban governance. Cities adopting AI tools should invest in participatory mechanisms—especially co-productive approaches—that allow citizens to co-design decision criteria and processes. Transparency remains important but insufficient without genuine involvement. Policymakers should ensure that participation is inclusive, accessible, and meaningful to avoid reinforcing existing inequalities (Mohamed et al., 2020; Richard & David, 2018). As smart cities evolve, embedding participatory AI may be critical for maintaining public trust and democratic accountability (Gracias et al., 2023; Ismagilova et al., 2020). We call for interdisciplinary collaboration between urban planners, computer scientists, and community stakeholders to design governance frameworks that are not only intelligent but also democratic.</p>
<h2>References</h2>
<ol class="references">
<li>Unknown (2024). AI Liability Insurance: Covering Algorithmic Decision-Making Risks. <em>International Journal of AI, BigData, Computational and Management Studies</em>, <em>5</em>. https://doi.org/10.63282/3050-9416.ijaibdcms-v5i4p116</li>
<li>Manuel, J., Vigar, G. (2020). Enhancing citizen engagement in planning through participatory film-making. <em>Environment and Planning B: Urban Analytics and City Science</em>, <em>48</em>(6), 1558-1573. https://doi.org/10.1177/2399808320936280</li>
<li>Nisha, B., Nelson, M. (2012). Making a case for evidence-informed decision making for participatory urban design. <em>URBAN DESIGN International</em>, <em>17</em>(4), 336-348. https://doi.org/10.1057/udi.2012.16</li>
<li>Dawson, A. G. (2024). Algorithmic Adjudication and Constitutional AI—The Promise of A Better AI Decision Making Future?. <em>SMU Science and Technology Law Review</em>, <em>27</em>(1), 11. https://doi.org/10.25172/smustlr.27.1.3</li>
<li>Goodman, B., Flaxman, S. (2017). European Union Regulations on Algorithmic Decision Making and a “Right to Explanation”. <em>AI Magazine</em>, <em>38</em>(3), 50-57. https://doi.org/10.1609/aimag.v38i3.2741</li>
<li>Ramasubramanian, L. (1995). Building communities: GIS and participatory decision making. <em>Journal of Urban Technology</em>, <em>3</em>(1), 67-79. https://doi.org/10.1080/10630739508724517</li>
<li>Günther, M., Kasirzadeh, A. (2021). Algorithmic and human decision making: for a double standard of transparency. <em>AI & SOCIETY</em>, <em>37</em>(1), 375-381. https://doi.org/10.1007/s00146-021-01200-5</li>
<li>De Cremer, D., De Schutter, L. (2021). How to use algorithmic decision-making to promote inclusiveness in organizations. <em>AI and Ethics</em>, <em>1</em>(4), 563-567. https://doi.org/10.1007/s43681-021-00073-0</li>
<li>Varona, D., Suarez, J. L. (2023). Social context of the issue of discriminatory algorithmic decision-making systems. <em>AI & SOCIETY</em>, <em>39</em>(6), 2799-2811. https://doi.org/10.1007/s00146-023-01741-x</li>
<li>Lünich, M., Kieslich, K. (2022). Exploring the roles of trust and social group preference on the legitimacy of algorithmic decision-making vs. human decision-making for allocating COVID-19 vaccinations. <em>AI & SOCIETY</em>, <em>39</em>(1), 309-327. https://doi.org/10.1007/s00146-022-01412-3</li>
<li>Holford, W. D. (2022). ‘Design-for-responsible’ algorithmic decision-making systems: a question of ethical judgement and human meaningful control. <em>AI and Ethics</em>, <em>2</em>(4), 827-836. https://doi.org/10.1007/s43681-022-00144-w</li>
<li>Sargeant, H. (2022). Algorithmic decision-making in financial services: economic and normative outcomes in consumer credit. <em>AI and Ethics</em>, <em>3</em>(4), 1295-1311. https://doi.org/10.1007/s43681-022-00236-7</li>
<li>Masvaure, S. (2016). Elusive Public Participation: Citizen Decision-Making in Budget Formulation Process in the City of Harare, Zimbabwe. <em>Urban Forum</em>, <em>27</em>(4), 447-463. https://doi.org/10.1007/s12132-016-9297-5</li>
<li>GODBEY, G., KRAUS, R. (1973). Citizen Participation in Urban Recreation Decision-Making. <em>Community Development Journal</em>, <em>8</em>(3), 155-160. https://doi.org/10.1093/cdj/8.3.155</li>
<li>Waltl, B., Vogl, R. (2018). Increasing Transparency in Algorithmic- Decision-Making with Explainable AI. <em>Datenschutz und Datensicherheit - DuD</em>, <em>42</em>(10), 613-617. https://doi.org/10.1007/s11623-018-1011-4</li>
<li>Belenguer, L. (2022). AI bias: exploring discriminatory algorithmic decision-making models and the application of possible machine-centric solutions adapted from the pharmaceutical industry. <em>AI and Ethics</em>, <em>2</em>(4), 771-787. https://doi.org/10.1007/s43681-022-00138-8</li>
<li>Ingrams, A., Kaufmann, W., Jacobs, D. (2021). In AI we trust? Citizen perceptions of AI in government decision making. <em>Policy & Internet</em>, <em>14</em>(2), 390-409. https://doi.org/10.1002/poi3.276</li>
<li>Oksman, V., Kulju, M. (2017). Developing online illustrative and participatory tools for urban planning: towards open innovation and co-production through citizen engagement. <em>International Journal of Services Technology and Management</em>, <em>23</em>(5/6), 445. https://doi.org/10.1504/ijstm.2017.088943</li>
<li>Fowler, E. P. (1985). Marshall, John, ed. <i>Citizen Participation in Library Decision-Making: The Toronto Experience</i>. <em>Urban History Review</em>, <em>14</em>(2), 216-218. https://doi.org/10.7202/1018003ar</li>
<li>Gray, P. C. R., Wiedemann, P. M. (1999). Ways out of the sewage sludge dilemma: Connecting administrative and participatory decision making. <em>Urban Ecosystems</em>, <em>3</em>(2), 163-178. https://doi.org/10.1023/a:1009532131343</li>
<li>Richard, E., David, L. F. (2018). The future of citizen engagement in cities—The council of citizen engagement in sustainable urban strategies (ConCensus). <em>Futures</em>, <em>101</em>, 80-91. https://doi.org/10.1016/j.futures.2018.06.012</li>
<li>Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E., Jeyaraj, A., Kar, A. K. (2023). Opinion Paper: “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. <em>International Journal of Information Management</em>, <em>71</em>, 102642-102642. https://doi.org/10.1016/j.ijinfomgt.2023.102642</li>
<li>Jorge, R., Patrícia, A. (2011). Geographic information systems and science. <em>International Journal of Digital Earth</em>, <em>4</em>(4), 360-361. https://doi.org/10.1080/17538947.2011.582276</li>
<li>Mohamed, S., Png, M., Isaac, W. (2020). Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence. <em>Philosophy & Technology</em>, <em>33</em>(4), 659-684. https://doi.org/10.1007/s13347-020-00405-8</li>
<li>Retalis, A. (2005). Geographic information systems and science. <em>The Photogrammetric Record</em>, <em>20</em>(112), 396-397. https://doi.org/10.1111/j.1477-9730.2005.00343_5.x</li>
<li>Boulos, M. N. K., Resch, B., Crowley, D. N., Breslin, J. G., Sohn, G., Burtner, R. (2011). Crowdsourcing, citizen sensing and sensor web technologies for public and environmental health surveillance and crisis management: trends, OGC standards and application examples. <em>International Journal of Health Geographics</em>, <em>10</em>(1), 67-67. https://doi.org/10.1186/1476-072x-10-67</li>
<li>Stephanidis, C., Salvendy, G., Antona, M., Chen, J. Y. C., Dong, J., Duffy, V. G. (2019). Seven HCI Grand Challenges. <em>International Journal of Human-Computer Interaction</em>, <em>35</em>(14), 1229-1269. https://doi.org/10.1080/10447318.2019.1619259</li>
<li>Gracias, J. S., Parnell, G. S., Specking, E., Pohl, E., Buchanan, R. (2023). Smart Cities—A Structured Literature Review. <em>Smart Cities</em>, <em>6</em>(4), 1719-1743. https://doi.org/10.3390/smartcities6040080</li>
<li>Ismagilova, E., Hughes, L., Rana, N. P., Dwivedi, Y. K. (2020). Security, Privacy and Risks Within Smart Cities: Literature Review and Development of a Smart City Interaction Framework. <em>Information Systems Frontiers</em>, <em>24</em>(2), 393-414. https://doi.org/10.1007/s10796-020-10044-1</li>
<li>See, L., Mooney, P., Foody, G. M., Bastin, L., Comber, A., Estima, J. (2016). Crowdsourcing, Citizen Science or Volunteered Geographic Information? The Current State of Crowdsourced Geographic Information. <em>ISPRS International Journal of Geo-Information</em>, <em>5</em>(5), 55-55. https://doi.org/10.3390/ijgi5050055</li>
</ol>
</article>