Digital Participatory Platforms for Co-Production in Urban Development: A Systematic Review

A renewed interest has appeared in citizen co-production of public services due to financial pressure on governments. While social media are considered an important facilitator, many digital participatory platforms (DPPs) have been developed to facilitate co-production between citizens and governments in the context of urban development. Previous studies have delivered a fragmented overview of DPPs in a few socio-spatial contexts and failed to take stock of the rise of DPPs. This article aims to provide a more comprehensive picture of the availability and functionalities of DPPs. Through a systematic review, 113 active DPPs have been identified, analysed, and classified within a citizen-government relationship typology. Almost a quarter of these DPPs demonstrate a realistic potential for online and offline co-production between governments and citizens. The article critically analyses the characteristics of these DPPs and explores their real-world applications in urban development. The article concludes with directions for further research.


INTRoDUCTIoN
Collaboration and participation of citizens in governments' activities at all levels has received increasing levels of attention in many disciplinary fields such as public administration and government studies, urban planning, public service design, computer science, and information technology (e.g. Bryer & Zavattaro, 2011;Gil de Zúñiga et al., 2010;Linders, 2012;Magro, 2012;Munthe-Kaas & Hoffmann, 2016;Sanders & Stappers, 2008;Slotterback, 2011;Verschuere et al., 2012). Much of this attention derives from the potential contribution of new social media, digital platforms and other ICTs to the interactions between (national and local) governments and citizens. Because of wider economic trends, welfare state retrenchment, devolution and new knowledge-sharing patterns, citizens' demands and governments' actions increasingly require two-way engagement and closer collaboration (Kleinhans et al., 2015). A renewed interest has appeared in citizen co-production of public services, especially in view of the financial pressures currently facing governments around 1. To provide a more (globally) comprehensive inventory of the availability and development of DPPS for various engagement purposes, their adoption, and actual use by government and citizens, ranging from information sharing to consultation and co-production; 2. To identify those DPPs in the inventory that demonstrate the potential for co-production between governments at various levels and citizens groups, by analysing and discussing their features, the kinds of problems and issues they are used for, and identifying potential case studies for future research on technology-enabled co-production.
For these purposes, we have conducted a systematic review for the identification of DPPs using both academic and practitioner literature, and a Google search in different languages to cover a broader 3 spectrum of countries around the world. The systematic review first and for all contributes to a much more comprehensive inventory of DPPs, which meet different purposes and goals and reflect varying degrees of intensity in the interaction and collaboration between governments and citizens. As such, this inventory goes beyond and adds to previous reviews of digital platforms for engagement and collaboration. It identifies examples of DPPs for co-production purposes, along with their essential technological features and real-world applications, showing the availability and readiness of technology and paving the way for future research on technology-enabled co-production which should guarantee new knowledge and understanding of how we can use the available technology more effectively.
However, because there are no hard limits between co-production and other forms and intensities of government-citizen interactions, the paper starts with a review of the literature on the different levels of collaboration and engagement between government and citizens that are fostered by DPPs. From this, a typology of citizen-government relationships is therefore developed in section two as the basis for the classification of all identified platforms within one of the levels, grounded in a working definition of DPPs and theoretical foundations of concepts such as co-production. In section three we present the research design that we have applied for the identification and review of DPPS. Section four moves on to discuss the results of the systematic review and to classify the platforms on the basis of the typology as developed in section two. Section five analyses the main elements and functionalities of the platforms that claim to allow for collaboration and co-production identifying their distinguishing features. The final section draws conclusions and provides suggestions for practitioners and for further research.

LEVELS oF THE CITIZEN-GoVERNMENT RELATIoNSHIP ENABLED By DIGITAL PLATFoRMS
Digital Participatory Platforms (DPPs) are defined here as a specific type of civic technology explicitly built for participatory, engagement and collaboration purposes that allow for user generated content and include a range of functionalities (e.g. analytics, map-based and geo-located input, importing and exporting of data, ranking of ideas) which transcend and considerably differ from social media such as Social Networking Sites and Microblogging (Facebook, Twitter and Instagram). DPPs thus allow for different levels and intensity of citizen engagement and participation. Many authors have tried to identify the levels of citizens engagement and participation in government activities through the use of digital technologies (De Souza & Bhagwatwar, 2014;Ertiö, 2015;Evans-Cowley & Hollander, 2010;Jones et al., 2015;Khan, 2015;Li & Feeney, 2014;Linders, 2012;McMillan, 2002;Mergel, 2013;Suen, 2006;Williamson & Parolin, 2013). Generally, three levels of engagement with an increasing degree of interaction are identified within the academic literature. They could be summarised as follows: • Information sharing: One-way communication from government to citizens. McMillan (2002) calls this Monologues whereas Linders (2012, p. 449) defines this level Government as a Platform, in which governments equip citizens with data needed to make informed decisions or design policies/services in a way that maintains freedom of choice, but stimulates a 'socially optimal' option; • Interaction: Two-way communication with dialogue between citizens and government representatives flowing both ways. McMillan (2002) calls this mutual discourse; • Civic engagement, involvement, collaboration: On this level, the two-way interactions go beyond basic information exchange to 'materialise' in policy measures, joint service delivery or other interventions. We will refer to this level as co-production, i.e. the public sector and citizens making better use of each other's assets and resources to achieve better outcomes and improved efficiency Loeffler, 2012, p. 1121).

Co-Production and the Role of Digitally-Mediated Practices
The Internet facilitates 'networked co-production' which emphasises the relationships between government and communities of citizens, i.e. users and non-users of specific services (Meijer, 2011, p. 599-600). DPPs can be particularly useful as content can be created in networks and communities, made available to all members of the community and stored in an accessible format to create an online repository for virtual communities (ibid., p. 601). As such, citizens and professionals can "bring different types of knowledge -the one general knowledge of the core (primary) process of the organization and the production of service, and the other situational or local knowledge" (Brandsen and Honingh, 2016, p. 430). Within the co-production level where citizens can identify, discuss problems, and propose solutions, concepts and practices such as PPGIS, collaborative mapping, Volunteered Geographic Information (VGI), crowdsourcing, and neo-geography play a fundamental role (Brown & Kytta, 2014;Goodchild, 2007;Silva, 2013;Wilson & Graham, 2013a). PPGIS describes technologies that aim at supporting public participation in a variety of contexts with the aims of inclusion and empowerment of marginalised population to inform environmental and urban planning (Babelon et al., 2016;Brown, 2012;Brown & Kytta, 2014;Panek, 2015). Brown (2012) states that collection of spatial data in PPGIS methods is agency driven as opposed to Volunteered Geographic Information (VGI) where citizens and individuals freely contribute their knowledge and information about a specific part of the city (Adams, 2013;Coleman et al., 2009;Goodchild, 2007). Such practices fall within the broader conceptualization of 'neo-geography' as defined by Wilson and Graham (2013b, p. 4) who underline that their "use of the word 'neo-geography' marks digitally mediated social practices through explicitly spatialised data/code practices".

A Typology of Citizen-Government Relationship
Adding on to the main levels of citizen-government relationship, some authors (e.g. Desouza and Bhagwatwar, 2014;Ertiö, 2015) identify more levels and sub-levels which further specify the role of, and information flows between, the actors involved in the citizen-government relationship. Ertiö (2015) for example identifies consultation as a sub-level of the information sharing level where information flows one-way from citizens to governments, and criteria power (ability of citizens to determine a policy or service) and operational power (ability of citizens to determine how a policy or service is carried out in practice) as the two sub-levels of civic engagement, involvement and collaboration (the author calls this level empowerment instead of co-production). Interestingly, Desouza and Bhagwatwar (2014, p. 37), in their four archetypes of technology-enabled participatory platforms, identify the citizen-centric and citizen-sourced data archetype "as an alternative medium for citizens to organize themselves to make a difference in their local communities." Linders (2012) calls this level Do it Yourself Government where citizens self-organize to achieve their own purposes.
This, in our opinion, should be the 'top' level of the citizen-government relationship in which citizens self-organize to produce solutions. However, at this level there is little or no interaction between citizens and government and interaction takes place only in cases where choice and implementation of the solution developed by the citizens still requires some government action, as Desouza and Bhagwatwar (2014) and Linders (2012) emphasise in their categorisations. Table 1 summarizes our typology of citizen-government relationship as drawn from international academic literature.
Despite a growing number of web-based and mobile-based platforms that enable information sharing and interaction between government and citizens, scholars have highlighted that use of DPPs is not yet interactive and is not being able to create two-way communication (Williamson & Parolin, 2013;Ertiö, 2015). To a certain extent, this is due to challenges, risks and other factors that influence government use and citizens adoption of social media and, in particular DPPs (e.g. Bertot et al., 2012;Casey & Li, 2012;Kavanaugh et al., 2012;Khan et al., 2014;Mergel, 2013;Picazo-Vela et al., 2012;Williamson & Parolin, 2013). However, building on and trying to expand previous scholarly work, this paper focuses on the availability of DPPs and on the specific features that characterise them, and shows that many interaction and co-production platforms have been already developed and are available and ready to use. Therefore, we have tried to identify a higher number of platforms than has ever been done before to understand development trends in several countries and contexts and to classify the platforms within one of different levels of the citizens-government relationship presented in Table 1. Other classifications are possible and because of this we believe it is essential to provide the full list of identified DPPs (see Appendix, Tables 3-7) as classified within all of our levels of the relationship. This will pave the way for future research on DPPs and especially on those that are classified in the co-production level, regarding their actual use, ability to foster co-production, involvement of citizens and implementation of ideas and solutions, reasons for success or failure. Below, we first discuss the research design that resulted into the list of identified DPPs and then explain how we classified them into the different levels of the citizen-government relationship (see Appendix, Tables 3-7).

RESEARCH DESIGN
For the identification of relevant DPPs we have employed a step-wise review approach with different sources. We started from the communitymatters blog (www.communitymatters.org) where a list of 50 tools for online public engagement was available. We checked all the tools that were listed on the blog to make sure that only those that were still online and accessible were included in our list.
We then searched the literature using the Scopus database to include all articles which discussed and presented DPPs for the engagement of citizens. We performed our search queries on 2 nd , 3 rd , and 4 th November 2016 with the keywords (participatory platforms), (collaborative mapping), and (PPGIS) in the fields "title, abstract and keywords" and limited it to the Social Sciences and Humanities Subject Areas (more relevant for urban studies, planning and geography) while not setting any Date Range limit. We did not use any search operator (e.g. AND, OR, " ") since this would have narrowed our search too much.
The keyword (participatory platforms) returned 552 document results, (collaborative mapping) 842 results, and (PPGIS) 114 results. For the keywords (participatory platforms) and (collaborative mapping), we also performed a manual selection from the search results based on titles. Studies were included if their title referred to one of the following subjects (government, citizen involvement, participation, engagement, planning, cities, PPGIS, crowdsourcing) even if the keywords (participatory platforms) or (collaborative mapping) were not included in the title itself. The documents that were Interaction Two-way communication with dialogue and feedback between citizens and government representatives.

6
deemed interesting from the title were downloaded and analysed further. We selected 54 documents for (participatory platforms) and 116 for (collaborative mapping). The keyword (PPGIS) returned 114 document results of which 29 had no full text available. We scanned all the remaining 85 articles to look for platforms mainly in the abstract, introduction, methods and conclusions sections. Finally, we performed a Google search in different languages to widen the scope and reach of our research outside the anglophone world for a wider analysis of the trends in the development and availability of digital applications and platforms. Other than English, platforms were searched in Italian, Spanish, Portuguese, French and Chinese. We used the national Google search engines for Italy, Spain, Portugal, Brazil, France and the keywords [public participation platform], [collaborative mapping platform], and [citizens engagement platforms] as they translate, not literally, to each language ( Table 2). We did search in Spanish in Google.com to cover Latin America and Mexico too. As far as China is concerned, a Chinese PhD student in our department performed the search for us. We used Google Netherlands and Baidu.com to performe the search and keywords [public participation platform], [public opinion platform], [community platform]. The search returned hundreds of thousands of results and we limited our review to the first five result pages for each of the languages which include more relevant and consulted examples.
We were thus able to cover many more countries than in any previous study. The results of this systematic review (summarised in Figure 1) aim to provide a very wide review and inventory of digital collaborative applications and platforms across the world. The list of all platforms identified through our review is available in the Appendix, Tables 3-7.
Following the approach employed by Desouza and Bhagawatwar (2014), the platforms were classified within one of the levels of our typology of citizen government relationship (see Section 2) on the basis of the relationship between the "agents", the knowledge/information flows among the agents, and the platforms' technological features. It is appropriate to focus on these three factors since they define "…aspects related to the collaborative environment of the platforms […] how outputs generated on the platform are used by public agencies and/or citizens, and the type of outcomes from the platform…" (Desouza & Bhagawatwar, 2014, p. 32).
It is important to stress that regarding the informing level we excluded platforms that act as government portals for information conveyance to citizens, since no actual government-citizens collaboration, interaction or two-way communication occurs in these portals (e.g. tax, welfare state, tourist information, public transit, parking information, Web-GIS Portals). As far as the sub-level private matters of the self-organization level is concerned, we have included only platforms that are concerned with neighbourhood matters that show potential of becoming related to matters of public interest within one of the levels of citizen engagement.
The analysis section focuses on DPPs classified as potentially fit for co-production, to fill the gap within the literature which states that use of DPPs is not yet interactive and is not being able to create two-way communication. In order to achieve this objective, we reveal which co-production platforms have been developed, the spatial context in which they have been applied and the characteristics that distinguish them from other kinds of DPPs. We focus on three elements: i) technological features which allow to perform certain tasks and reach a high level of engagement; ii) the kinds of problems and applications they have been used for, and therefore tested upon, which allows to identify empirical case studies across the world for future research; and iii) the pricing patterns which may compromise the ability and willingness of government and citizen groups to adopt digital solutions for their coproduction efforts. These elements provide fundamental information for adoption and use of such platforms by government agencies and citizens and allow to answer some questions such as: what are users allowed to do with one specific platform? Where and what has it been used for? How much does it cost? Further research on empirical applications and case studies will answer other and more detailed questions such as: how did the platform foster idea generation, resource pooling and discussion between governments and citizens? To what extent have better outcomes and efficiency been achieved, in relation to the objective?

DESCRIPTIVE RESULTS
Based on the research design, we have found a total of 113 digital platforms that classify within the identified levels of the citizens-government relationship (see Appendix and Figure 2). Our review has identified platforms from 21 countries around the world (see Figure 3) with the USA (n = 12), the UK (n = 3), Sweden (n = 3) and The Netherlands (n = 3) showing the higher number of co-productive platforms (25 in total). Other countries that have developed a co-production platform are Australia, Finland, France, Kenya. However, our inventory is by no means fully comprehensive and it is highly likely that many more platforms exist, and further work can expand the current list. For every platform we provide the name, URL (an extremely important element for future work by other scholars, guaranteeing replicability), a brief description with main purposes, geographic coverage (which refers to the country where the platform was developed or headquarters are located and differs considerably from where the platform has been used and applied), pricing, availability of documented case studies (again, fundamental for future work by other scholars), level of the citizens-government relationship, and main technological features (see Appendix, Tables 3-7). This last element, together with real world applications and availability of case studies, is one element upon which we have determined the specific level within which a platform classifies. More specifically, the availability of discussion forums, collaborative mapping tools, geotagging and geolocation tools, reporting functionality, voting and ranking options, submitting and commenting on new ideas, exporting and analytics are all elements that contribute to make a platform more suitable for co-production purposes. Unlike other studies (e.g. Desouza & Bhagwatwar, 2014), we have not included a column on information and knowledge flows as this is intrinsic to the levels as explained in Table 1, Section 2.
From the lists in the Appendix, as highlighted in Section 3, we have excluded those platforms that act as government portals for information conveyance to citizens within the informing sublevel. This kind of platforms are nowadays extremely widespread; yet, considering the development of technology and the aim of this article they are not relevant here since they do not connote a participatory platform and citizens can only gather and collect information through them. Desouza and Bhagwatwar (2014) present a wide array of such platforms in their model Citizen-Centric and Government Open Data. However, we have found four interesting platforms (Civic Insight, OS City, Open City Chicago, and Tell us Toolkit) that represent exceptions to the more widespread informing platforms and have been included for their general ability to visualize spatial data, query databases, and inform on different elements and on different spatial contexts. They therefore differ from municipal Web-GIS portals or similar services and are worth mentioning.
Consulting platforms (n=22) that require citizens to provide their views, comments and preferences through consultations and surveys exist but usually sustain no options for the government to provide feedback, thus lacking an interactive and co-productive potential. Interaction platforms are the most widespread (n=51) and show the greatest geographical diffusion (e.g. China, Brazil, France, Ecuador, New Zealand among other countries). Among these, very much widespread are the so-called reporting platforms (e.g. Fix My Street, Fix Ma Ville, Colab, Get it Done) that facilitate interaction between citizens and the local government on a specific practical issue to be solved (citizens report problems and the local government informs them when it has been fixed).
Co-production platforms (n=25) are also quite widespread with many examples from around the world (e.g. Block by Block, Commonplace, coUrbanize, Crowdmap, Maptionnaire, Urban Interactive Studio, TransformCity). Thus, these results expand previous studies (Babelon et al., 2016;Desouza & Bhagwatwar, 2012, 2014Ertiö, 2015, Evans-Cowley & Kubinski, 2015 and reveal emerging trends towards developing interactive and co-productive platforms that enable online and offline outcomes. In the next section we will focus exclusively on co-productive platforms to understand their technological features, pricing elements, and the kinds of problems and projects that are examined on these platforms that foster a mode of governance where the roles of citizens and governments converge towards co-production. As regards the self-organization platforms (n = 11), we have decided to keep some platforms that fall within the sub-level private matters since they are more representative of the level itself and concerned with neighbourhood matters that show potential of becoming related to matters of public interest (see also Table 1). However, we have excluded many platforms that would classify in this sub-level which do not seem to be related to matters of public interest (e.g. Craigs's List, Marktplaats, Bla Bla Car).

ANALySIS: FEATURES oF PoTENTIAL Co-PRoDUCTIoN PLATFoRMS
This section focuses exclusively on the identified DPPs that appear to allow for co-production between (local) governments and a range of stakeholders involving citizens, businesses, and other organizations. Since our definition of co-production emphasizes joint resource contributions by both the public sector and citizens, in order to achieve better outcomes (Bovaird and Loeffler, 2012), we classified DPPs as co-production platforms if they show potential in terms of their characteristics, features and applications to foster joint resource mobilisation and collaboration between (local) governments and stakeholders, including joint solutions, designs, delivery schemes and budget priorities for urban spaces, public facilities, etc. Therefore, redefining how public agencies and urban stakeholders connect, interact, negotiate and make decisions according to a paradigm of co-production in governance is of key importance. However, despite the fact that digital technologies and platforms have pervaded our daily life, many issues and limitations continue to exist and should be taken into account when using them, such as digital divide (Bertot et al., 2012;Burkhardt et al., 2014;DiMaggio et al., 2004;Norris, 2001;Pizaco-Vela et al., 2012), costs of technology, data storage and ICT infrastructure, training, and in terms of learning the new technology.
The interesting geographical pattern of distribution and diffusion of such platforms (see Figure  3) shows that the USA is the main pioneering country in co-productive platforms. This pattern could perhaps be traced back to the history of community development, planning and engagement that characterises the USA context more than others (Angotti, 2008;Davidoff, 1965). Given the current availability of technology, more citizens groups and cities in different countries can now employ them and undertake efforts for DPP-led co-production.
The following set of co-production platforms (see also Appendix, Tables 3-7) will be discussed in more detail: Bang the As explained in Section 3, these platforms will be analysed with regard to their technological features, the kinds of problems and solutions that are examined and proposed on these platforms, their real-world applications, where available, and pricing patterns. More in-depth future research is needed on co-production DPPs and related case studies to identify the nature, outcomes, engagement processes, and policy implications of co-production platforms.

Technological Features
Digital platforms for co-production purposes sustain a wide variety of features that allow for different user behaviour and collection of ideas, solutions, knowledge, discussion and collaboration between public and private actors. Our systematic review has identified all such features that can be summarised as follows: opinion maps, surveys, discussion forums, budget allocation, simulation design, voting and ranking of ideas, analytics, map-based and geo-located inputs for collaborative mapping such as comments, pins, and other geographical features, crowdfunding, exporting in different formats for further analysis (shape files, csv, kml), importing and media uploading, sharing on other social networking sites such as Facebook and Twitter. The simultaneous availability of many of such features is a fundamental element to classify a platform as fit for co-production. It is interesting to notice that co-production platforms build on many and different features that are available for other platforms by including them in one single platform. The most commonly-found features that represent the more essential ones for co-production are: collaborative mapping, map-based comments and ideas submission, voting and ranking options, media uploads, and analytics. The identification of technological features that characterise and connote co-production DPPs allows to clearly distinguish them from platforms which have other engagement purposes while at the same time highlighting what is needed, and what has already been developed, in terms of technology in order to be able to co-produce actively. This, in turn, provides government agencies, citizens and stakeholders in general with relevant information to choose a platform for co-production and collaboration purposes.
A fundamental feature of the co-production platforms is the spatial map-based visualization of issues, initiatives, and projects at stake that allows citizens to comment upon, design and co-produce new options and alternatives in a geo-located way and with specific reference to the spatial context. The map-based and geo-visualization tools have important implications for engagement practices since they allow participants to be more specific and accurate in both the discussion and solutions of problems, providing them with an intuitive way of representing issues, objects and solutions. This is fundamental in consideration of the real-world contexts and purposes to which they are applied, as discussed in the next subsection. These tools are able to increase the understanding of the issues 11 at stake, feasibility of solutions, and the spatial relationship between different elements and parts of the spatial context (Marzouki et al., 2017).
This, together with a built-in option for analytics (e.g. number of votes, ideas, participants, spatial analysis), is the main feature that distinguishes co-production platforms from platforms within other levels and for engagement purposes. Important examples of platforms with a user-friendly interface for a map-based feature are Maptionnaire, Carticipe, Commonplace, Transformcity, and Bang the Table. Special mention is for two DPPs that appear to be particularly interesting. The first one is Block by Block whose features allow citizens to create and design parts of their city simulating in a 3D environment what the new solution will look like. This is an outstanding example of how platforms can engage citizens in the effort of designing new solutions for public places. The second example is the Geojson platform for collaborative mapping that is based on open source technology and allows users to map, add properties and information, share their maps and ideas with other users, and export them into different formats for further analysis. There are other examples of open source platforms (Crowdgauge, Crowdmap, Map Server, Shareabouts) which represent an important opportunity for citizen groups and governments to collaborate without having to bear extra costs for the technology.

Kinds of Problems and Real-world Applications
Co-productive platforms are being applied to find solutions to several kinds of problems within the context of real world applications. However, there is a common denominator among all of the digital co-production platforms' case studies that we have analysed as part of our systematic review: the environment and places we live in and use in our daily life, be them urban, rural, coastal, with a specific use such as a bus station or a public open space such as a square. The focus is on co-producing with citizens, sharing ideas and solutions that can enhance the way such places are used and enjoyed in order to improve their lives and answer their needs.
The real-world applications can be of two distinct types: the first one is more practical, small scale and oriented towards redesigning a specific, spatially-bound object or service, such as a market, a university campus, a bus station, or a park. The second type has more the nature of future-oriented policy making or planning related to the production of a general planning vision for the city wherein citizens co-produce ideas in the fields of pedestrian and cycling mobility, cultural heritage, affordable housing, public transportation needs, and so on. Digital platforms are not limited and designed to fit just one of these two typologies. Many are the examples of platforms that are used for both of them (e.g. Crowdbrite, Commonplace, Mapping for change, Citizens sticky world). We will now focus on each of the two types in more detail.
In the first type, we find telling examples from Bang the Table with short-term development exercises for open space improvements along the beach through artistic installations, new pathways, boardwalks and benches, improvement of parks and recreation facilities. Block by Block is also a clear example. Based on the Minecraft game and in collaboration with UN-Habitat, it allows for reproduction of the environment and space in a 3D virtual world for citizens to plan it and design it. It has been applied to several examples all over the world and especially in distressed areas such as Kosovo, Haiti, Palestine, for instance for the design and upgrade of a transport hub, a new skate park, and a city market. Some of many other interesting examples and potential case studies for future empirical research in this first type are Citizinvestor, Commonplace, Co-urbanize, Ideascale, Urban interactive studio. Real world applications concern solutions for the location of new bike racks, plaza upgrade, playground renovation, improvement of a university campus, new concepts and ideas for a zoo, redevelopment of a factory building, new sporting village and a community stadium.
The second type of real world applications, which are more oriented towards long-term planning and general policy principles, involves examples of institutional participatory processes. An interesting example is Carticipe for the municipal and metropolitan plans of various cities in France such as Lille, Grenoble, and Avignon. This platform was used to co-produce ideas and proposals within the domains of public open spaces, sustainable mobility and bicycle lanes, public services and new retail 12 and commercial activities. Similarly, Crowdbrite was used within the process of developing the Las Vegas Master Plan. Last but not least, the location of new affordable homes and the development of neighbourhood plans (a new practice part of the current English spatial and urban planning system) are examples of real world applications found in Commonplace, Co-urbanize, Citizens sticky world, and Mapping for change.
The problems and real-world applications therefore concern several domains of the living environment in urban as well as more rural settings. Public spaces and facilities are the objects of such efforts of co-production through digital platforms with the underlying aim of delivering a new and improved design and a greater usability to meet the needs of the communities.

Pricing Patterns
As for the pricing pattern, there seems to be a connection between the availability of a higher number and more advanced technological features (and therefore co-production potential for users) and the pricing element. In fact, the majority of platforms (16 out of 25) classified within the co-production level have several explicit pricing plans and solutions that vary according to the level of service that is provided. The other nine platforms have no pricing plans, five of which embrace the open source philosophy. No difference is observed between platforms that serve for immediate and short-terms solutions and those which instead support more long-term planning applications. Pricing and financial affordability have implications for the ability and willingness of government and citizen groups to adopt digital solutions for their co-production efforts. If for large and richer cities and communities this might not be an issue, smaller municipalities and poorer communities might indeed perceive it as an obstacle.
Pricing plans can vary from a few dozens of Euros per month to as much as GBP 1,200 per month as in the case of Creative Citizens Sticky World. All in all, there seems to be a specific business model behind the platforms for co-production (and to some extent informing and consulting platforms) that is intended to meet a public demand and the need of governments to engage with citizens with state of the art digital tools, so exploiting an expanding market (Sifry, 2014) whose value is estimated to be of around 700 million US dollars' worth of investments (Gordon & Mihailidis, 2016). However, it is not simply a matter of availability of technological features that determines the price. Platforms for interaction are generally free to use and so are platforms for self-organization, while many of the informing and consulting platforms included in this article have pricing plans. The reason for this seems unclear and further investigation may be needed from business scholars to determine why this happens.

CoNCLUSIoN
In the current era of almost ubiquitous Internet accessibility, increasing attention and resources are devoted to the role of new information-and communication technologies in establishing meaningful and democratically legitimate citizen engagement. Due to economic trends, welfare state reforms, devolution and new knowledge-sharing patterns, there is a growing interest in two-way engagement and collaboration of governments and citizens (e.g. Ertiö, 2015;Kleinhans et al., 2015;Linders, 2012;Parrado et al., 2013;Williamson & Parolin, 2013). While social media are considered as an important facilitator in this respect, a plethora of digital participatory platforms (DPPs) has been developed to enable collaboration and co-production between citizens and governments. Previous research has revealed the promise of the Internet to "enable co-production on an unprecedented scale" (Linders, 2012, p. 446), but reviews have yielded a fragmented picture of DPPs in only a few specific socio-spatial contexts like the United States of America and some parts of Europe (see e.g. Desouza & Bhagwatwar, 2012, 2014Atzmanstorfer & Blaschke, 2013;Evans Cowley & Kubinski, 2015). Moreover, Internet-facilitated co-production has not been systematically studied yet (Clark et al., 2013;Linders, 2012;Meijer, 2011). 13 In this context, this paper aimed to provide a more (globally) comprehensive picture of the availability and functionalities of DPPs and to identify platforms that demonstrate a realistic potential for co-production between government at various levels and citizens groups. Using a three-step systematic review approach, 113 currently active DPPs were identified, analysed and subsequently classified on the basis of a typology of citizen-government relationships, ranging from information sharing to self-organisation (of citizens). This review has shown that almost a quarter of the identified DPPs can be classified as co-production platforms even though development efforts of such platforms are mainly concentrated in the United States while applications concern more countries such as Kosovo, Haiti, Palestine, Nepal, South Africa, Mexico and many more. This is in contrast with previous research that has identified a lower number of collaborative or co-production platforms (Desouza & Bhagwatwar, 2014;Ertiö, 2015, Evans-Cowley & Kubinski, 2015. However, taking into account the development and availability of technology, efforts towards digitally-enabled co-production could be undertaken by many more governments and communities around the world and further investigation should be devoted to understanding how we can make the most out of the available state-of-the-art technology, within the context of community development traditions. The subset of 25 co-production platforms has been analysed with regard to their technological features, pricing patterns, real world applications and addressed problems and solutions that are examined and proposed on these platforms. Our systematic review has not only contributed new knowledge by the extensive inventory, but also by identifying real world applications and issues which these platforms are used to address. In everyday practice, these platforms are either used for practical solutions for spatially-bound problems, objects or services in citizens' living environments or targeted towards future-oriented vision, planning or policy making of local areas, neighbourhoods, but also cities (master plans and local community plans).
Based on the review and inventory, a few general recommendations for policymakers can be offered. First of all, the review has identified quite a number of up and running co-production platforms. Policymakers, especially in English-speaking countries, should resist the temptation of building their own applications, and rather opt for an already established DPP that has been through numerous test and validation rounds, thus saving substantial time, money and energy in setting up a new platform for digital engagement or collaboration. Second, the range of available functionalities, from easy to more complex, should remind policymakers that not all members of target communities will be sufficiently technologically savvy to use all available functionalities of DPPs. While citizen users will already be a specific part of the general population, technological abilities will differ even within this group. This should be taken into account while applying a DPP to a specific socio-spatial context. Finally, the variety in applications of DPPs should remind policymakers that, while paying attention to the platforms is important, these should not be considered as 'stand-alone' objectives. Rather, DPPs should be perceived as instruments to enable public sector institutions and citizens to make better use of each other's assets and resources, for the sake of better 'offline' outcomes and improved efficiency (see also Bovaird & Loeffler, 2012), but not for the sake of technology itself.
Finally, clear avenues for further research can be provided. First of all, scholars around the world can expand the list of platforms (Appendix, Tables 3-7) and update the description of features, which is enabled by providing crucial platform data (including URLs) to create an up-to-date knowledge base of DPPs. A next step may be to further study actual take-up rates of various DPPs, to uncover reasons for either limited or highly substantial use of DPPs that claim to facilitate co-production. Third, co-production has been associated with strong expectations regarding its potential "to reinvigorate voluntary participation and social cohesion in an increasingly fragmented and individualized society" (Brandsen & Honingh, 2016, p. 427). However, many authors have observed that Internet-facilitated co-production has not yet been systematically studied (Clark et al., 2013;Linders, 2012;Meijer, 2011). Considering the availability of documented case studies for most co-production DPPs, it is our aim to conduct more in-depth research which should identify the nature, process, outcomes (both online and offline) and policy implications of digitally-enabled co-production.

NoTE
An ongoing project and repository of mainly offline participatory and deliberative governance experiences is available on the Participedia platform (www.participedia.net) (Fung, and Warren, 2011).

ACKNowLEDGMENT
The research leading to these results is developed in the context of the SmartGov Project (Advanced decision support for Smart Governance). It has received funding from the Joint Programming Initiative (JPI) Urban Europe, i.e. the program ERA-NET Cofund Smart Cities and Communities (ENSCC), under the European Union's Horizon 2020 Program. We are grateful to Xin Li, PhD student at OTB -Department for the Built Environment, TU Delft, for helping us with the search in Chinese. We are grateful to the three anonymous reviewers who provided insightful comments and suggestions which allowed us to improve our manuscript.