Sketching out the research agenda: open data, governance and developing countries
Building on the outcomes from the ‘Fostering a Critical Development Perspective on Open Government Data’ workshop held in Brasilia in April, and interactions with Jose Manuel Alonso (Web Foundation) and Fernando Perini (IDRC), I’ve been sketching out some initial ideas for a research agenda to take forward the exploration of how open data initiatives can work positively for development.
We’re planning to have a proper public draft of a research agenda ready to share during July, and we’ll be discussing it during on the the workshops at the 2012 Data.gov and World Bank International Open Government Data Conference (July 10th - 12th Washington DC), as well as through other face-to-face forums and here online. Based on these discussions we plan to share a ‘call for proposals’ in late July or early August, inviting researchers in the Global South to propose case study research looking into local, national or regional open data initiatives and their development impacts. Then in October we’ll be meeting in at the Berkman Centre at Harvard to review proposals received, and to put together a team of expert mentors who can offer support to the successful research teams. Working with the selected research projects we plan to establish a research network focused on exploring the impacts of open data in development contexts, with a series of online and face-to-face research network activities over the next two to three years. We hope to make many of those network activities open to the wider community also, and to be sharing regular updates on this blog.
That much we know. However, as we develop the research outline we’ve got a lot of different elements to consider, and so I thought it would be useful to share some of the current (draft) thinking, open to your feedback. You can add comments to this post (look for the comments link), or drop me a line: firstname.lastname@example.org with any thoughts, feedback or questions.
Setting the scope: Towards a framework for researching the use of Open Data to secure better governance in developing country contexts
Open Data is potentially a very broad field, and so we need a conceptual framework to help identify the areas we will focus on, and how we will map out the field.
The Brasilia workshop highlighted that it is important not only to look at open data from the supply side, when governments or other institutions launch open data initiatives, but to also look from the demand side, at cases where grassroots groups, or intermediary organisations, are seeking access to data in order to secure some set of development outcomes and to improve some system of governance. In order to understand the full potential of open data in development it is important for us to explore cases that start both from the supply and from the demand side.
Much of the existing discourse focuses on Open Government Data. This includes an implicit assumes that all the data relevant to securing outcomes of interest from Government, yet in practice this is rarely the case, particularly in development contexts where a wide range of government, NGOs, international agency and private actors may be involved in processes of development. Key datasets relevant to governance may come from many different types of institution, or event crowdsourced from citizens, rather than being officially ‘government data’ (that is, data generated by or owned by government)
Whilst Governments are likely to remain central actors in our exploration, our core interest is not Open Government Data from (as traditionally defined), but the use of open data for governance. Here, we describe some of the key element of this framework.
> A focus on open data
Firstly, in setting our scope we need to define in detail what we mean by open data. As a new phenomenon, we are primarily interested in data which meets the Open Definition (http://opendefinition.org/okd/), being accessible and technically and legally open), but we also want to recognise that relevant datasets might be open by degree, or in more partial ways (Smith & Elder, 2010), and so we should not exclude from our view data that meets some, but not all, of the Open Definition criteria. A minority of such cases in our fieldwork may help in highlighting how technical and legal openness and easy accessibility influence the use of data in governance processes.
There are many different kinds of data that our enquiry may touch upon, including data about government operations (from financial budget and expenditure data, to voting records and meta-data about legislation and legislative processes, and performance data on public service delivery), data released through ‘targeted transparency’ policies (Fung et. al, 2007) about companies & markets (from environmental performance and safety statistics, to regular financial reports from listed companies), and data about citizens (such as census records, educational and health statistic, migration data and other socio-economic information).
The open datasets considered in the research may be ‘big data’, suitable for large-scale statistical analysis, or might be small-scale datasets, from which individual actors can directly extract relevant facts. The datasets will vary in terms of comprehensiveness and currency, with some datasets providing a backward look at past policy implementation or public service performance, and others providing ‘real-time’ information, for example, on public transportation
We also need to understand open data in connection with the technical components of open data initiatives beyond raw datasets, from data structures and data standards, to data catalogues and platforms, and wider frameworks relating to copyright and privacy issues in open data. Many of these broader components are being developed on a global level, driven by leading open data initiatives. In our exploration it will be necessary to explore and understand how these components create opportunities and challenges for the implementation of open data initiatives in developing countries.
> A focus on the process of governance
Theconcept of governance is used in a wide range of contexts, from global financial markets to community-based water management. In general terms, governance can be defined as “the process of decision-making and the process by which decisions are implemented (or not implemented)” (UN ESCAP, n.d.). Lynn et. al. (2000) describe governance as referring to “the means of achieving direction, control, and coordination of wholly or partially autonomous individuals and organisations on behalf of interests to which they jointly contribute.”
Political governance may refer to different parts of the political system of a country, from the definition and monitoring of budgets in national or local institutions of government, to decision making in the legislative process and election processes. Economic governance is often related to the specific firms and markets that are regulated by the government (such as financial and extractive industries) and/or services that governments deliver, subcontracted or subsidize such as transport, education and health. Social governance is generally focused on specific social issues, such as the empowerment of specific marginalized groups, such as women, illiterate groups, youth, poor community or racial minorities. In order to understand how open data initiatives are embedded in political, economic and social contexts, we intend to explore in greater detail the process of governance in different areas, and the changes that open data initiatives are creating.
In the first phase of this research effort, it is likely that we will need to focus on some specific governance themes and issues (for example, budget monitoring; land ownership; local community empowerment) in order to build a more comparable and in-depth understanding of specific domains. Although we are likely to highlight a few areas in the open call, we also would like to build on the presented proposals as a reflection of the interest of the research community.
> A focus on emerging outcomes
Open data can impact on governance in a range of different ways. Arguments in favor of open data highlight a number of different outcomes that the release of open data may lead to. These include:
- Supporting greater transparency and accountability
- Stimulating innovation, efficient markets and economic development
- Promoting inclusion and empowerment of groups often excluded from the policy process
The extent to which open data initiatives realize these emerging outcomes will depend on a wide range of factors – relating to the qualities of the open data initiative, and a range of wider contextual factors. Identifying the factors that enable or constrain realization of these emerging outcomes in developing country contexts is at the heart of our research.
The emerging outcomes of open data have been put forward as both intrinsically and instrumentally valuable. For example, greater transparency, like greater democracy, is often taken to be a intrinsic good in society, although it is also treated as an instrumental value to promote better governance or a range of other social goods (Fung et. al., 2007). As well as looking at how far greater transparency and accountability result from open data, or the dynamics of open data promoting inclusion and empowerment, in each case the research explores we will also be asking about any substantive development impacts that were anticipate from the open data initiative, and cases will explore the evidence on how far those impacts have been realized, and the enabling and constraining factors in moving from the emerging open data outcomes, to creating change in specific areas, or on specific issues.
> Diverse, but comparable cases
Putting these together, we can think about the our cases in terms of the nature of open data used, their interaction with the governance systems in developing countries, and their emerging outcomes and impacts. This should give us a broad framework where cases may fit, and allowing us to cluster and explore comparisons between cases.
Applying the framework
Based on this general framework, we intend to find a range of cases (likely between 5 and 10), and to construct a mixture of flexible, and more formal, tools to capture data within them - supporting cross-case analysis, whilst ensuring that local research also generates value for the local context.
The idea is to announce an open call for proposals with the following characteristics:
1) Case study centered on the governance of relevant developmental issues
Whilst we anticipate using a range of methods at the overall project level, the bulk of the research will happen through in depth case studies, conducted over time, in settings across the developing world. Case studies are well suited to “investigate a contemporary phenomena in depth and within its real-life context”, coping well with “technically distinctive situation[s] in which there will be many more variables of interest than data points” (Yin, 2009).
Based on the general framework suggested above, the cases should centered on understanding the (i) the characteristics of open data currently available within the specific case and its relation with global initiatives that aims greater data openness in this specific issue, (ii) the context and structure of the governance in the specific area of development, including a diagnostic of some of the main strengths and weaknesses of governance area under study; (iii) an in-depth analysis of the emerging outcomes of open data initiatives within the specific area of governance paying special attention to the factors that foster or hamper greater transparency and accountability, better economic efficacy and efficiency and greater inclusion and empowerment of marginalized groups. The focus of cases studies should be on describing and analyzing the complex reality of focus open data initiatives, rather than performing program evaluations of open data initiatives.
Case studies should employ multiple forms of data collection, and we anticipate cases will use a mix of documentary analysis, interviews, surveys, participant observation and action research to explore a specific supply or demand driven open data initiative. We’re particularly interested in finding ways to visualize the networks of actors, datasets, technical artifacts and other elements involved in open data initiatives, and to identify data collection and analysis methods that help us deal with the complexity of interactions involved in open data in developing countries.
2) Policy and intervention oriented (at the local and global level)
Value will be given those cases with a strong policy relevance, both in terms of its local and global application. Case studies may take place in a single country, or taking a comparative approach to look at a particular issue of open data in governance (e.g. budget monitoring) across a number of countries. Case research is expected to take place over 6 to 12 months for the first round of supported projects.
Projects may include elements of action learning, or applied research, with the project research team engaged directly or indirectly in the open data initiative being studied.
We also anticipate that projects will have a strategy for local dissemination of findings to input into local policy and practice debates.
3) Mentoring and peer-support
We will be looking to provide each project with a mentor and an offer of peer-support within the research network. This may focus on the practicalities of an open data initiative (where an expert mentor could offer support to the development of an initiative), or may focus on support for the research, data collection and analysis process.
The structure of the mentorship scheme will be addressed in another future note.
4) Beyond individual case studies: building cross-cutting M&E indicators
We will also be looking to develop some other cross-cutting data collection instruments and forms of analysis to help us answer key research questions. These might include policy reviews of national open data policies; gathering metrics on datasets use and re-use; secondary analysis of documented cases of data use; and drawing on existing indicators or current data collection efforts (such as the Open Data Census) or the Web Index, to build up a more global picture of the developmental impacts of open data.
In addition to exploring the different contextual factors within each particular case of open data use, we are interested in how wider national (or regional) contextual factors may impact upon the realisation of developmental impacts from open data. The Web Foundation have looked at six key sets of contextual factors for and open data initiative: political, legal, organizational, technical, social and economic. These categories can guide us in developing appropriate cross-cutting measures, and descriptive accounts, of the context in which each open data initiative is operating.
The design of this part of the study will be addressed in a future note.
Case studies in summary
Taken together then, we would be looking for proposals that include:
- the development of case studies of open data use in developing countries that start from both ‘supply’ and ‘demand’ sides, placing a central focus on data used in a specific area of governance
- in depth case studies exploring an open data initiative (supply or demand driven), its interaction with existing or emerging governance structures and highlighting observed outcomes (positive and negative). They may include analysis of the networks of agents, artefacts, agendas and actions involved in creating, using and mobilising open data over time.
- a focus on cases in progress (open data initiatives already implemented or under implementation) involving datasets that meet (to a large extent) the Open Definition. We are looking for cases that can inform policy in developing countries around the adoption of open data for good political, economic and social governance and developmental impact.
- the overall project will cover a variety of types of data, processes of change, and domains of impact. Each case will have a specific focus, but we anticipate that together will cover a range of political, economic and social impacts, processes of change, and types of data.
We need your views
Everything above is incomplete and draft. It’s not fixed, and will be evolving throughout the coming month(s). We will be talking to people, reviewing the literature, and testing out lots of possible ideas, and so we would welcome your thoughts and input too.
If you have any ideas, questions or suggestions (from pointers to relevant literature, to substantive points on the outline above), then please do get in touch. You can leave a comment on this post, or e-mail email@example.com
Thank you for taking the time to read and contribute to this project. Keep an eye on the front page of the website for further updates and announcements.
Fung, A., Graham, M., & Weil, D. (2007). Full Disclosure: The Perils and Promise of Transparency (p. 282). Cambridge University Press.
Joshi, A., & Houtzager, P. P. (2012). Widgets or Watchdogs? Public Management Review, 14(2), 37-41.
Lynn, L. E., Heinrich, C. J., & Hill, C. J. (2000). Studying Governance and Public Management : Challenges and Prospects. Journal of Public Administration Research and Theory, 10, 233-261.
Smith, M., & Elder, L. (2010). Open ICT ecosystems transforming the developing world. Information Technologies and International Development, 6(1), 65-71. Retrieved from http://itidjournal.org/itid/article/viewFile/489/214
Yin, R. K. (2008). Case Study Research: Design and Methods (p. 240). Sage Publications, Inc.