The role of open government data and co-creation in crisis management: Initial conceptual propositions from the COVID-19 pandemic

1.Introduction

There is a growing body of research on the topic of open government data (OGD) and research has explored how the usage of OGD may bring about numerous benefits (Janssen et al., 2012; Kucera & Chlapek, 2014; Safarov et al., 2017; Ubaldi, 2013). One benefit that is often discussed in relation to OGD is its ability to drive data-driven innovation and the development of new services (Janssen et al., 2012; Jetzek et al., 2014). For example, research has shown that OGD can drive and enable new digital co-creation processes for digital public service development, which ultimately have the potential to create public value (Khayyat & Bannister, 2017; McBride et al., 2019). While there are numerous approaches to conceptualizing and understanding co-creation, in this paper, co-creation is understood as “a process through which two or more public and private actors solve a shared problem, challenge, or task through a constructive exchange of different kinds of knowledge, resources, competences, and ideas that enhance the production of public value” (Torfing et al., 2019, p. 55). Unfortunately, research that empirically examines the OGD-driven co-creation process is still rather limited, providing a clear opportunity and need for research.

It is from this need for research that this paper emerged. It offers new understanding into how OGD enables and drives the co-creation of new services and, in the process, creates public value within the context of a crisis – specifically the COVID-19 pandemic. The research started from several observations: in previous crises (such as floods and forest fires) OGD had seen use (Azevedo et al., 2015; de Mendonça et al., 2015; Shidik & Ashari, 2014), that during the COVID-19 pandemic OGD was being released in large amounts, and COVID-19 related OGD were being reused consistently. During the earliest stages of the pandemic research was conducted that explored the adoption and implementation of data-driven innovations and services during the COVID-19 pandemic (Brem et al., 2021; Dahlke et al., 2021; Grinin et al., 2021; Sampaio et al., 2022; Nikiforova, 2020, 2021; Alamo et al., 2020; Maione et al., 2021; Khan, 2021), but this research often focused only on new services, datasets released, or the potential value that has been created from these services. Research that explored the role that OGD specifically played in the crisis management process was missing.

While there are several different understandings of a crisis, for example both a political scandal or a natural disaster could be considered a crisis, this paper is concerned primarily with crises that have a potential to directly and negatively impact large number of individuals; this would include natural disasters, conflicts, or – like COVID-19 – public health crises. The common factors for these crises being that there is a large divergence from normal operating conditions, there is risk to wellbeing, and that these risks must be managed quickly and, often, with only limited information available. In such a situation, OGD may play an important role by aiding in the creation, collection, and dissemination of information, thus creating new opportunities for service co-creation, potentially aiding in the management of a crisis.

To generate new insight into how OGD was used during the COVID-19 pandemic as a crisis management tool a comparative exploratory case study was conducted, focusing on three Central and Eastern European (CEE) countries: Czech Republic, Estonia, and Latvia. The CEE region is united by a shared history, traditionally lower levels of administrative capacity, and the relative newness of their public administrations (Bouckaert et al., 2011; Randma-liiv & Drechsler, 2017). Within this context, technology is often embraced to subsidize missing capacities, unsurprisingly, many CEE countries are viewed as leaders when it comes to the use of digital technologies by the public sector (Randma-Liiv & Drechsler, 2017). In the case of the selected countries for this research, all three belong to the very high e-government development index grouping in the United Nation’s e-Government Development Index.11 Taken in totality this provides an ideal context for exploring how OGD could be used during a crisis – states that are already used to using digital technologies in their administrations, often to overcome strategic shortages or skill gaps in their administrations, which may provide incentives for administrations to engage in co-creation.

With the cases selected, two primary research questions were formulated and used to drive the study:

In conducting the research to answer these questions, three clear contributions emerged. First, this paper provides new insight into the role that OGD may play during times of crisis and offers initial conceptual propositions about the role of OGD in the crisis management process. As a second contribution, this paper provides new empirical exploration into how OGD drives the co-creation of new services. Third, the paper empirically demonstrates the potential that OGD has when it is released and used. Ultimately, this paper aims to be of use both to practitioners who may be directly involved in the management of a crisis or in the management of an OGD initiative, for academics interested in studying the impacts of OGD or crisis management, and also for citizens to see how they can play an active role in co-creation processes and thereby create value.

2.Background literature

2.1Crisis management

A crisis can be defined as “a serious threat to the basic structures or the fundamental values and norms of a social system, which under time pressure and highly uncertain circumstances – necessitates making critical decisions” (Rosenthal et al., 1989, cited in Rosenthal & Kouzmin, 1997, p. 280). Crisis management can be further understood as “the processes by which an organization deals with a crisis before, during, and after it has occurred” (Christensen et al., 2016, p. 888).

A crisis is a strenuous time for any public administration, as they often have “prepared neither themselves nor the public for appropriate responses once tragedy strikes” (Rosenthal & Kouzmin, 1997). Additionally, bureaucratic institutions that are slow by design are now required to work in an increasingly adaptive way (Janssen & van der Voort, 2020) with time-sensitive and critical decisions needing to be made – often without discussion. This forces institutions and administrators into the realm of the unknown, where laws, regulations, and standard procedures are no longer valid (Boin & Lodge, 2016). In these situations, it could be said that crisis management would be better handled by organizations that are able to be more adaptive or flexible with their arrangements (Boin & McConnell, 2007; Comfort, 2007). One way that has been highlighted in the literature to improve adaptability and flexibility is through the adoption and usage of digital technologies, particularly those that allow for data and information to be gathered, analysed, and utilized by decision makers (Comfort, 1993; Moynihan, 2008; Pan et al., 2008). If true, then one key component to any crisis management strategy must involve clear processes for obtaining, releasing, and using data and information.

Figure 1.

Crisis management process (Pearson & Clair, 1998).

When looking at the actual process of managing a crisis there are several ways it may be conceptualized. For example, in the context of the COVID-19 pandemic, Zheng et al. (2021) draw on Fink’s crisis lifecycle model that includes four stages: (1) prodromal, (2) acute, (3) chronic, and (4) resolution, which maps out the entire crisis process from start to resolution. That model is based, at least in part, on research from Pearson and Clair (1998) that conceptualized the crisis management process that integrates viewpoints from “psychological, social-political, and technological-structural perspectives” (p. 59). For Pearson and Clair (1998), the crisis management process (shown below in Fig. 1) is influenced by the environmental context and administrative crisis preparations, and consists of the planned responses to a crisis as well as the individual and collective reactions to a crisis. For this paper, the model proposed by Pearson and Clair (1998) was selected as the basis for understanding the crisis management process. How this model relates to OGD and co-creation is further explicated in the following section.

2.2OGD, OGD-driven co-created services, and COVID-19

2.2.1OGD and the COVID-19 pandemic

OGD is data provided by a government, under a reusable license, in human- and machine-readable format, released without discrimination or cost to the public (Ubaldi, 2013). Research dealing with OGD has often found that there are many benefits associated with the use and release of OGD, such as economic benefits, transparency, accountability, innovation, etc. (Hansen & Schrøder, 2019; Janssen et al., 2012). However, as Janssen et al. (2012) have highlighted, in order for OGD to actually create some value or benefit, something must be done with it. During the COVID-19 pandemic OGD was used. It created value.

One of the most comprehensive reviews on the use of OGD during the COVID-19 pandemic emerged from a cooperation between TheGovLab and the OECD, which released a report titled “Open Data in action. Initiatives during the initial stage of the COVID-19 pandemic”. That report argued that the COVID-19 pandemic “increased the demand for access to timely, relevant, and quality data” and that “during the initial stage of the COVID-19 pandemic, many governments were active in promoting OGD, both in terms of releasing data and re-using them to build different types of data products” (Gonzalez-Zapata et al., 2021). The studied applications and services built on top of OGD were, primarily, focused specifically on the public health aspects of the pandemic and were primarily citizen focused and oriented towards providing updates and situational awareness (Gonzalez-Zapata et al., 2021). Though many of these services were being supported, built, and maintained by governmental organizations, many non-governmental stakeholders also played an active role. An interesting example of this early in the pandemic was seen in Italy (Fig. 2), when an OGD user remarked on 10 March 2020 that the government had made data available, numerous developers were already contributing, and that it was something previously unseen at such a scale.

Figure 2.

Tweet showing interest in OGD and co-creation at earliest stages of the pandemic. Source: https://twitter.com/Zizzamia/ status/1237258789973188608.

Figure 3.

OGD-driven co-created public service system. Source: (McBride et al., 2017).

Similarly, academic organizations also began to quickly collect, gather, clean, and provide new insight based on OGD from around the world. For example, the University of Oxford launched a Government Response Tracker, Johns Hopkins University quickly created a map to easily visualize the spread of COVID-19 around the world, and journalist organizations such as the Financial Times and the New York Times used OGD to conduct statistical analyses and provide daily updates to concerned readers about COVID-19. OGD was being released, OGD was being used, services were being co-created from OGD, and OGD was creating value at a global scale.

2.2.2OGD-driven co-created services

One clear way that value can be created from OGD is through the development of new services, and this is especially the case when services are co-created. There is increasing research focusing precisely on services of this nature – OGD-driven co-created services (Jarke, 2019; Khayyat & Bannister, 2017; McBride, Aavik et al., 2019; Mergel et al., 2018). An OGD-driven co-created service may be defined as a service that is “producing public value, that rely on OGD, and have been developed by stakeholders from different groups coming together in a co-creative process” (McBride, Aavik et al., 2019). The OGD-driven co-creation process can be conceptualized within a broader ecosystem (Dawes et al., 2016; Harrison et al., 2012), that consists of the ICT tools used to build the service, the data, the co-creator(s), and the environmental factors that influence this ecosystem (McBride, Toots, et al., 2019).

OGD is often viewed from an ecosystem perspective, with Dawes et al. (2016) clearly outlining the dynamics that exist within such an ecosystem. In this paper, the focus is on the OGD co-creation ecosystem, rather than OGD more broadly, and is shown above as Fig. 3. This ecosystem model was developed as part of an international research project, developed iteratively, and empirically tested and validated with several cases (McBride et al., 2017). The figure shows that the co-creation process requires a co-creation system that consists of ICT tools and a group of stakeholders who will perform the co-creation, is influenced by organizational beliefs, regulations, funding, and previous experience, and that by “co-creating” a new OGD-driven public service it is possible to create value from OGD.

2.3Summary and initial conceptual propositions

Reflecting on the existing knowledge and relevant background research a clear line of argumentation begins to emerge. First, crises are hard. To overcome these challenges organizations must have data for the decisions they need to make, and they must have the capacity to make decisions based on this data. As demonstrated during the COVID-19 pandemic, OGD can play an important role in the crisis management process. However, for the future, this process must be better understood, institutionalized, and strategized as it is likely the case that OGD may be used as a tool to better navigate future crises.

Starting with the process outlined in Fig. 1, it is possible to identify two concrete areas where OGD and co-creation with OGD occur in the crisis management process – in the stakeholder responses and in the planned/unplanned responses. When a crisis occurs, individual and collective reactions are triggered (the co-creation of new services would fall under this dynamic) additionally there are the planned and ad hoc responses (the release of OGD would fall under this dynamic). Thus, the model can be extended within the context of OGD as shown above in Fig. 4.

Figure 4.

Open government data driven co-creation during crises. Source: Authors, extended from (Pearson & Clair, 1998).

The effect that OGD will have on the crisis management process will also likely be influenced by the environmental context, for example, does the organization have experience with releasing OGD or co-creation?

Based on this initial overview of the literature, and in attempting to synthesize the initial insight from the theoretical and contextual background, several themes and arguments can be identified:

• • first, the literature argues that OGD can help governments respond to crises by decreasing barriers for information exchange and increasing the rate and span of information exchange (Comfort, 1993; Moynihan, 2008; Pan et al., 2008);

• • second, the availability of OGD during crises may drive the development of new ecosystems that focus on the co-creation of OGD-driven services (Toots et al., 2017), especially during times of crisis;

• • third, in this newly created ecosystem, stakeholders will naturally emerge to engage in OGD co-creation together with the government to help manage a crisis (Scognamiglio et al., 2022);

• • fourth, during a time of crisis, these services are likely initiated by the release of OGD from the public sector, driven and developed by non-governmental stakeholders, and likely focused externally towards information dissemination (Gonzalez-Zapata et al., 2021).

These arguments will form the basic framework for analyzing the cases conducted for this research and it is through this discussion and exploration that the research questions may be answered so that the paper’s contributions may emerge.

3.Methodology

For this research, a holistic comparative exploratory case study approach was used (Yin, 2018). Exploratory case studies are especially useful when there is a need to develop new hypotheses and propositions, particularly when the issue of study is a contemporary issue with limited empirical information available (Eisenhardt, 1989; Yin, 2018, Chopard & Przybylski, 2021). In this paper the unit of analysis is the usage of OGD during COVID-19 pandemic, not each individual application or solution. Rather, this paper is concerned with the overall ecosystem and process within each country (Czech Republic, Estonia, and Latvia).

While the case selection was justified and discussed previously, it must be highlighted that in the case of both Estonia and Latvia authors of this research had direct involvement with the country’s OGD initiatives surrounding the COVID-19 pandemic. While this does raise potential issues for the objectivity of the research, it is also well established that having direct insight and observation of the object of study offers important insight that may not otherwise be available to the scholarly community (Yin, 2018). To prevent any potential for bias numerous sources of empirical evidence were used to triangulate the findings. Furthermore, during the data analysis for each case the data was reviewed individually by each of the authors and the conclusions drawn compared to further improve objectivity in the analysis of the results.

The data utilized included: the OGD provided related to COVID-19 within each respective country, physical artifacts (the OGD-driven co-created services and applications built during the pandemic), government documents, press releases, statistical reports, news articles, participant involvement and direct observations. Data were collected and archived over the more than two-year duration of this study (March 2020 to September 2022). To analyze the data, several approaches were used. For the physical artefacts and the OGD released particular attention was paid to what data were being released, the size of the data, number of datasets available on the portal, the datasets being used by the application, the contributors to the application (for example, by monitoring public GitHub pages), and the usage of the application. Textual data, e.g., news articles and press releases, were analyzed for mentions of either OGD or the co-created applications. By analyzing data in this way, it is possible to construct a timeline of the pandemic and its management, including key events such as political and policy changes as well as releases of OGD or the launch of a new co-created application.

4.Case overview

5.Discussion

This paper began by arguing that OGD was likely able to play a role in the management of crises by enabling opportunities for OGD-driven service co-creation. However, the process in which OGD could be used in the crisis management process was not yet well defined in the literature. This paper remedies this gap by conducting an exploratory and comparative case study into the topic and, because of this exploration, it has been possible to generate new insight and knowledge. As a starting point for this analysis, Table 1 has been prepared, which lists the services identified during this study, the stakeholders involved, and whether the service continued to function throughout the duration of the study. The included services only represent services that were co-created during the pandemic, and those that utilized OGD, while every attempt has been made to be as comprehensive as possible, it is always possible that other services exist that meet both requirements and were not included.1111

To better structure the discussion and the analysis of the cases, the four arguments offered in the background section of this paper are used and presented in the following paragraphs.

The first argument emerging from the literature was that OGD could help governments respond to crises by decreasing barriers for data and information exchange as well as increasing the rate and span of information exchange. Based on this research, there does appear to be support for this argument. In all countries usage of the OGD portal saw clear increases when COVID-19 data were made available and the related datasets also saw large amounts of interaction (e.g., in numbers of downloads). In all countries studied again there were demonstrable examples of how stakeholders outside of the government utilized OGD in some form to either generate models for decision makers or to disseminate information. Thus, it does appear to be the case that OGD does play an important role in enabling new forms of information dissemination during a crisis.

The second argument emerging from the literature was that the availability of OGD during a crisis may drive the development of new ecosystems that focus on the co-creation of OGD-driven services. This finding appears to be only supported in part, and with an important caveat.

For an OGD-driven co-created ecosystem to be sustainable, continual government support and involvement is important.

While, in theory, governmental organizations do not need to play an active role in the co-creation of new OGD-driven services, their support does appear to be critical in maintaining these ecosystems over time. As seen in Table 1, in a clear majority of services studies government was a key or participating stakeholder. This has been seen in previous research as well, where research has shown that “government-led initiatives tend to survive longer than citizen-led initiatives after controlling for multiple mediating factors” (Cingolani, 2021). This observation was also seen in the studied cases. In Estonia, for example, there was a large amount of support from the government at the beginning for co-created solutions, once this support disappeared the usage and maintenance of the co-created systems declined. Similarly, in Latvia and Czech Republic, where governmental support was minimal from the beginning for co-created applications, the impact and quality of the solutions tended to be lower. This also seems to highlight the temporal component for both OGD co-creation ecosystems and public value during crises where, at the earliest stages of a crisis, there is a higher level of co-creation and support, it has a plateau, and then begins to decline over time as the crisis resolves and support for the ecosystem begins to dwindle.

The third argument emerging from the literature was that in OGD co-creation ecosystems, stakeholders will naturally emerge to co-create new solutions together with the government to aid in the crisis management process. In all three studied cases there were clear examples of how the release of OGD activated interested stakeholders to start co-creating OGD-based solutions together in partnership with the government. Thus, there does appear to be some support for the idea that in a crisis, when OGD is released, co-creation can naturally emerge between stakeholders. However, in all three cases, the actual number of people working on such solutions was low, and those involved are those who have either previous experience with OGD, or those who are already directly involved somehow with crisis management via their occupation. Thus, the success of a given OGD co-creation ecosystem seems to be additionally highly dependent on the presence of individual motivated actors with a higher level of digital skills and competences, particularly those that enable them to effectively work with OGD.

The final argument emerging from the literature was that during a crisis co-created services were likely to be built on top of OGD, driven and developed by non-governmental stakeholders, and likely focused externally on information dissemination. In all three countries studied during this research support was seen for this proposition. However, it is also the case that the government was often playing an active role in the provision of OGD, and the non-governmental stakeholders were more involved with the design and development of the service. This finding seems to hint that it may be the case that certain stakeholders have certain proclivities towards specific roles in the co-creation process. When looking at the actual nature of the services themselves, in the studied countries, most services studied were focused on information dissemination, for example dashboards. Thus, this argument may potentially be reframed as follows: during a crisis governmental and non-governmental stakeholders will engage in co-creation, enabled by OGD, and that the OGD-driven co-created services will primarily focus on information dissemination and visualization contributing to the understandability of the data/information.

Drawing on the above findings, it is possible to answer the primary research questions of the paper. The first question asked what is the role of OGD in crisis management. It can be said that OGD does help public sector organizations with crisis management and that OGD helps by providing a way to increase information availability and dissemination. This is clearly supported by the conducted research. In all three countries OGD was released and made available to the public via the country’s national OGD portals. OGD that was released was used to develop solutions that could be presented to both decision makers and the public to help provide an overview of the current crisis situation, for example, by displaying the number of hospitalizations, the total vaccination rate, or the number of people infected in certain specific geographic regions.

The second research question asked how does OGD influence the co-creation of services that assist in crisis management? To answer this question, it can be said that OGD enables new co-creation ecosystems, and that these ecosystems allow for different stakeholders to cooperate and co-create new services that solve problems directly related to the management of a crisis. This answer is also supported clearly by all three studied cases. The best example of a co-created solution comes from Estonia, whereas the examples in Latvia and the Czech Republic appeared to have lower levels of uptake. There are at least two reasons for this that are apparent from the research. First, in Estonia there was a strong governmental commitment at the earliest stages of the pandemic to utilize citizen co-created solutions, due to missing governmental capacity, this political support was not present in the other two studied countries. Second, in Estonia there were a larger number of stakeholders involved, spanning academia, the private sector, government, and the media, whereas in the other two countries studied, the number of stakeholder groups included was smaller. This demonstrates how previous government experience with co-creation, government commitment to co-creation, and higher levels of involvement from non-governmental stakeholders leads to better co-creation outcomes.

Figure 5.

Crisis management with open government data availability and government co-creation involvement continuum. Source: Authors.

Finally, based on the initial exploration of the cases, it is possible to identify a matrix that maps together OGD, co-creation, and crises. This matrix is shown below as Fig. 5 and consists of two primary dimensions. The first is the involvement of government in the OGD co-creation ecosystem and ranges from inactive to passive to active. The second dimension is the availability of OGD, this ranged from no OGD availability to partial openness to full openness. Important to note for this dimension is that it must also consider the quality of the data that is being provided, for example if a dataset is not accurate, poorly documented, not up-to-date, or not maintained, but it is available, it must not be considered fully open. Depending on where a specific government is within this matrix when responding to a crisis, a different level of impact is likely to emerge, with a different level of sustainability of the co-created services. Importantly, as highlighted in this discussion section, there also appears to be a temporal nature to a country’s path through this matrix. At the earliest stages of a pandemic, when a government is releasing and actively involved in releasing information, it may rapidly shoot up to the top right corner of the matrix, thereby creating a higher level of impact. However, over time as involvement slows, it may start to fall towards the bottom of the matrix, demonstrating a decrease in impact.

One potential explanation for this temporal component can be found in the previous literature on crisis management. At the start of a triggering event, there is not much information available, and many stakeholders are involved as the government begins to handle the crisis. At this stage, many stakeholders may get involved and try to assist – this is especially true when limited information is available. As the crisis goes on, and more data get released as OGD, some initial co-created services will emerge. However, as the crisis continues to go on, and the government gains more control of the situation, the government may begin to reassert itself as the core stakeholder and become passive or inactive in the co-creation process.

6.Conclusion

The results of the research presented in this paper make important contributions for both crisis management and digital government related research by generating new understanding on the relationship between OGD and crisis management. These contributions are made in three core dimensions. First, the paper directly integrated OGD into the previous literature on crisis management, identifying where and how OGD can be used by crisis managers and argues that in addition to increasing the availability of information, OGD is also used in the co-creation of services that subsequently create value for decision makers and citizens alike. While previous research has explored ideas relating to the use of digital technologies for the co-creation of services (Rodriguez Müller et al., 2021; Lember et al., 2019), such research did not directly explore how co-created services could be used as a crisis management tool. Second, the paper builds on existing arguments in the extent literature on OGD and crisis management and tests their veracity through an empirical comparative case study. The third contribution is related to its empirical demonstration of how OGD was released and used in co-creative processes to create important services.

For scholars of crisis management, this paper highlights the potential benefits that engaging with non-governmental stakeholders may have for the management process, particularly in areas dealing with information and data availability. As the availability of data increases, openness becomes increasingly prevalent, and familiarity with data analysis increases it is increasingly likely that co-creation and OGD will become a core tool in the crisis managers toolbelt. Similarly, for scholars of OGD and digital government, this paper provides important insight into important possibilities of OGD reuse, but also highlights the importance of context and governmental support for co-creation.

In addition to the above-mentioned academic contributions, there are also several important recommendations that can be made for practice:

Though this research does make a number of new contributions to the literature, there are several opportunities for future research that must be explored. During the analysis of the cases two potential risks came to light when OGD was used as a crisis management tool. First, when governments released data in an open format at the earliest stages of the crisis they were often accompanied by mistakes in the data, often due to inaccurate data collection methods or improper data handling techniques. This has an unfortunate effect of lowering trust in the government’s ability to properly manage the crisis. Thus, there appears to be a dynamic at work between a government’s desire to release data in an open format, and their need to appear in control of the situation and in the accuracy of the data being used to drive their decision making. The “quality over quantity” principle seems to remain valid and crucial for the overall success of OGD initiative in the crisis management context (Sadiq & Indulska, 2017). Second, when governments opened up data during the pandemic it was no longer possible to control the narrative or communication strategy for how the crisis was being managed. Future research should explore these dynamic struggles between the desire of the government to use OGD as a crisis management tool, with the desire of the government to maintain trust and control. Similarly, in all three cases there were unclear processes in place for how to engage in co-creation, officially, with non-governmental stakeholders. Further research could be conducted that outlines how to best integrate non-governmental stakeholders into the planned response of government to a crisis.

Another important opportunity for future research is in relation to OGD-driven co-created services themselves. In the current research most of these services were focused primarily on visualization, but there are likely other forms of services that could be utilized and developed to assist in crisis management It is important to establish an understanding of these different forms and how they may be created. In the research several different stakeholder groups were identified as having taken part in the co-creation process, however, exactly which role each stakeholder group played in the co-creation was not well elucidated. It may be the case that certain stakeholder groups are predisposed for certain roles in the co-creation process, for example perhaps the government mainly plays a supervisory role as the data provider, but future research should explore this in more detail. Finally, as a third aspect of future research, there is a clear need to both expand the comparative aspect of this study, moving outside of just one administrative context, and to additionally explore the usefulness of OGD in different crises (for example, financial crises).