Open government, civic tech and digital platforms in Latin America: A governance study of Montevideo's urban app ‘Por Mi Barrio’

Digital technologies have a recognised potential to build more efficient, credible, and innovative public institutions in Latin America. Despite progress, digital transformation in Latin American governments remains limited. In this work, we explore a peculiar yet largely understudied opportunity in the region: pursuing digital government transformation as a collaborative process between the government and civil society organisations. To do so, we draw from information systems research on digital government and platforms for development, complemented with governance theory from political science and conduct an interpretive in‐depth case study of an urban reporting platform in Montevideo called ‘Por Mi Barrio’. The study reveals three mutually reinforced orders of governance in the trajectory of the project and explain how the collaboration unfolded over time: (i) a technical decision to use open platform architectures; (ii) the negotiation of formal and informal rules to make the project thrive and (iii) a shared, long‐term ideology around the value of open technologies and technical sovereignty grounded in years of political history. Using a contextual explanation approach, our study helps to improve our understanding on the governance of collaborative digital government platforms in Latin America, with specific contributions to practice.


| INTRODUCTION
Governments in Latin America face high pressure to transform, with the challenge of balancing critical demands: to deliver better, more inclusive, and relevant public services for citizens and business, to become more transparent and responsive while reducing costs, increasing efficiency and reinventing their operations (ECLAC, 2021;OECD et al., 2020).Digital technologies, including platforms, data analytics and cloud computing, have often been portrayed as enablers of public sector modernization and as solutions to these challenges (OCED, 2019;World Bank, 2021).
However, the great expectations about digital-enabled transformation of governments remains limited (Benbunan-Fich et al., 2020), and Latin American countries are not the exception.The opportunity to transform an otherwise old-fashioned and over-bureaucratized logic has been linked to the pervasive platformization of services and opening-up governments to third-party innovators (Bonina & Eaton, 2020;Brown et al., 2014Brown et al., , 2017;;Fishenden & Thompson, 2013).
During the 2010s in Latin America a series of government-led innovation labs were launched in Uruguay, Argentina, Chile or Brazil to innovate public service delivery with new technologies, methods and open data (Acevedo & Dassen, 2016;Bonina & Eaton, 2020;Ferreira & Botero, 2020).In parallel, an expanding civil society played an increasing role as collaborator or even driver of digital transformation in public services, transparency, and innovation in the region (Bonina & Scrollini, 2020;Fumega & McNaughton, 2019).Civil society organisations in Latin America have tapped into open data, open platforms, and open-source software as bastions for socio-political change and government transformation (Carranza, 2017;Gutierrez, 2018;Scrollini, 2017).The salient role of civil society is peculiar to the region, and contrasts with trends elsewhere (Davies et al., 2019).We use the growing role of nongovernmental organisations (NGOs) as an empirical motivation to study digital transformation in the public sector in Latin America, and to add more generally to the otherwise dominant government-centric view that prevails in the field (Belanger & Carter, 2012;Benbunan-Fich et al., 2020;Cordella & Paletti, 2019;Fountain, 2001).
In this article, we position our research on open digital platforms for government services as distinctive digital technologies that can enable organisational change, and therefore government transformation (Avgerou & Bonina, 2020;Brown et al., 2014).Specifically, we are interested in understanding how the process of a joint NGOgovernment digital platform for public services was structured, negotiated and implemented in Uruguay.Because we are interested in the process of this collaboration, we develop a conceptual framework building on governance theory from political science (Kooiman, 2003(Kooiman, , 2008)), that we adapt and complement with information systems (IS) research concerning digital government, platforms, and platform governance (Belanger & Carter, 2012;Benbunan-Fich et al., 2020;Bonina et al., 2021;Cordella & Paletti, 2019).We therefore understand governance as a process that goes beyond state actors, to present a three-level model of interdependent orders, namely: the technical, institutional and meta-governance.In this framing, we investigate the following research question: how is a collaborative digital platform for public services governed in the city of Montevideo, Uruguay?
Empirically, we conduct an in-depth, interpretive case study of a digital platform called Por Mi Barrio in the city of Montevideo in Uruguay.Por Mi Barrio is a platform where citizens can report urban problems or complaints, started and driven by a local NGO and integrated in the government systems in mutual agreement.
Taking a contextual, sociotechnical explanation approach (Avgerou, 2019, p. 981), we use the three-level governance framework to organise parsimoniously and to communicate clearly the complex empirical phenomena that we study.We find that Por Mi Barrio was a collaborative effort sustained over time due to three mechanisms identified in the interdependent orders of governance.First, we find that at the technical level (first order), the open platform architecture chosen to integrate Por Mi Barrio into the government systems was central to make the collaboration possible in a short timeframe.The use of open standards and its modularity were critical to experiment and build incrementally on open government data without having to deal with an otherwise big process redesign.Second, we find that the presence of rules (formal and informal) combined with an incremental implementation, facilitated a process of collaboration of seemingly different actors with distinctive logics (the government and the NGO).We also find that despite the existence of these rules, the electoral timings and differences in managerial approaches brough frictions that jeopardised the possibilities for the initiative to scale.Third, our findings show that the existence of a history of shared ideologies towards open government, open technologies, and technology sovereignty, was a significant governance factor at the macro level that reinforced the design choices at the artefact level, and that contributed to navigate the frictions that emerged at the institutional level.
Our paper brings forward novel insights with a number of timely implications for IS theory and practice.First, we offer a theory-informed understanding to unfold a process of collaborative governance between multiple stakeholders in digital platforms ecosystems (Bonina et al., 2021), that underscores the mutually reinforcing roles of open technologies, institutions, and ideologies.In line with the interdisciplinary nature of the information systems, the theoretical lens we adapt from political science and extend with IS research highlights different levels of interaction among information technology, data, actors and the associated social, geographical and technological contexts (Struijk et al., 2022).Ours offer a novel, context-sensitive understanding of a complex governance phenomena, which extend otherwise platform-owner centric analysis of ecosystem governance in commercial settings (Ghazawneh & Henfridsson, 2012;Jacobides et al., 2018;Wareham et al., 2014).Second, our work emphasises the importance of politics and non-state actors for digital transformation, and therefore extends the literature of digital transformation and digital platforms in government beyond government-centric views (Brown et al., 2017;Cordella & Paletti, 2019).Third, we contribute empirically to expand the emerging scholarly studies of digital platforms for socio-economic development and public good in Latin America and the Global South (Bonina et al., 2021;Nicholson et al., 2021Nicholson et al., , 2022)).We offer peculiar nuances in the region valuable for policy formulation, including notions linked to technology sovereignty, which are used as policy drivers in many countries but understood and operationalised differently (Foster & Azmeh, 2020;March & Schieferdecker, 2021).Practically, we suggest that governments in the region should open up innovative channels to work with small organisations, including not only NGOs but also SMEs and social enterprises to foster digital transformation for good.We believe the lessons from this study are also relevant to other cities in the region, especially in countries like Argentina, Colombia, Chile, Costa Rica, Brazil, or Mexico that have a relatively robust civic technology ecosystem, comparable legal and normative orientations, and interest in open government.This paper is organised as follows.The next section offers the conceptual basis of the paper; it critically reviews the literature on digital platforms in government, platform ecosystems, NGOs in Latin America, and governance.It then presents a framework that guides the analysis.The paper follows with the empirical strategy and the details about the in-depth case study conducted.Section 4 presents the narrative of the case, Por Mi Barrio, followed by the case analysis.The last sections address the discussion and research contributions.Concluding remarks follow.

| REVIEW OF THE LITERATURE AND CONCEPTUAL FOUNDATIONS
Because we want to understand how a joint NGO-government digital platform solution was governed in the specific context of Latin America, in this section we build our conceptual lens to organise the empirical material parsimoniously and to clearly explain our findings.To do so, we review relevant work in IS research concerning digital platforms and governance.Since we take a contextual approach, we also review the peculiar characteristics of the context of Latin America and Uruguay that are relevant for our inquiry.We combine these insights that result in a three-level governance model that we build inspired in socio-political theory (Kooiman, 2003;Kersbergen & Waarden, 2004).

| Digital platforms as enablers of government transformation
Historically, in the multidisciplinary literature of e-government, information technology (IT) use in government has been associated with several types of benefits, including cost-savings, better planning decisions, transparency and responsiveness to the citizens' needs (Belanger & Carter, 2012;Benbunan-Fich et al., 2020;Kraemer & King, 1986).
Under the more recent label of digital government, digital technologies are widely seen as an enabler of far-reaching reform of the public sector (Avgerou & Bonina, 2020;Fishenden & Thompson, 2013), and as a catalyst of an upto-date innovation (Brown et al., 2014;Kankanhalli et al., 2017).The pressure to deliver reform and transformative effects are more profound in the discourses on digital transformation in government in the Global South (Avgerou & Bonina, 2020), with Latin America not being the exception.
Digital platforms and their ecosystems as well as their underlying data and open architectures have emerged as a promising avenue to realise the unserved transformational potential of digital government (Bonina & Eaton, 2020;Brown et al., 2017;Cordella & Paletti, 2019;Fishenden & Thompson, 2013).Digital platforms have their own characteristics that can enable transformation in public services.In this paper, we follow Bonina et al. (2021), who build on Cusumano et al. (2021), and define digital platforms as a "type of IT artifact with distinct properties" (p.3), which are part of "a socio-technical setting where they interact with an ecosystem of immediate stakeholders and a broader set of economic, organizational, institutional and spatial forces."(Bonina et al., 2021, p. 13).Research into digital platforms suggests three types: (i) transaction platforms enabling exchange of information or services between users or user groups; (ii) innovation platforms enabling innovation of apps and services and (iii) hybrid platforms, those that combine both transaction and innovation types (Bonina et al., 2021;Cusumano et al., 2021;Gawer, 2022).
We take an innovation platform perspective, and focus on the underlying properties of open architectures.Open platforms in government provide modules that can enable an almost infinite scope of applications.For example, open standards allow new services to be built combining data and functionalities through application programming interfaces (APIs) as illustrated in the literature of digital government (Brown et al., 2014;Cordella & Paletti, 2019;Fishenden & Thompson, 2013).Open government data platforms rely on open standards and make a specific illustration of digital-enabled transformation in government.Open government data platforms organise and make available large volumes of datasets and APIs for third-party innovators to build new public services.Transportation apps are common illustrations of city governments choosing to open their data instead of building in-house services for their local commuters (Bonina & Eaton, 2020).A digital platform service for urban complaints is another example, as the one we cover in this work.An innovation platform perspective is useful because, as we expand later, it allows us to focus on the properties of open standards and open data for government transformation as well as its governance features.

| Government platforms ecosystem in ICT4D: Collaborative governance and multiple stakeholders
Studies on digital platform for development (Bonina et al., 2021;Madon & Schoemaker, 2021;Masiero & Arvidsson, 2021;Nicholson et al., 2021) suggest that digital platform ecosystems encompass a broader set of actors, dynamics and relationships compared to those documented in commercial and Global North settings (Cusumano et al., 2021;Jacobides et al., 2018;Uzunca et al., 2022).These include traditional actors such as governments, enterprises and NGOs as well as emerging ones, such as partnership, networks, users and other hybrid organisations (Heeks, 2017;Jha et al., 2016;Qureshi et al., 2021).Often, relevant literature in IS and digital government literature have suggested that it is in the realm of the government to orchestrate solutions (Addo, 2022), achieving mixed results (Brown et al., 2017;Cordella & Paletti, 2019).While relevant, the orchestration lens is insufficient to unfold processes of governance that are not government-centric but a collaboration instead, like the one we study in this work.
There is also extant work in information systems research on platform governance.This focuses largely on commercial applications and the governance that the platform owner deploys on its immediate relations.The work of Ghazawneh and Henfridsson (2012), for instance, addresses how developers and their innovation activities are governed by the platform owner sitting at the centre of a wider digital platform ecosystem.Theirs and related studies (i.e.Wareham et al., 2014) offer an approach that theorises governance expressed on the tools and regulations (or boundary resources) that a digital platform owner may exert to cultivate service innovation in commercial industries.These studies, while useful, investigate how platform owners manage governance relations with third party complementors (Chen et al., 2022), and are therefore insufficient to investigate processes of governance and collaboration among a broader set of stakeholders, including non-commercial ones.
The multi-stakeholder nature of digital platform ecosystems in developmental settings points to the need of indigenous collaborative governance arrangements.Governance arrangements in multi-stakeholder settings are often challenging given the different capabilities, powers or opposite interests of relevant parties.Madon and Schoemaker (2021), for example, provide evidence on the clashes of value between seemingly different stakeholders in their study on a 'platformization' of digital identity services for refugees in Africa.Specifically, their study uncovers the tensions that emerge between the perceived value for the government and humanitarian organisations working on the services, in opposition to the perceived value for the refugees themselves and society at large.
Within IS more broadly, the studies of health information systems (HIS) in developing countries have looked at similar challenges from an institutional lens (Sahay et al., 2009;Sarker et al., 2019).HIS have historically been characterised as having multiple stakeholders with divergent interests and multiple systems adopted in uncoordinated ways (Braa et al., 2004;Nicholson et al., 2022;Senyoni, 2020).Substantial institutional work conducted in the field of HIS has confirmed the suitability of understanding politics, institutions, and the value of incremental approaches (Sahay & Walsham, 2006;Sahay et al., 2009;Senyoni, 2020), and the need to link micro processes with larger ones (Madon et al., 2007).We build on the relevant mediating role of institutions in the threegovernance framework that we present later.
While relevant, however, most empirical research on HIS has been largely conducted in African countries, India (Kimaro & Nhampossa, 2005;Madon et al., 2007;Sahay et al., 2009;Sahay & Walsham, 2006;Senyoni, 2020), and also the Eastern Mediterranean Region (Sahay et al., 2020).Our focused literature review found no studies of HIS in Latin America, which is a limitation to apply the highly contextualised insights of these studies.In addition, HIS contexts are well known for its challenges of coordination of multiple stakeholders in highly fragmented settings (Senyoni, 2020).Unlike HIS, civic technology applications including open government platforms like the one we study, are not characterised by the highly fragmented nature of health systems (Davies et al., 2019).This makes our work more manageable from a political science lens perspective.
Digital platforms shall not be understood as a panacea for development.Critical views have already pointed to detrimental effects, including exclusion, vulnerability of users, increased inequalities, or excessive surveillance (Bonina et al., 2021;Madon & Schoemaker, 2021;Masiero & Arvidsson, 2021;Pak et al., 2017).These critical perspectives are useful in our study to uncover specific tensions that emerge in governing a digital platform like Por Mi Barrio.

| The significance of Latin America: NGOs And technology sovereignty
In our work, we study a particular digital platform ecosystem-that of the government services.There, civil society has had a significant position.While the Latin American civil society played a vital role through the social movements of the 1980s in Latin America to defeat authoritarian regimes (Dagnino, 2003), in the last two decades, civil society organisations have focused on fostering local governance and innovation in democratic and political spaces.Latin American civil society is under pressure to articulate a varied range of global governance issues that affect their local environments, ranging from holding institutions to account and promoting transparency to delivering services to meet education, health, food and security needs and giving power to the marginalised (World Bank, 2021).Driving technological innovation has also gained space among civil society, to address voids not attended by governments or the private sector (Ingram, 2020).
In parallel, there has been a major rethinking of the digital ecosystem as one that is different and shaped by local idiosyncrasies.The Global South and Latin America have their specificities concerning the challenges of digitization and datafication processes from those emerging in Silicon Valley or the global North (Aguerre & Tarullo, 2021;Milan & Treré, 2019;Ricaurte, 2019).For example, notions of 'digital universalism' have been debunked by the experiences of appropriation of ICTs in countries such as Peru (Chan, 2014).The capability to design pathways that meet the needs of local populations regarding ICT uptake is part of a growing technological sovereignty approachthe latter understood as the autonomy of a national government or community to make their own, free decisions affecting citizens and businesses within the digital domain.Technology sovereignty therefore includes norms and regulations covering data, software, ICT infrastructure, and public services (March & Schieferdecker, 2021).Since 2013, civil society, local digital entrepreneurs, journalists, academics, and some government institutions, have worked together to accomplish a progressive agenda in the region connected with open data and open government initiatives (Scrollini, 2017).These experiences have been documented in cases such as Quito (Pozo Donoso, 2017), Bahia Blanca (Mirofsky & Bevilacqua, 2017) and Montevideo (Carranza, 2017).At the same time, open source, crowdsourced and open data initiatives from civil society have been made visible, using visual mapping tools, and applications to fight human rights violations, gender based violence and environmental injustice (Gutierrez, 2018;Hunt & Specht, 2019).While digitalization has expanded the agenda for NGOs globally there are different possibilities for these organisations in places such as the United States, or Northern Europe that tend to receive financial support from local and national donors and institutions.This is not the case for Latin American NGOs pursuing this agenda, as they rely mostly on regional and international donors funding (Aguerre & Tarullo, 2021), necessarily linking them to transnational processes.The growing datafication concerns, as well as government transparency, have been a paramount focus for civil society in the region, and in Uruguay specially.We now turn into our overarching framework.

| Theoretical framing of the study: Three orders of governance of digital platforms for public services
We choose a theoretical lens on governance from the field of political science as it has been invoked as a strong and vital link to better understand platform governance theory and practice (Gorwa, 2019).We select this lens as it allows us to unpack two fundamental issues of this study: a broader systemic view on the context in which platforms are developed, and the role of political power in the shaping of this process.Governance approaches for this study are significant as these theories generally invoke a process view, rather than a static analysis.In the political science field, governance theory is considered to provide a conceptual framework for characterising and analysing changes in how governments operate and interact with actors over time in the provision of goods and services with publicgood attributes (Bevir, 2011;Brousseau et al., 2012;Stoker, 2018).A multidisciplinary definition of governance implies a paradigm shift in the understanding of government where hierarchical configurations are transformed, to encompass changing social and institutional interdependencies that are established in the search for solutions to complex social problems among different actors.Fundamentally, a governance lens enables to address policies and politics beyond the state (Schuppert, 2015).
Whitin political science, Kooiman (2003) presents a relevant lens that we adopt in this paper to unfold how different stakeholders design, interact, and govern a digital platform for public services, i.e., the process quality of this experience and the evolution of the artefact.Kooiman (2003Kooiman ( , 2008) ) featured as a socio-political governance is a perspective to understand different moments, actors, ontologies and epistemologies of policies and politics.He develops a governance model of three orders.The first order is oriented towards problem solving and opportunity creation.It is the most grounded of the three and is directed towards the provision of public goods at their most operational level."First-order governance deals with day-to-day affairs.It takes place wherever people and their organizations interact in order to solve existing societal problems and to create new opportunities."(Kooiman, 2008).This governance level is related with the stocktaking and ordering of circumstances for those experiencing a problem; the identification of relevant interactions and their relations; the surfacing of the problem; The second order of governance is related to institution building, with the understanding that these arrangements are not made at an abstract level but are embedded in specific institutional settings that shape the interests and objectives of the actors and the norms, while adopting organisational modalities.Consistent with Kooiman (2003) and IS research, the notion of institution that we adopt in this work is that of intermediary agreements and rules that shape corporate and organisational arrangements (North, 1990;Powell & Di Maggio, 1991).The second order is therefore compatible with long standing work in information systems research, which suggests power battles (Bloomfield & Hayes, 2009;Danziger et al., 1982;Kraemer & King, 1986) and institutions (Fountain, 2001;Iannacci, 2010;Sahay et al., 2009;Sarker et al., 2019;Senyoni 2020) mediate digital government transformation, while they tend to explain the mismatch of expectations between set goals and actual results.Following Kooiman's conceptualization of this second order ( 2003) we operationalise this level by addressing the types of interactions between the actors, and the formal and informal rules that steer and guide these actions within organisational and temporal frames.
The third order corresponds to meta-governance (Kooiman, 2003), or systemic governance (Stoker, 2018) understood as the macro rules that bind together different constellations of actors and stakeholders, based on structural ideologies that guide the modes of interaction between them.This order "(…)is already subsumed in the first two(…)" (Kooiman, 2003,p. 2).The influence of high-level agreements and shared cognitive frameworks at the meta level, are at play in this stage.Building on IS research on ideologies and politics in digital government and computerization movements (Avgerou & Bonina, 2020;Barrett et al., 2013), we assume that digital platforms in government settings are embedded in intrinsically political dynamics.The third order, therefore, includes broader shared or contested ideologies in a particular context, which have a longer duration.More precisely, we follow Avgerou and Bonina (2020), which define ideology as "a shared system of beliefs, values, and norms that contribute to maintaining existing social structures, articulating resistance, or mobilizing action for change" (p.71).The peculiarities of the shared values in open data, or the views on technology sovereignty that may be present in Latin America (Section 2.3) are especially relevant on this level.Figure 1 provides a visual representation of this framework.
F I G U R E 1 Three orders of governance to understand a collaborative digital platform for government services.Source: Authors.
All three orders of governance co-exist, and are dependent on each other (Kooiman, 2003(Kooiman, , 2008)).In Figure 2 we address these three orders-IT artefact, institutions, and ideologies, against the backdrop of a temporal frame.The temporal dimension is important to understand the stages of development of a sociotechnical project like the one we study empirically.The case involves different actors in a specific context, where the negotiation of synchronicity, and a shared sense of opportunity (Granqvist & Gustafsson, 2015) become relevant dimensions to factor into the analysis.

| RESEARCH CONTEXT AND METHODS
Because we aim to understand change and governance relations over time using a contextual explanation approach (Avgerou, 2019), we designed an interpretive in-depth case study of Por Mi Barrio's project.This strategy was chosen because of its appropriateness for asking in-depth how and why questions, in which no experimental control is possible (Myers, 2009;Yin, 2003).We followed an information-oriented selection of the case (Flyvbjerg, 2006), considering its significance and its background at the regional and international level (Peixoto & Fox, 2016).This makes Por Mi Barrio a suitable and relevant subject to derive very rich accounts on the phenomenon.

| Study setting: Montevideo, Uruguay
The open government movement has been unique in Latin America and Uruguay specifically (Fumega & McNaughton, 2019).The existence of a shared language, a community that since 2012 meets annually in a regional conference (Abrelatam/ConDatos), strong civil society networks, technical capacity, and the engagement from governments have all led to the growth of a distinctive Latin American open data movement with increasing focus on delivering digital transformation for societal changes (Fumega & McNaughton, 2019;Open Knowledge Foundation, 2017).Uruguay stands out as a leading country not only in the open data movement, but also in terms of digital agendas and digital government transformation.For years, it has embraced the use of IT as central to its economic and social development, being the country with the highest ratio of the US dollar/per capita exports of software in Latin America (Sotelo & Rizzi, 2021).Uruguay also comes as the second top performer Latin American country in the comprehensive United Nations' E-government Index (United Nations, 2020) and 26 globally.The Network Readiness Index (NRI), a global index that measures the degree of readiness of countries to exploit opportunities offered by IT, places Uruguay within the top 49 performers globally, and second in the region (NRI, 2020).
The case we study takes place in the city of Montevideo, the capital and largest city of Uruguay.With approximately 1.4 million inhabitants (about one-third of the country's total population), Montevideo has the highest position in the Human Development Index (HDI) in the country (OTU OPP, 2019), and is placed within the top three of Latin America (UN, 2020).Montevideo is also well-known for the open data networks and government initiatives that have been developing since the late 2000s (Bonina & Eaton, 2020).

| Data collection
Our research design included multiple and complementary data collection strategies (Langley, 1999(Langley, , 2009)), and covered a ten-year period of analysis.Following an interpretive approach (Walsham, 2006), our data collection can be grouped in four different phases, each with its own purpose (see Table 1).During the exploratory phase we conducted three initial exploratory interviews, lasting 90 minutes each, to identify common themes and critical issues.
The second phase of fieldwork (June-November 2017) corresponded with the intensive data gathering.An interview guide was developed for this phase informed by our insights gathered in the first phase.The interview guide covered four themes: (a) tracing the history of the initiative and key milestones of the project; (b) understanding how collaboration took place between the different actors involved, including technical, social and political; (c) understanding the value that Por Mi Barrio had for the different actors and their respective institutions, and how that perception changed over time and (d) comprehending the sources of friction around the project and their consequences on the implementation and uptake.We followed a snowball sample approach (Noy, 2008) to identify key informants in this phase, and conducted 20 in-depth interviews.Phase 2 also included an online, short survey of users of Por Mi Barrio in Montevideo designed and administered by the authors. 1 The survey was sent to all users of the platform in Montevideo, who met the criteria of having used the platform at least once (1160 users).The survey had a participation rate of 27% and 316 unique responses during the third trimester of 2017.The survey contained 15 short questions that covered levels of satisfaction with the service, areas of improvement and whether users would recommend it.
Phase 3 involved two workshops to discuss preliminary results of the research and to confirm relevant patterns.
The first one was conducted in the city of Buenos Aires with the participation of 24 invited specialists (November 2017).The second workshop in (December 2018) included 22 invited participants, including key stakeholders of the project (government, donors, and the NGO).The final phase included two additional interviews to validate analysis and write up.We gathered and reviewed extensive archival documents during all phases of data collection.Extensive observations in vivo were also collected in most phases (one of the authors participated in various regional open data meetings, held informal conversations, and recorded events during 2014, 2015, 2017 and 2018).Appendix A provides a detailed summary of the multiple data sources and Appendix B lists the affiliations of the 27 interviews conducted.

| Data analysis
To analyse the data, we started by constructing a detailed timeline of the main events, episodes, project features and contextual factors covering the period 2008-2020; we used it as a visual mapping tool to help uncover relevant relations, address the three-layered heuristic model and to inform the case narrative.In parallel, we followed coding techniques adopting an interpretive stance that privileged the worldview of participants (Klein & Myers, 1999;Walsham, 2006).We began data analysis by open coding relevant textual segments of fieldwork data on values, motivations, practises, struggles, technical features, rules, actors and their objectives and assumptions.We continued with more substantive coding to elicit the key insights to determine how, why and under what circumstances the T A B L E 1 Summary of data collection phases and main uses in the study.observed evidence happened (Avgerou, 2019).Informed by our theoretical lenses, we organised the materials aligned with the three governance orders (Table 2): technical, institutional, and meta-governance.
To analyse the first order, we examined how the different actors coalesced around the technical artefacts and underpinning characteristics of open government platforms.References to standards, technical architectures and modules were especially salient here as these projected the solutions envisaged by these actors to address the problem of urban data reporting through an open web-based platform.In the second order, we focused on the institutional elements that conditioned the different actors and teams involved in the project execution within the legal and political framing.For example, we investigated those attributes appearing in formal rules (i.e. as reported in formal agreements or legal documents) but also informal arrangements as reported in the interviews.The political and organisational dynamics between different sectors within the IM and the electoral timing are part of the institutional analysis (second order).As to the ideological elements, we tracked references to digital sovereignty or enduring political claims about the values of open technology at this level.Crucial for the analysis was understanding the relations between the different orders: how the technical, institutional, and ideological configurations emerged, interacted, and eventually changed over time.A final point was to make sense of our findings within existing literature (Eisenhardt, 1989).While described as stages, these steps were highly reflexive and involved an on-going process of going back and forth between empirical and theoretical concepts.The richness of the data and the multiplicity of sources that we drew upon were vital for this process and resulted in a complex interplay between observing perceived patterns and its theorising that we later present in the analysis and discussion.

| Main actors and roles
As a collaborative platform, four main actors were involved in the initiative (Table 3)  initiative.The fourth partner was the Neighbour's Ombudsman Office, an independent office set up in 2006 to attend complaints about companies and organisations.Because of its proximity with the citizens, they were included in the project as a strategic partner to deliver capacity building and engagement in the use and dissemination of the project.

| Implementing Por Mi Barrio
The origins of Por Mi Barrio can be traced back as early as 2012, when the newly created organisation DATA was formed and led the First Regional Conference on Open Data (Abrelatam).It was then, in a region that was beginning to embrace the open government movement, that the idea of applying FixMyStreet with funding from AVINA Foundation began to take shape.Back then, there was a strong need to prove to governments that there was a the funding of Por Mi Barrio that a formal binding agreement between DATA and the government had to be signed as it wanted to avoid yet another failed digital service for urban citizen complaints.This was not easy, even in a relatively small city and there was an 'urgency' to be met as with many projects relying on international funding: it was a process designed against the clock.(…).
(Donor, our interviews/ID#15) In 2013, a Cooperation Agreement between the IM and DATA was signed, that formally started the collaboration on the project's technical interphase.During the end of 2013 and early 2014, the technical assemblage demanded work and coordination between the IT teams from DATA and the government.The implementation implied a learning curve for both sides from a technical and a management perspective, but in general there were no serious frictions.In July of 2014 Por Mi Barrio was launched.
After its launch, the next phase of the project involved promoting the platform for a broader uptake among citizens.To do so, DATA had originally planned to work with the Ombudsman's office as a partner to boost citizen engagement, mainly through the use of social networks.While training workshops had been organised between 2014 and 2015, there was no willingness (Peixoto & Fox, 2016) to promote the initiative and by mid-2015 the workshops and talks had ended.Partly, this is explained by the fact that the Ombudsman in office that had agreed to partner with DATA left his position when the workshop and outreach phase started.Another barrier surfaced when in May 2015, municipal elections took place.Even though the same party stayed in office, the Mayor who originally signed the Agreement was no longer in its position and new teams took over various digital transformation policies.The continuity of the project was not questioned, which was a milestone for the project.The political transition, however, did affect the internal level of commitment to the initiative; its appeal declined as other priorities emerged, and the community engagement efforts, both internally in the different sectors and towards the different 'boroughs', were interrupted.

| Main results and outcomes of Por Mi Barrio
The experience of Por Mi Barrio is one that challenges the traditional notions of success of IT artefacts through adoption rates and numbers, to one that emphasises a broader understanding about the 'transformation' involving the main actors, their institutions and their collaboration strategies.Por Mi Barrio did not represent an increase in overall complaints about urban infrastructure, a concern that had been raised originally by maintenance sectors of the IM and retained a small and stable proportion of users over time.It peaked at 3% of complaints but was stabilised to 1%, despite the high levels of access to devices and connectivity in a city as Montevideo.It did not generate optout mechanisms from other channels of the IM either, as the telephone prevailed.
From the users' perspective, Por Mi Barrio was well received.Overall, 52% rated the experience of using Por Mi Barrio as 'Very Good' or 'Good'.On the other hand, 77% indicated that they would use the tool again, and 74% say they would recommend it to another neighbour.The areas for improvement that were identified showed some of the problems that surfaced in making the service work seamlessly.For example, users noted the lack of follow-up and accountability on the evolution of the report that was often not made visible on Por Mi Barrio.This was a source of frustration for both DATA and the IT team at the municipal government.The IT sector did not have authority over the maintenance sector of the IM to enforce the use of the system to report the progress.Users also reported that there was a vast part of the population that was unaware of the service.From our data, we infer this was owed to a lack of publicity given to Por Mi Barrio on the government channels.
One notable outcome of Por Mi Barrio was the national and international prominence that it achieved over time.
During 2014-2018, other countries tested the implementation of Por Mi Barrio (or similar adaptations).For example, Por Mi Barrio was brought to Costa Rica (municiaplities of Osa and Palmares), the capital city of the Province of Cordoba in Argentina, and to the municipality of Rivera in Uruguay.It received awards for its design, such as the 'Digital Cities' competition and was featured in by the World Bank as an exemplar of highly institutionalised response (Peixoto & Fox, 2016).
Por Mi Barrio also triggered organisational changes.In the case of the government of Montevideo, it resulted in the incorporation of an open platform interphase (Open311) into their existing systems, SUR, with two outcomes.The first is that it enabled future applications connected to SUR through this interphase, opening a door for further digital innovations.The other is that by implementing Por Mi Barrio, the government expanded their system SUR to all the service and maintenance areas, accelerating a process that would have otherwise taken years.Por Mi Barrio contributed to the Government of Montevideo to later develop its own digital service for citizen complaints, both as a web (in 2018) and as a mobile app (2019).The experience of Por Mi Barrio was regarded as the catalyst to building the new, strictly governmental web application.
Por Mi Barrio was also a source of transformation for DATA.It served to lay the groundwork for other collaborative public service projects that happened later, such as the multi-awarded health platform "A Tu Servicio" (Bonina & Scrollini, 2020).Barrio-SUR, which is the system of the municipality, but we actually created a standard protocol(…) What does that mean?It means that if tomorrow another organization comes and wants to join, the integration is already in place.
(IM, Technical leader IT, our interviews/ID#19) We were conceptually clear that we had to define an APIand that's where we found Open311 (IM, Technical leader IT, our interviews/ID#18) Our data reveal that both implementation teams-the government and the NGOs-understood (and shared) the benefits of the modular and open design at the government level, including its cost-effectiveness, quicker deployment, and ability to mix and match with other components.
(…) ours was free software, it was run by civil society, in short, like all the other pluses about this project, it was matched the expectations of the municipality (DATA, IT, our interviews/ID#2) Since PMB, the private sector has approached PMB to use the same service.The API Open 311 is available to the people of PMB and to anyone who needs it.
(IM, Technical leader IT, our interviews/ID#7) According to our interviews with members from the government, the API was an ideal artefact that allowed them to clearly define the roles and responsibilities of the two main actors involved in its deployment.
Operationalising an open platform architecture based on open data included a vision of technical transparency, but also one of institutional adaptation: As one of the IM's IT leaders expressed: If I promise you a fancy system to manage reports but it does not help to manage the problem… well, that is not useful to us.PMB helped us to see how to prioritize the users' concern and not the management per se (IM, Technical leader IT, our interviews/ID#03) Por Mi Barrio was designed as an open technology that through its modular, interoperable features enabled the NGO DATA to monitor the citizen's claims against government responsiveness.It also enabled the Municipal government to obtain participatory citizen data from outside, feeding it into its system (SUR) without compromising its workflow or informational integrity.
The first order of governance emerges in the negotiations around the definition of the problem and its solution through an open platform, the selected IT artefact by the donor, the NGO and the government in a shorter temporal time frame, since its implementation was achieved in less than a year.The sense of urgency was a moment of fundamental opportunity creation to materialise the platform.It was shared by the developers' teams and the higher management of DATA and the IM to accomplish the project, fulfil the commitment with the donor and grasp the political window for the outgoing Mayor.The creation of this API materialised a need to change a situation whereby citizen data reporting was not yet clearly developed as a possibility by the municipal government and where there was a deficit of civilsociety-government collaboration in these digital endeavours.

| Second order: Institutional rules and negotiations
The roles and responsibilities that were synchronised to make the IT artefact work refer to the second order and affect the understanding of how these arrangements unfolded in practice against the backdrop of institutional dynamics and actors' expectations.
Por Mi Barrio was an activator of many internal processes at the IM, though it is by no means the only factor.It is not only something that facilitated and moved things on the users' side, it moved things inside the IM, with its services and open data scheme, it also touched upon the role of civil society (DATA leader, our interviews, #ID05).
The implementation of Por Mi Barrio helped the municipality to define the responsibilities and what roles corresponded to the outside party, in this case DATA.This platform: (…) allowed [us] to define the rules of the game (IM, Technical leader IT, our interviews/ID#27) The institutional frameworks that made Por Mi Barrio an open data driven, collaborative platform, were not neutral or free from friction between the major institutional stakeholders.An empirical peculiarity of Por Mi Barrio was that, contrary to other civic technology platforms, there was no single 'platform owner'.This meant that there was an institutional 'experimentation' within the project, as new collaboration rules, formal and informal, had to be put in place for the main partners to work together.Even though DATA had a greater control of the digital artefact towards the public, for the platform to be responsive and to enable the high institutional commitment originally expected by the donor, the complaints were directly funnelled and integrated into the government of Montevideo's own reporting system SUR.The Municipality of Montevideo aimed to define these rules and responsibilities so that each party was focused on their share and there was no mingling nor confusion.Despite the 'co-creation' jargon in the civic tech environments, our interviews suggested that governmental actors were not prone to use this term, as it did not capture the different interests these actors expressed for the artefact.
We identified two types of arrangements to embed the rules into this project, formal and informal.A predominant formal rule is the Agreement between DATA and the Municipality of Montevideo.It served as a basis to set up the operational plan, and to build confidence between the parties: The most important thing we did in PMB was to have an agreement signed by the IM, which was a condition of AVINA to fund this It was an agreement that stated that we had access to the systems not only to enter reports but to then take those reports and give them back to the system and to the people.
( Agreement was also integrated into DATA's own institutional practises.The organisation subsequently used it as a template that was applied and adapted in other projects that they implemented with different governments.For the municipal government, the Agreement was particularly an asset for the IT department, since it empowered the technical sector to confront internal challenges emerging within the organisation.Most notably, it was a rule to help them overcome the temporary opposition to the project that emerged from other government offices (i.e. the maintenance sector), as well as from competing alternatives proposed by other third party organisations, including a private firm.
Our agreement with the IM was photocopied and the names were changed for other experiences within the IM.That amuses me a lot, the 'hippies of civil society', we made an agreement, a sustainable project three years ago and we opened the door to private for-profit entrepreneurs to do the same and it didn't work.The hippies are still alive and the entrepreneurs didn't work.
(DATA Leader, our interviews/ID#05) Though useful, the Agreement was no silver bullet, and other less formal arrangements between the main partners were practised.Not all parts of the government were on board working alongside an external partner, let alone being more 'exposed' to scrutiny and complaints by the citizens.Incrementalism was a vital strategy to quell these resistances: only a few service areas of the Municipality of Montevideo were included at the beginning, with an expansion plan to roll into others over the months.Incrementalism, as an informal agreement embraced by both project partners, enabled the project to be perceived less as a threat to specific service departments (in particular, cleaning services) that thought they might be subject to an 'avalanche' of complaints.The introduction of Por Mi Barrio allowed the IT sector of the municipality to extend this system to other areas previously not contained within SUR, accelerating the expansion of their own system.
By this I mean that, for example, lighting, which had its own complaints system, would gradually migrate to SUR.So, what we were doing with DATA was that every time we made a new category available in SUR, they prepared everything to enable the new item on the platform.
(IM, Technical leader IT, our interviews/ID#13) Additional collaboration tools and rules that shaped the outcome of the project were system audits, as well as best practises and consultancy reports that were undertaken in the civic tech environment during those years.These enriched the thought-process and decisions of DATA and the IT department of the Municipality of Montevideo for the preparation, launch and early phases of implementation.2 (…)we had several lessons learned but something that is still happening today is that the reports are badly categorized, right?So the municipality decided to put someone to check the category and make sure it is correct and print it out.
(DATA IT, our interviews/ID#2) Despite the definition of roles and responsibilities, the Agreement, and the incrementalism of the approach undertaken, these did not trigger the buy-in from other sectors in the municipal government.For example, our interviews suggested that the Communications Department of the municipality did not support the initiative beyond the purely contractual clauses.Most importantly, they did not place Por Mi Barrio as an official channel for citizen complaints on the government website, despite the official technical integration of the service.An additional institutional barrier that limited the outreach of the project to more users was that the government did not reach out officially to the different political leaders in the eight districts of the Municipality of Montevideo.Instead, the effort was left to communicate with you later, because nobody from the IM's Mayor Office has told us anything".
This lack of engagement left DATA exposed to attend different institutional fronts, internally with the IM but also with the Neighbour's Ombudsman Office, that did not play the role that DATA and the donor had originally envisaged.These difficulties point to challenges concerning institutional drivers, government inertia and the management of incentives.Raising the project's profile and expanding its user base would have benefited the escalation of the initiative.

| Third order: Sustained ideologies over time
The It is a project that seeks to make public management transparent, but at the same time it helps to generate collaboration between the state and society.(DATA co-founder, our interviews, #ID01).
(…) in the middle and upper management they understand that it (open data) is an interesting initiative, with its value, and at a certain moment it is formalized in departmental resolutions and that's where it starts to take shape.
(IM, Technical leader IT, our interviews/ID#18) Another example that emerges from our interviews is how the donor (AVINA) emphasised that it wanted to innovate by not repeating prior experiences where there was no accountability involved in the reporting solutions.
Devising a project that had the commitment and the participation of the government of Montevideo, with an active involvement at the time in international fora such as the Open Government Partnership was a means to connect different layers of trust sustained through ideological components.
'Technological sovereignty', as an ideological force, is another feature of this third order of governance.
When Por Mi Barrio was conceived, the debate around technology autonomy had spiked with the denunciation by Edward Snowden of global mass surveillance by the United States government through the PRISM system.In Latin America, many actions were taken both at regional and national level to cut off interdependence with U.S connectivity infrastructure.The discourse around technological sovereignty that was espoused by the actors in Por Mi Barrio was firstly tuned to the platform layers of data and protocols, but also, as an ideological drive to steer away from proprietary software.A case in point is that there was an international private company hoping to sell a web complaints solution to the Municipality.DATA used the argument of 'technological sovereignty' to finally convince the IM to adopt Por Mi Barrio and the integration into the SUR system and against a competitor from the private sector.Who would be better able to provide such a service but the own IM with its IT department comprising nearly 60 IT developers at the time of the launch, and DATA who had secured knowledge transfer from My Society as well as funding to develop its own learning process?
That was a tough resistance because a company would come and try to set up a shop, and that's when we had to fall back on civil society legitimacy… and technological sovereignty approach.
(Data co-founder, our interviews/ID#01) The technical sovereignty as an ideological feature was also about the development of capabilities-that is, developing and owning the necessary skills to operate, expand and adapt an information system: Sovereignty, as I understand it in this case is to be able to manage your own systems and data, and it is based on your capabilities.
(IM, Technical leader IT, our interviews/ID#27) These arguments were taken by the more ideologically aligned sectors in the IT department at the IM.It was not just a discursive feature: in Uruguay, the Law 19.179 mandates the adoption of open-source software for all public institutions (2013) in the acquisition of IT solutions for governments.The artefact itself of Por Mi Barrio was therefore aligned with the legal norms as well as with the ideologies of the developers and key policy leaders.

| DISCUSSION
This work has aimed to address how collaboration unfolds around a digital platform across sectors, organisations and through time.Our findings show three levels of governance that reinforce each other, but with their own dynamics and results.Por Mi Barrio navigated a process of collaborative governance that included compromises and realignments over time which were not exempt from friction and tensions.As per its outcomes, the NGO-driven digital platform triggered new ways of working in the government and contributed to spur their organisational changes.
More specifically, the collaborative work that Por Mi Barrio entailed prompted the later expansion of the Municipality's own SUR system to all other urban maintenance areas.Por Mi Barrio was also a catalyst of other local innovations: the initiative meant the first web interface and channel for urban citizen reporting in the municipality and promoted the launch of the municipality's own web interface.More substantially, Por Mi Barrio enabled the government to experiment with open standards and walk the talk validating these legal norms in concrete artefacts, as well as to expand it to other initiatives in the municipality.In what follows, we discuss our findings considering our proposed framework and existing work.development with relative smoothness and with a shared sense of opportunity (Kooiman, 2003) for the main actors involved.
The ideology around open government was a differential enabler to make the initiative thrive compared to other digital projects in Latin America.The findings are aligned with previous work on digital transformation in government and its surrounding politics that challenges the rationality behind digital transformation in government (Addo & Avgerou, 2021;Avgerou & Bonina, 2020;Danziger et al., 1982).Avgerou and Bonina (2020) show that in Mexico, a solid, shared and sustained government ideology on free trade at national level was bigger and stronger to sustain a digital innovation programme and to overcome changes in government teams.Like theirs, our findings show that a common and sustained ideology towards open technologies and open government were instrumental to operationalise a collaborative project with external stakeholders and that triggered a high institutional response (Peixoto & Fox, 2016).
The shared ideologies, in the form of a meta-governance, have also been studied in related domains, such as the open-source movement (Barrett et al., 2013;Stewart & Gosain, 2006).The latter includes a collaborative, less hierarchical and anti-mainstream approach to contemporary developments compared to proprietary modes of knowledge production.Similar to those studies, ours show that the crucial factor facilitating collaboration was the commitment of an NGO to transparency as its primary value, rather than profit as the main driving force.Por Mi Barrio shows a trajectory that is rooted in decades of technological feats addressing concerns on digital sovereignty (Sabiguero et al., 2016) and which could only be materialised when there was trust and a vision of opportunity that made these efforts coalesce into the IT artefact.
The predominant role of ideologies in the case at the meta-governance level suggests a recursive relation with the other governance orders.Our findings suggest that the longstanding shared ideologies conditioned how the initiative was implemented.At the same time, these ideologies projected a preference for the future.Finally, while the case shows that there was ideological convergence among the key actors of Por Mi Barrio, there were differences in the responsiveness towards the initiative at different levels of the government.Despite the powerful role of ideologies, the institutional level of governance challenged the platform expansion and failed to support the consolidation of network effects.

| Governance norms and political battles
Our case showed the relevance of holding a normative document (the agreement/memo) to secure a political compromise for the project to go ahead which was validated by the ideological convergence of the different actors involved.The formal agreement was a game changer for the donor to invest in the project.This is significant in Latin America because often, these formal compromises are difficult to craft and to enforce.The formal agreement, as a rule, acted in the case as an organising and binding force that helped minimise transaction costs (Ambrozini & Martinelli, 2017).As a vital governance instrument (Malhotra & Lumineau, 2011), this legally binding document assisted in reducing friction and in accelerating the technical work undertaken by teams from both organisations to develop Por Mi Barrio.This finding is aligned with the literature on transaction cost economics and interorganisational relations, which highlights that there are multiple means of coordination and control among different actors.The ex-ante moment of the signing of these agreements, however, leaves a space for negotiation which is filled by the relationship between parties.In this way, the relational aspect is crucial for the minimization of risks (Ambrozini & Martinelli, 2017).The agreement was an instrument serving the ideological calibration of actors in the project and is a central piece to understand the boundaries of meta and institutional governance in this case study as it adopted a specific mode that organised both main actors' teams, expectations, and timings.
Despite formal rules being important in our case, our findings also show these are not sufficient to prevent frictions or projects to fail.In line with political economy analysis in the region (Stein & Tommasi, 2006), public policies tend to fail not because of their substance, but because of the politics of the process of implementation.In digital government, political changes may simply derail entire projects because of the lack of leadership buy-in, probing too unrealistic the assumption of linear and irreversible changes (Iannacci, 2010).The implementation of Por Mi Barrio is an exemplar of the capabilities that are needed to operate such a multi-level project.Managerial and political skills prove to be essential for collaboration to unfold and make such a project a viable alternative.The governing needs and capacities of the second governance order (Kooiman, 2003) reinforced the role of the main actors involved, which may both hinder and promote digital transformation projects (Gegenhuber et al., 2022).There is a significant need to coordinate efforts at an institutional and technological level.This proved to be critical for Por Mi Barrio to grow its user base with secondary alliances and a larger network of partners, including the Ombudsman Neighbours' Office and grassroots organisations.
Political battles and leadership transitions at the level of the Ombudsman office, but also at the IM with the electoral timing are a reminder of how institutional trajectories pose constraints and can challenge scaling an innovation in ICT4D contexts (Nicholson et al., 2022;Sahay et al., 2009;Sahay & Walsham, 2006).
While it is difficult to assess the managerial competencies of the sector leaders, the temporal factor was significant.Although it helped the artefact to coalesce and synchronise the IT teams from both organisations, the 'window of opportunity' (Granqvist & Gustafsson, 2015) was partially lost with the electoral timing and new political authorities.Significantly for the NGO the delays also affected its capacity to steer human resources to the project as the timeframe and funding for the initiative was also running out.Though the institutional level (second order) challenged the project's feasibility the most, it is at this level where greater competencies were achieved for both the municipal government and DATA in future projects concerning digitization and public service/interest delivery.

| The broader digital platform ecosystem governance
Por Mi Barrio is an example of the broader stakeholder relations included in digital platforms for development.The case we study challenges a platform owner view of ecosystem governance or orchestration.For example, extant literature on government platform orchestration suggests an ex-ante design and configuration orchestrated by the government in the production of public services (Cordella & Paletti, 2019).While this approach is useful and valuable from a government-centric point of view, it does not help to explain what we find in our case.Instead, our findings support a rather dynamic and strategic process of platform governance and orchestration (Addo, 2022) that unfolded over time, negotiated between the multistakeholder project partners.Our study reinforces the role that non-state and non-profit actors have in shaping the governance of platform ecosystems in developing country settings.
Our findings highlight the peculiarities of civil society within the platform ecosystem.National and subnational governments from the Global South have relatively more challenges and obstacles than their counterparts in the Global North (Addo & Avgerou, 2021;Avgerou & Bonina, 2020;Danziger et al., 1982); but civil society may have more knowledge, flexibility as well as legitimacy (Scholte, 1999) in these contexts.This allows the push for the implementation of democratising agendas with digital technologies.Our findings show that Por Mi Barrio underscores the role of transnational networks of civil society actors that can have access to resources and work with public actors (Quack, 2016).Though some ecosystem governance mechanisms in this case were enforced by both the government and the NGO, the role of civil society and donors in defining the design of the artefact and some rules for the interaction emerges as a distinctive feature of the case.
Our work also shows the vulnerability of civil society working with these issues in the region.For example, DATA as an NGO with a shared past with many leaders in the IM was unable to advance on its own with the outreach of the project.(Belanger & Carter, 2012;Benbunan-Fich et al., 2020).The three orders of governance model that we develop in this study also complement previous theorising of digital platform governance for socioeconomic development and within ICT4D contexts.In line with previous studies of digital platforms for development (Jha et al., 2016), ours show that to thrive, a convergence of efforts by different actors is required to sustain and expand a project like the one in Por Mi Barrio.Unlike those studies, we do unpack in detail the technical, and political/ideological dynamics of the different actors as critical elements in sustaining the project.
The framework also proposes a suitable and needed approach for linking micro, meso and macro-orders in studies of global governance, and therefore extends relevant work in IS and ICT4D (Madon et al., 2007;Nicholson et al., 2022;Sahay et al., 2009).
Second, we contribute empirically to expand knowledge on digital platforms for development in the Global South, and Latin America in particular.The rich case we present in Uruguay, while bound in the peculiarities of its context, can shed light to various other similar processes that other cities experience in the region, including the institutional settings that embed artefacts and values (Galperin, 2004).In doing so, it also nurtures research contemplating the sociotechnical imaginaries and their influence on policy making in different geographic contexts (Jasanoff & Kim, 2013) and the sources underpinning contemporary digital transformation.
Third, the contextual explanation we propose in this study contributes to extending the predominant and narrower view of context as a social domain (Avgerou, 2019), to account for a sociotechnical phenomenon (Sarker et al., 2019).The three orders of governance that we suggest are helpful to explain clearly and systematically how collaboration is facilitated or hindered; but also, to unfold the nuances that may be specific to the Latin American culture and context.The contextual factors we find extend the social dimension (in the relations between power, politics, and open government ness ideologies), to also account for the features of the technical artefact, and the geographical context (Montevideo and the Latin American).
Fourth, our study contributes more generally to the literature of governance of digital platforms ecosystems in public services.Research in management has only recently started to pay attention to the entrance of Big Tech platforms such as Google, Amazon and Microsoft in highly regulated services such as health and education to find that they pose risks to broader societal and human rights (Ozalp et al., 2022).The understanding of governance collaborations with alternative players such as NGOs in public services like the one we explore in this work is a contribution to move away from the commercially driven focus that prevails in digital platforms research (Bonina et al., 2021;Nicholson et al., 2022).Moreover, our work aligns with that of public administration that examines trends towards collaboration as part of a broader move to address the diffusion of knowledge and institutional capacity among different actors (Ansell & Gash, 2007;Bianchi et al., 2021).Distinctively, our study addresses a difference with respect to their government centric approach, as the originator of the project is not the government but a civil society actor.Our research brings in additional clues to understand the different roles of civil society as agents of digital transformation.Their roles have been varied, with different degrees of involvement as well as resistance against Big Tech in Global South contexts (Aguerre & Tarullo, 2021;Alvarado Garcia & Le Dantec, 2018;Rodima-Taylor, 2022).Our work has exposed how their knowledge claims and practises, as well as their institutional capacities, can influence public policies concerning digital-enabled transformation.
Finally, our study also contributes to shedding light and nuances on the increasingly used, yet ill-defined concept of technological sovereignty (Caravella, 2021).It does so by contextualising the development of the project in a country and region that are not at the heart of contemporary debates on technological sovereignty, as is the case with Europe.In the case of Por Mi Barrio this sovereignty is understood as capability, agency and autonomy to create technologies and to adapt them to their needs, rather than as an autarchical project where local actors control all aspects of a technology.

| Contributions to practice
Our study has important implications for practice.For policy makers and the civic tech ecosystem including donors, we suggest the need to propose policy changes in the region to enable more collaboration with small social enterprises or NGOs to contribute to digital transformation.More precisely, we suggest creating funding schemes for gov tech such as the UK's GovTech Catalyst Fund as well as new regulations to include more NGOs and small and medium enterprises (SMEs) into digital government transformation.
We also suggest policy makers and regulators in the region to expand the policies on interoperability in the public sector to not only include the technical aspects but to address semantic, organisational and legal interoperability (European Commission, 2022).Our work shows the critical enabling role that working with an open standard such as Open311 entailed for digital transformation phenomena.We suggest thinking more substantially about the role of datain the semantic layer of digital transformationas a critical factor for digital transformation enterprises.
Embedding interoperability into design principles of IT artefacts and government platforms may be a crucial measure to successfully develop digital ecosystems around open government.
Our study is especially relevant for local government settings.The application of such initiatives at a municipal government level leaves a greater margin for innovative governance arrangements which may include collaboration with other sectors, particularly those with a focus on the public interest (Gutierrez, 2018).We warn, however, both public administrators and complementors in the region of the existence of a technical but also political learning curve for the teams.Having good translators to implement digital transformation projects at the technical level was a crucial feature in our study.Equally, developing skills to navigate the broader institutional tensions is fundamental.Particularly applicable to other local settings in the region with less expertise of working with governments, we recommend taking an incremental approach and weaving international networks to help to navigate complexity and achieve synergies and benefits.The incremental approach followed in Por Mi Barrio in Montevideo may serve to both quell concerns of over-demand and exploitation, and it is also a more realistic approach in terms of human and financial sustainability for projects that are usually operating with low budgets.
The research also underscores the need to rethink different types of alliances between actors.Grassroots organisations could have helped DATA and the IM to reach out to citizens in more marginalised sectors.A second type of alliance concerns the repositioning of actors that have worked with open data and data access to address some of the imbalances of data concentration in large platforms as strategic partners for governments and local SMEs.A final type of alliance is the strengthening of collaboration between NGOs in the south and the north alike.These networks are relevant for technical, political and economic support, and to eventually increase the chances of success in their projects.Our case also shows that the civil society in Latin America has a trajectory of regional networks and alliances to leverage their work, visibility and resource mobilisation which may be used to further enhance collaborative governance approaches aiming to foster digital transformation with a public interest lens.The initiatives that have been monitored in the region (Fumega & McNaughton, 2019), point at the difficulties of scaling these platforms massively.

| CONCLUDING REMARKS
Governments in Latin America increasingly turn to digital platforms and their ecosystems to transform how they operate and innovate their public services.The need for understanding arrangements between different types of actors to drive digital platforms in government remains outstanding to foster change.In this paper, we provided a framework of how architectural design, rules, practises and ideologies play over time, to advance our understanding on platform governance in public services, contextualised to the reality of the region.Collaboration between actors in this work has been framed with a governance lens to capture the different levels of issues at stake-technologies, rules and ideologies, as well as their interplay over time.Crucial for our study were the sustained values and ideologies around open government platforms that crystallised in open services and open data to enable the initiative to thrive.But a highly responsive initiative such as Por Mi Barrio (Peixoto & Fox, 2016), had problems to scale and reach a critical mass mainly due to the conflicts that emerged at the institutional level (second order of governance), and the limitations that encompass the managerial and political capacity, the lack authority across sectors and actors, and vision of its leadership.If digital transformation is to be understood as a driving force for improved government services, technological fixes on their own will not address mid-term institutional change and longer-term meta governance strategies (Galperin, 2004;Sahay et al., 2009).
As with any study, we recognise ours has its limitations with an open road for future work.To further develop the contextual explanation approach of this article, future work could design comparative case studies in the region and elsewhere.For example, forthcoming work could compare Por Mi Barrio to the UK's counterpart Fix My Street to enrich our understanding of the Latin American peculiarities.An empirical limitation we encountered was to obtain more feedback from the sectors in the government that showed less interest or even hostility towards the initiative.
Even though in our fieldwork we tried to unpack as much as possible how the collaboration between different stakeholders unfolded, we do understand that there were project dynamics within the government that brought specific tensions.Future studies could explore these in more detail also in relation to other digital transformations.Finally, an important voice that we did not cover is that of citizens and potentially marginalised users.Although we explored levels of satisfaction and areas for improvement that citizens shared in the survey we conducted, our study did not focus on the citizen's interests or the potential trade-offs or detrimental effects that the platform may bring about.Future work could untap how similar digital platforms make citizen voice more inclusive, or to the contrary, the inequalities that may be exacerbated by leaving out those that have no access to digital technologies (Pak et al., 2017).Bringing these voices into these experiences is of uttermost importance in a region with high markers of inequality.
Latin America is facing an era of difficult times ahead, with increased polarisation, surveillance, and political and economic uncertainty in various countries.Despite the relevance of the open data, transparency, and open government agenda for the region during the 2010s, its prominence and influence is on the decline.The power of big tech platforms such as the Google, Twitter or Facebook via apps like WhatsApp are increasingly threatening indigenous innovations for democratic and public services in the region.A flexible, purposeful, and practical approach to work and govern digital platforms collaboratively with local actors seems crucial for the region, where pockets of innovations may emerge from local governments willing to address issues of autonomy-such as those covered in this paper.

FUNDING INFORMATION
We gratefully acknowledge funding received from Avina Americas during 2018 to conduct part of this research work.
the decision on a solution; and drawing a boundary around it.It has typically implied an operationalization that addresses the technologies as a way of doing things ('Techn e').Building on IS research, we operationalise this level with the characteristics, design, and implementation of the artefact, where the problem definition and direction for a solution and the value proposition around the IT artefact (digital platform) is enacted.The technical artefact represents the different actors' vision on problem-solving and opportunity creation, as well as a wish to change an unwanted situation with this instrument.

4
| ' POR MI BARRIO' : FEATURES, TRAJECTORY, AND MAIN RESULTS Por Mi Barrio (PMB) (https://pormibarrio.uy/) is an open-source web service that was launched in Montevideo in 2014 with the aim of facilitating city dwellers to report urban problems (such as potholes, litter, lighting).In simple terms, Por Mi Barrio offers an easy to use web and mobile app where citizens can report and check the status of repairs of urban problems.It is a local adaptation by the Uruguayan NGO DATA of the FixMyStreeta software developed by the UK-based social enterprise My Society in 2013.Different from their UK counterpart, DATA adapted the software to integrate Por Mi Barrio with the municipal Government of Montevideo's complaint management system called 'SUR'.The integration was possible thanks to an open platform featurethe Open311, an open standard that enables interoperability between applications widely used in urban planning (Rudmark, 2020).Por Mi Barrio was deployed within an enabling context for digital transformation in Uruguay.Since the 2000s, the state policy has been oriented at stimulating social, political, and economic transformation through digital technologies as well moving towards greater transparency.Examples include the creation of a National Agency for e-Government and Information Society (AGESIC) in 2006, and the enactment of the Law on Access to Public Information in 2008 that fostered the later development of open government and open data portals.Other policy milestones are shown in Figure 3 below.
. DATA, a local NGO leading Por Mi Barrio, was created in 2012 with half a dozen members on a part-time and mostly volunteer basis.Its founders had been advocates of the open government, open data and open software communities in Uruguay and in the region, with a history of working together with the government.The Municipality of Montevideo as the main government partner with a long tradition of developing in-house information systems, and a capable and relatively large IT department (Bonina & Eaton, 2020).The IT staff at the Municipality of Montevideo was also known for its commitment to open source and open data, being the first city in Latin America (and probably globally) to have an open data policy and portal launched in 2009.A third relevant actor is the AVINA Foundation, an international philanthropy organisation that funds civictechnology applications to foster innovative models of citizen participation and advocacy in the region.AVINA sponsored DATA with seed funding (100 K USD) for DATA to afford costs of human resources and infrastructure to deploy the F I G U R E 3 Por Mi Barrio: main project features and enabling conditions (2008-2020).Source: Authors.T A B L E 3 Actors involved in Por Mi Barrio, roles and vision/value of the platform.
civic demand for open data, and not just a concern of the elites.The context of Montevideo with tech savviness, commitment to open technologies and open data were appealing to both the NGO, the donor and the government.For the NGO, pursuing the project was attractive for three reasons: (i) the project was aligned with its mission, (ii) the founders had strong links with the open data policy and the civic movement in the UK and (iii) it meant securing AVINA Foundation as a solid partner and much needed financial sponsorship.The government, in turn, was keen to spur their online complaints system based on open data, open standards and at no cost to taxpayers.AVINA, the donor, established as a condition for

5
| CASE ANALYSIS: THREE GOVERNANCE LEVELS OF POR MI BARRIO 5.1 | First order governance: Open platform architecture Technically, Por Mi Barrio was built on an open architecture, both in terms of content (i.e.citizen complaints and reports that are also visible to users), as well as at the protocol layer for the modular interphase built with Open311.The Open311 protocol enabled the interoperability between the front-end user complaints addressed by DATA and the management of the complaints which falls under the scope of the SUR management system of the Montevideo government.This interoperability was a crucial condition for the project's funding and an output for the Municipality of Montevideo and DATA.The relevance of the open standard for the replication and extension of the initiative was well highlighted in the project.As the technical leaders in the government put it: So this [Por Mi Barrio] is good as an experience, and it should be replicated, it is one of the policies we have, of open data, and not only open data, but open protocols… We did an integration of Por Mi DATA Leader, our interviews/ID#5) Finally, what is agreed is that they (DATA) develop the platform based on Fix My Street, we (the IM) develop open services based on our policy that make the Unified Complaints System (SUR) visible to the outside world and with this principle we made a formal agreement.(IM, Technical leader IT, our interviews/ID#7) The Agreement established that DATA would oversee the development of the digital service, and the Municipality of Montevideo would commit to work on the integration between Por Mi Barrio and their own, in-house reporting system SUR.It played two important roles for the NGO and the government: it helped DATA to validate its product and therefore served to endorse Por Mi Barrio as a solution that was innovative, viable and aligned with the norms and values of open standards, open data and accountability espoused by the Municipality of Montevideo.The implementation of Por Mi Barrio cannot be understood without the ideological convergence of the main actors involved in the project.Open technology as a catalyst of social and political change had been a conviction among the actors involved in Por Mi Barrio, which can be traced to initiatives concerning open data and open-source software as early as the 2010s.The main actors involved in Por Mi Barrio (including the government and the NGO but also the donor) knew each other and participated regularly in regional and global fora on open data and open government.It was in these fora where principles and standards were developed, and lessons and best practises shared.For example, in one of these events, Fix My Street provided inspirational value for DATA to imagine the indigenization of the platform to generate an innovation in Uruguay.More significantly, the shared political ideologies help to explain why Por Mi Barrio survived the change in government in 2015.As it is well known in Latin America, changes in government with the entry of new teams are a hallmark of IT projects to fail or become discontinued in public administration.In our case, the shared beliefs at both technical and broader political levels were crucial for the continuation and later expansion of the project: The [government] profile was basically engineers, technicians.In general, we [DATA] dealt very little with political actors.(…) I think the people in charge of this [in IM] were also people with a particular leadership, and who came very strongly from the free software and open-source community… They had a very progressive vision of technology in the broad sense of the term; people who said "ok, how is technology going to genuinely benefit people and benefit the municipality"… (…) people who were keen to reform processes.(Data co-founder, our interviews/ID#01) Rooted in similar beliefs, these ideologies were widely shared among DATA and some sectors of the Municipality of Montevideo.However, the implications of open standards and transparency had different consequences for both actors.On the one hand, open standards implied increased scrutiny, and on the other, it entailed the possibility of more responsiveness and ultimately, a new channel for citizen participation.

6. 1 |
The recursive role of the open government ideology Our findings reveal that the macro-level of governance (meta-governance) played a pivotal role in shaping and reinforcing the preferred technical choices in the case.The longstanding shared ideologies on open government platforms and open technologies were instrumental to navigate a complex political scenario marked by the indifference of some sectors, as well as a period of elections during the early phases of the project.The level of cohesion achieved at the meta-governance can be explained by an ecosystem of actors around the platform concerned with open data, government transparency and the democratising potential of open IT.At a pragmatic level, this was manifested in the platform's technical design, notably in the deliberate choice of interoperability, which encapsulated the value of open data, and open standards.The platform materialised the first order governance on the technical For example, shared values towards open standards and open government platforms shaped the IT artefact design choices in the specific case of Por Mi Barrio, but also conditioned future choices.As a collaborative project, Por Mi Barrio helped to reinforce the broader discourse on open government and technological sovereignty by setting a precedent and the condition that future digital public services shall also be developed based on open platforms and open technologies.
Summary of characteristics and attributes of theoretical framework applied in the case.
The unstable conditions are embedded in the ecosystem of Latin America where civil society organisations working on digital innovation, development, and human rights with technology face challenges to develop independent and longer-term projects (ECLAC, 2021).While in Uruguay there were legal provisions on open software, open data and access to information, the implementation of these norms with Por Mi Barrio surfaced the tensions and challenges of public administrations.Por Mi Barrio was not only about data and standards, but one with the potential to change the dynamics of power.
7 | RESEARCH CONTRIBUTIONSOur research makes several empirical and conceptual contributions.First, our study contributes to displaying and unfolding how local governments can work with and support organisations of civil society to build and deploy open government platforms for public services.We build an encompassing socio-technical perspective of digital transformation in government by building a framework that combines governance theory from political science with IS research.This framework addresses successfully the lifecycle of a digital innovation and combines problem solving strategies around the technical artefact with the institutional forces that surround it as well as the more enduring ideological forces.As a result, the conceptual framework we propose in this paper responds to the call for advancing theory-building in digital government research