Nicholas Gates – Open Source in Local Governments: Lessons from across the EU #FOSSBack

This talk features Nick Gates, a senior policy adviser at Open Forum Europe, discussing EU-funded research on open source software adoption in local governments. He presents findings from a report detailing a five-case study analysis, revealing gaps in the existing research regarding open source in local governance. Gates highlights the collaborative dynamics developers use to create software solutions and emphasizes diverse adoption methods across regions. The report identifies three key archetypes of collaboration involving local governments, service suppliers, and community roles. Gates encourages further exploration and dialogue on improving open source adoption practices within local public sectors.

Hi everyone. Uh so my name is Nick Gates. Uh you can also call me Nicholas. Uh I'm a senior policy adviser with Open Forum Europe. Uh we are a nonprofit think tank based in Brussels, Belgium. Um I'm based in Vienna, so I work remotely. Um but here uh today I'm representing the open source observatory. So I'm going to talk about some EU funded research that we did

on open source adoption and reuse in European local governments. Um and I should mention that Ozor is uh funded uh by DG Digit. Um and we uh work on this project in collaboration with a consultancy called Wavestone. Um so what we did is we did a five uh case study report. Uh it was published uh late u last year in September. Uh and you can download it

on our website. I'll give you a QR code to do that in just a moment. Um, so this is myself and Johan Lekar. We're the two main authors of the report. Uh, though we also uh had support uh from Simon Vaver and Jiannisades. Uh, if you want to download this report um, please feel free to go ahead and do so. You can use this QR code u,

and we encourage you to read through it. If you would like a more uh kind of digestible version, we also have some summaries available on the website and we also have longer in-depth narrative versions of the case studies that are annexed in this report. So the case studies that I'll talk through are all available as kind of longer case studies with uh nice fun quotes and uh

detailed findings etc. Uh so really what we were trying to accomplish in this research was to start a conversation and to fill a gap. Um what we found was that there wasn't really any systematic research of open local governments. Uh we use the term local governments very deliberately by the way. I'll get into that in a moment. Um and we just wanted to fill the gap. So

there was systematic literature available on foss adoption and reuse. Um but not at that level. And so we wanted to kind of start that conversation um and provide a framework for kind of thinking about how we can investigate this uh area of open source adoption and reuse uh in more depth. So here's the problem context exactly as I said it. Um we uh found that on the

local level of government um there's a lot of adoption and use of open of common open source software solutions and uh there is uh collaboration on their development and maintenance but it's not really well stood well understood in terms of public sector organizations. Uh and so we wanted to create new knowledge in that area uh and start the conversation around different archetypes for how open source collaboration

is happening at this So uh we did a simple literature review. Um we uh reviewed over 60 papers, analyzed over 30 papers uh specifically in depth and then incorporated about 15 of them into the final written document. Um so we have a detailed thematic analysis available um which kind of summarizes uh some of the problems and challenges in the literature that I mentioned. Um I personally uh

did a mapping of over a thousand projects, policies, institutions and activities um at uh the European uh kind of local level. Um also the regional level though ultimately we all only focused on local um use cases. Um and then we shortlisted around 20 to 25 case studies um resulting in the selection of these five. Um and I'll explain that selection in just a moment. Um we also

conducted over 19 interviews uh across uh 21 different persons uh for the five case studies. Um and uh they all made their way into the final report. So here were some of the archetypes that uh we kind of hypothesized in the leadup to this study because we wanted to get a wide and kind of representative sample of different types of collaboration happening in local governments. This is

not all of the archetypes that made it into the final report. I'll highlight those in a moment. Um but they were ways that we were kind of thinking about what are the kind of ways that um open governments come to adopt or reuse or collaborate on uh software at this level. Um so one archetype that is very familiar to many of you is that software can end

up in a foundation. So it becomes very widely reused. It's very popular. Um and it actually spins out into a foundation which is responsible for its sustainability and then other governments go oh that's shiny. We want to use that. Um there's kind of this international but less adopted kind of software that's out there that people in the know um but not is not necessarily in a foundation.

uh it has a reputation but um it's more known among developers or um it's kind of being used but at uh very different levels of the stack um maybe not at the service delivery layer. Um there's also kind of local open source solutions uh that are being developed by non-public sector bodies and reused those that are being developed by public or quasi public bodies. I'll get into

the qua public in a moment there's intercity collaborations one of which we highlighted in the report ultimately around kind of a common open source software solution. uh there's very local solutions that are developed in a city but aren't being reused anywhere and uh we actually found through this project that people uh decide to open source their software for many different reasons. It's not just around uh adoption

and reuse. Uh and then there's also regional adoption and customization of existing general uh projects for local needs. So what we really found was that open source collaboration at this level is incredibly diverse. There's not a single way that people come to it. uh and there's uh kind of cascading effects of how different uh groups uh intersect to uh help uh use open source software at that

level. So we really wanted to kind of unpack this and this was just kind of our first uh entry point into kind of surfacing case studies that might be interesting. Um ultimately we focused on five case studies um and we chose five case studies that we think represented a diverse cross-section uh of those different archetypes and spoke to different collaboration dynamics. And so I'm going to talk

you through each of the five cases and then kind of what they told us in terms of um kind of uh the archetypes for open source collaboration and then we'll uh kind of uh unpack uh where we might do further research. Uh so one which is very well known is called console democracy. Um some of you might have heard it. So console emerged in Spain uh around

2015. Um you know I think there was kind of a movement uh in kind of the the mid2010s towards citizen participation software. Um and you saw kind of a very rapid adoption. Console was at the very beginning of that. So deedum uh which will also have a presentation later today um kind of came in the wave of that 2015 election in uh Spain. Uh and there was

a lot of kind of progressive democracy-minded um uh parties that were emerging that wanted to bring democracy direct to the people. And one way they did that was very quickly rapidly uh iterating and developing an open source software solution that enabled people to participate in democratic processes at the local level. Um so uh console kind of very much thrived uh in this period. Uh Deedum kind of

emerged as as kind of a fork of console. Um and they've both been very successful. Um but we also saw kind of how uh you know local political elections also influence open source software. So in particular even more so than at the national level you're really subject to kind of um uh the waves of political momentum and support that can drive a project. And so after 2019

when the government changed uh in the city uh uh of Madrid, we actually saw kind of uh a decline in support for console. And this was kind of exacerbated by the onset of the co9 pandemic in 2020. Um and so you saw kind of a falling off of the project so to speak. Um but what you're seeing now is a very deliberate effort by the uh console

democracy foundation which was initiated in 2019. um as that uh support Wayne to really kind of bring the codebase back together, kind of stitch together all of these efforts and kind of like rebuild the community that was really thriving um in that early period. So period. So despite some challenges in governance and funding, um what we saw here was a really successful example of an international foundation

uh really trying to support and in this case even revive uh kind of a community uh bring in effective governance structures um and and kind of uh you know uh build bridges uh to allow um public sector organizations to collaborate in diverse ways. Um and this case study was really interesting to us because it also allowed it to be kind of an umbrella. Um you could do

a case study of case studies. So, we spoke to the folks at Code for Romania, and that was a really interesting example of, as I'll get to with another case study, a civil society organization providing direct support, almost serving like a system integrator to help uh install and customize a solution for several local governments. We saw a municipal association uh called Kosla in um Scotland. It's the

Scottish Association of Municipalities that was working to basically provide console as a service and kind of maintain it on behalf of all the municipalities in Scotland. Um we also saw a very active kind of um project coming in and contributing to the governance and financial sustainability of console and that was uh Munich, the city of Munich in Germany. So um there's many different ways to unpack what's

going on in console, but I think it's a really fascinating example uh of many of the different collaboration dynamics that I highlighted earlier. Another one is uh OS2 Borger Pes and Medborg Pace. Um so this is kind of a library software. Uh it's based on Ubuntu. Um and it is packaged for uh libraries uh at the municipal level. Uh so this originated in uh Copen not in

Copenhagen in Arus in Denmark. Uh and it was kind of iterated on as a a a kind of collaboration between the supplier Magenta um and OS2. Um there was a lot of kind of uh complications in terms of uh different philosophies that each of those institutions brought to the use of this software. And so what we observed here was kind of uh how suppliers can become very

dominant and play an outsized role in uh local open source ecosystems and how that can you know support but also come into conflict with uh the efforts of kind of public sector institutions in some cases when there's differing opinions around governance and and sustainability and things like that. So really what we saw is kind of OS2 which is this municipal association seeking to drive open source support

for local IT solutions um for local IT open source for local open source solutions um uh as kind of playing a role in kind of uh disseminating uh this project to to a lot of different municipalities. But we also saw um some governments choosing to work directly with magenta. Um and then we also saw interesting crossber dynamics when eventually Sambuk which is the kind of counterpart uh

to OS2 and wants to support adoption of uh open source solutions in Sweden also interested in adopting the same software. Uh and there were some dynamics there where Meeno was collaborating across borders with the the folks at Sambuk as well. Um yeah we had a lot of lessons that we learned here but um I guess the number one yeah is the role of the suppliers. Um but

also kind of that municipalities need to do continuous checks and reviews of concerned projects. Um that they should be wary of kind of procurement processes. Uh so market size and potential turnover in relationship to uh the number of service suppliers where um suppliers uh where services are procured from. Um and yeah I think uh just a lot of really interesting dynamics to unpack. Uh this is another

example of uh kind of that local kind of custom uh uh open source solution that emerges. So uh GMIO was uh organized and then eventually kind of uh run by uh operator ICT which is what I'll call a quai public sector body. So kind of a a state-owned IT corporation. Um and they did all of the original kind of uh coding and development of Galio as a

smart data platform to provide data services to to the city of Prague. Um and then eventually they chose to open source the solution because they found that it actually helped to um kind of make collaboration within the city easier ostensibly um and also because it um helped to retain and attract developer talent um people that were very interested in also participating in open source communities. And so

what we haven't seen in this case is a lot of kind of uh sharing and reuse uh outside of uh the borders of the city uh of Czecha though um uh the city of Prague. Um but there is kind of limited crossber sharing within Czecha specifically with the city of Bohemia region which uh with the Bohemia region which of course makes sense given that um Prague is

located in in the Bohemia region and there needs to be some collaboration in terms of integrated transport planning there uh which is one of the use cases. Um, so what we really saw here is a really great example of a a public sector body kind of working in a very agile and iterative way. Um, really kind of a great example of building trust within city stakeholders. Um,

and and kind of facilitating horizontal and cross- departmental collaboration uh through the open source and kind of licensing of its software um even where they weren't really focused on or interested in um kind of international use cases though they they did get some uh kind of uh interested organizations approaching them. Um so yeah I think this can show that uh successful open source software solutions at the

local level often can come out of solving very kind of um specific local needs. Um and it's ultimately as you see with console uh that type of innovation that can get reused internationally. We haven't seen that in this case but um I think it's a very prominent example of how that might emerge. Um here's another very interesting example. Um this is a solution um for parliaments called

uh parliament. Um, so this was initiated and is still to this day maintained by a civil society organization in Slovenia called Danisenovan. Uh, and Danisenovan is is one of the more interesting organizations I've encountered in this space. They they basically seek to promote democracy as part of their mission. Uh, and they kind of function very much like a civil society organization. Um, so many of their staff

are non-technical and focus on promoting democracy and facilitating adoption of tools and and and uh trainings and capacity building efforts. Um but we also have a strong technical team at Donisodon who kind of contributes to uh the development of uh parliamentator and kind of supplies it as a service to um interested parliaments. Um we see this specifically at the local level across uh Slovenia. Um so there

was kind of widespread adoption across uh a number of municipalities in Slovenia. Um but then we also saw it at the national level in Slovenia and then eventually kind of crossber sharing and reuse. um looking at governments in in for example Bosnia and Herzgoina um and Croatia. Um so this is a really great example of how kind of civil society and kind of a nonprofit driven service

supplier model can kind of step in uh and help support open source adoption and reuse in kind of lowcapacity governments. Um it also kind of speaks to kind of cultural resistance and political challenges um and how that can kind of be overcome through uh kind of a more integrated service supplier model. But um we also were really encouraged to see that there's uh ways of structuring those

relationships in ways that are not profit driven and and not highly extractive. Um so you don't un ultimately end up seeing a lot of um kind of let's say technical capacity building in kind of the beneficiary institutions but you do see a lot of support for the solution and kind of uh appreciation of the uh the dynamics of open source software collaboration and Uh and then the

last one is uh digit transit. Uh so Digi Transit is in many ways a very traditional open source collaboration in the sense that they they find uh some code online in this case uh open trip planner and they try and kind of build their own solution that builds on top of that. Um and what they really sought to do was to kind of use this kind of

core of integrated transportation logistics software to kind of build a custom solution for the Helsinki region. Um and we also saw a lot of kind of horizontal collaboration between uh a number of different actors. I won't go into all the names just not to confuse you. um in kind of the Helsinki region and then eventually uh sharing kind of reuse within uh Finland uh so by other

municipalities um but also uh across borders including with Estonia for example. Um so here we saw a really uh kind of compelling use case around how uh you know uh a city can kind of manage a burgeoning open source software community. how they can develop like clear documentation and onboarding processes, really start building a community within uh kind of uh national u borders um and kind of

really use standards to enable broader adoption um and then participate in true code development where you see public sector institutions um uh really kind of working upstream and and collaborating in a in a very um consistent way that is uh commenurate with traditional open source software collaboration dynamics. So uh what does all this tell us? Basically we think that there are kind of three archetypes. So we

uh focused in on three buckets. Uh the first one is local governments. The second one is around the service suppliers and then the third one is around communities. And we looked at how all of these different kind of stakeholder groups interact with one another um to promote um uh kind of the adoption and reuse of open source software at the local level. So in terms of local

governments themselves uh and we didn't focus on municipalities or cities or villages because there's a very kind of diverse range of different uh governance types uh we focused on what was kind of external to them what was internal to them and then what was kind of quai internal for lack of a better term. Um so external capabilities that really relates to kind of uh the pooling of

resources through foundations or local government associations like OS2 or through trusted external service suppliers uh like magenta. um internal capabilities. We see this very prominently in the cases of Digi Transit and in Glaio. That means kind of building the capabilities to collaborate and participate in open source software communities um within the government. Um but that requires substantial and long-term investments. What we saw was really um prominent

here was the use of kind of um IT corporations or companies that were kind of owned by the government. And uh you also see that at the national level with Pagopaya in Italy for example. And this seemed to be a very interesting model um for some of the bigger local governments uh in particular. Um and that that's kind of relates to kind of the the qua internal

capabilities. So these like kind of organizations that can kind of come on board um some of them are listed as joint stock companies as in the case of operator ICT and are subsidized by local governments and kind of kind of do this collaboration as an extension of of the state. Um and there's differing kind of um uh findings in terms of how different governments are able to

translate that back into kind of internal capability um within uh the city government itself. Um but that's some of the examples of of archetypes that we saw emerging. And again or not again but I I would like to emphasize that I don't think any of these um archetypes are um very strictly defined. So you can't just say this is one archetype. They're kind of overlapping. And I

I think it should be more of a useful heruristics or for starting to think about um assessing or evaluating u what's going on in a given uh jurisdiction. Uh then we have community based archetypes. So we talked about with console the role of international foundations. Um this is a model that can really be brought in to help kind of ensure the health and long-term sustainability of a

community. Um and in the case of console democracy for example, we actually found that it really helped to to save that project. Um, and while their efforts kind of flagged a little bit, um, they're really kind of doing all of the effort. They're becoming a digital public good, they're getting out there and, uh, exploring new fundraising channels to really kind of revive, uh, the community, bring some,

uh, elements of the codebase back together, and promote more collaboration between all of the different cities that are, uh, reusing, uh, that software. Um, national provider oriented associations. So, uh, this would be, for example, like uh, municipal associations. Um there's also development oriented associations. So those that enable pooling of resources and capabilities like Donovan did to uh kind of support um the interlinking and kind of collaboration

amongst these different governments that were adopting this uh solution for Uh and then supplier based archetypes. So uh again with the Donian of Don example, you see kind of a what I would call a civil society service supplier. Um but you also see local government-owned uh service suppliers. So that's the example of operator ICT or uh HSL in in Helsinki. Um and then you also see private

service suppliers like magenta. Everybody is familiar with this kind of example. Um but there's you know different uh relationships that governments have with uh with service suppliers. Some of them can be um very kind of hierarchical and top down. Uh in the case of magenta it was kind of a very much a relationship based on kind of mutual exchange. Um, and so there's actually a lot of

participation in kind of the the community uh that that is supplying the solution. They don't just kind of customize and run. Um, so yeah, I think it's really worth looking at and unpacking all of the different dynamics across these different groups. Um, I won't go through all of these in depth because uh I think I'm running out of time a little bit, but if you go into

our report, we have a number of recommendations that kind of surface based on uh our kind of evaluation of these different case studies and how the um archetypes overlap. And so I encourage you to kind of uh dig in and read those. Um but you know things like explicitly defining policy goals um kind of firsthand is really important. Being able to self assess and we hope that

uh kind of using our kind of uh basin framework you can you can do that. Uh helps you map out kind of barriers and challenges and then think about what the different relationships are and the incentives that are involved are. Um you can inventory what capabilities are needed. Uh and then hopefully you kind of think about these different kind of logics and how they intersect to to

begin thinking about what resources you might need and how you might solve a specific challenge for your local use case. Um again uh you know not again but uh designing solutions with interoperability and reuse is uh very important. Um dedicating central funds to support development and maintenance. Uh this was really uh prominent particularly in some of the examples of uh you know kind of public sector institutions

doing this themselves. um striving towards an open and kind of agile culture uh in procurement and development of new software solutions. Um actively considering how to support development and maintenance. Uh the role of local government associations um paying uh attention to the conditions necessary for service suppliers uh to to kind of maintain sustainable business models. You see different ways this is happening for example with magenta in

Denmark or with uh Donovan in Slovenia. Um and then just considering what role kind of existing bodies or associations can play in supporting uh open source adoption, development and collaboration. I was really heartened in particular uh with the console case study around uh Kosla uh which has a very small staff of like one and a half people um that are kind of maintaining and providing uh an

instance of console democracy to all of the local municipalities in uh Scotland um and doing so in a very costefficient Um and then finally just considering who is kind of the the steward um supporting kind of these open source solutions long term um figuring out how you can build that into the planning um and then also in actively involving uh relevant service suppliers in technical governance and

coordination. So you know this is kind of a bit of a a scattershot series of kind of recommendations. There's a more detailed um subset of them across different archetypes in the report. Um I hope you'll consider going in and reading the case studies. I think they kind of map out interesting collaboration dynamics uh and that we've been able to kind of surface some of the the interesting

things that we were able to find through our research. Um ultimately what we hope this is is uh kind of a taxonomy so to speak for starting to think about open source collaboration but one that we hope to build on. Um, we haven't specifically identified further research that we can do in this area, but we really want um to kind of begin kind of uh maybe using

some of this research to kind of map out other case studies or maybe even uh to work with a local government that could say uh evaluate uh what they're going to to do in the future um based on some of these dynamics and see if that assists in kind of uh surfacing some insights around how they could do the planning. So, um, yeah, I I hope that,

uh, you all took something away and I'm happy to answer any questions you might have. >> Any questions for Nicholas? >> Hi, thank you very much for the talk. Um, I'm wondering, um, did you also look into where the impulse for the starting this project comes from? Because I'm wondering about this very much. I'm a fellow vianese. Um, and like I would be very curious like where

where where this would actually start. Would it start on a political level or or from like within the administration or from outside actors? I guess like all of your case studies um answer this a little bit differently, but is there a common thread here? >> Um, yeah, I would say the most common thread is that there needs to be some level of political will. And when I

say political will, I mean like from a policy maker. So uh that can start from the outside in the sense of uh a politician or a policy maker goes to a conference and they're approached and uh or a bureaucrat approaches them with a compelling idea. Um but really kind of the the through line here was that you need kind of senior leadership bought in um very early.

Uh and so you saw this uh most directly with console like a very uh specific example of like we want to bring democracy to the people. How are we going to do this? Oh, we can build this solution and we can do it in 3 months. Um but there's also kind of yeah examples uh within say like operator ICT of how they were able to kind of

really just kind of uh onboard the government and say that uh putting an MIT license on uh this software is actually going to make it easier for us to collaborate across the rest of the government because a lot of us are working in silos. Um there's also examples of uh Donna doing some of the capacity building and training um with kind of policy makers and officials to

help them understand what the solution is, what it means that it has an open license, how they can leverage it, um how they can request features, backup stream, uh that kind of thing. Yeah. >> Hi, thanks for the talk. Um are you aware or maybe have done yourself a similar study for academic organizations >> uh like universities or what do you mean >> Yes. >> Uh no

we haven't. Um but I would be really uh interested to do that and uh I think uh we could probably yeah map out similar archetypes I could imagine at the university level. Um I would say yeah the the proliferation of uh OSOs is becoming more advanced in uh uh the national public sector. We see this through the EC OSPO network. Um and we also uh hopefully we'll

see kind of even more of that with the new open source strategy um that will promote kind of OSOs or OSPO like institutions to support adoption and reuse at that level but it kind of all originated in universities and so um I think this would be a really compelling por uh piece of research. Yeah. >> All right. Thank you Nicholas. Thank