Workshop Report: Exploring the integration of data-intensive analytical skills in public affairs education



About the Workshop

Data, Technologies & Context: How the practice of government is changing

Preparing Government Professionals for a New Context

Advancing Policy Informatics in Public Affairs Education

Next Steps

Workshop Participants

Preparing Government Professionals for a New Context

The examples above illustrate the new contexts government professionals at all levels will face in the coming years. They will need to understand and manage different types of data, create new organizations or programs focused on analytics, select tools to support computation or public engagement, and engage in data-intensive policy analysis and evaluation. Policy informatics, as defined above, can act as a useful bridge between research and practice in the areas of technology, data, analytics, and policy analysis and management. Academics need to study current and emerging practices in their research in order to shape curricula in public policy and management to meet the needs of the practice community. In turn, the practice community can take advantage of insights from research to help address pressing public problems and should be able to rely on our graduate programs to produce professionals who are well-versed in advanced design and analytical approaches to public sector governance and management.

Challenges for Practice

The presenters and workshop participants discussed the ways that current data, analytical, and technical capabilities are changing the nature of government as a ‘user of data.’16 Several themes emerged from the presentations: (1) public problems vary widely in content and complexity and thus have different data and analytic needs, (2) stakeholders play, and will continue to play, various roles in the administration of program and policy making process, and (3) public affairs graduates, regardless of specialty or career goal, will need to work with a variety of technical and policy specialists.

Finding and using relevant data

Finding “good” data that will answer important practice and policy questions and does not cost a lot has been a perennial problem for government. Government is one of the largest collectors of data on a vast range of topics, but data are usually collected for specific purposes related to programmatic needs or compliance with rules or statute. The data are dispersed across different departments and protected by various laws regarding collection, access, management, and use. Today new sources of data from outside of government (e.g., social media networks, sensor data, or text) can often be combined with government data. Practitioners need to know how to find data sources across departmental and jurisdictional boundaries, understand the limitations of the data they find in respect to its intended use, and assess whether the use of new data sources is feasible or to what extent combining different classes of data produces the kinds of analysis they need. And, they will need to answer these kinds of questions while operating within and balancing the legal and ethical parameters of appropriate use of government information.

Applying tools and analytic techniques that fit the situation

Complex policy problems can be big or small, broad or narrow, acute or chronic, unique or ubiquitous. Different analytic tools, computational techniques, and technologies will fit different situations. The data and tools available to address an immediate crisis tend to be those readily at hand, even though they are probably incomplete or otherwise flawed. By contrast, developing a major piece of legislation is likely to occur over a longer time period with the opportunity to search out or even collect appropriate data and apply a variety of analytical techniques to test different policy choices. Some techniques, such as visualization, pattern matching, or geospatial analysis may be applicable in many kinds of policy domains and settings. Others, such as the integrated models Millennium Institute uses to forecast the challenges of sustainable development are tuned to a certain kind of problem that demands complex understanding of different scenarios now and in the future. Practitioners will need a diverse ‘tool kit’ of tools and techniques and have a good understanding of the benefits and limitations of each.

Communicating and engaging with a range of stakeholders

Workshop speakers emphasized the importance of being ‘good communicators,’ particularly about translating messy, complex problems into more meaningful and manageable areas for discussion with stakeholders and leaders. Translating the data, analysis, and models in ways that ‘keep the right balance of detail’ for the decisions at-hand, but conveys the limitations, assumptions, and holes in the data is an increasingly important skill. The recent open government movement, combined with new governance practices, promotes engagement of a range of stakeholders as essential to the policy making process. Policy modeling tools are very useful for engaging experts and lay stakeholders in the design and implications of various policy options. This kind of engagement helps create understanding, creativity, and buy-in. However, explaining to average citizens the efficacy of different policies or the limitations of data and technology in making choices is different than explaining it to experts in the field or to legislators. Workshop participants emphasized that the ability to communicate clearly and meaningfully in these different situations is an essential skill for responsible policy informatics work. Public affairs graduates need to be able to identify and address questions about the ensemble of technologies, data, and policies so that they are better able to manage new programs, innovations, and experts who use these technologies.

Working across specialties

Using technologies, analytics, or modeling to address problems requires the ability to assemble and work in multi-disciplinary teams. Public managers, data analysts, subject matter experts, and policy makers need to work together in situations where their different kinds of knowledge and expertise can be focused jointly on problems. Some actors will have more technical expertise in coding, mathematics, visualizations, modeling, and technology, others in policy relevant information, or organizational and implementation considerations. Policy informatics takes all these views into consideration and thus helps not only to identify the different kinds of expertise that are salient to a problem but to see how they can complement or conflict with one another in various policy scenarios.

Challenges for Teaching

Policy informatics provides an opportunity to teach students the importance of understanding data in the broader social, economic, and political context. Workshop discussions focused on four areas for consideration: (1) the importance of teaching for ‘the real world’, (2) providing students with a broad appreciation for data and information, (3) providing access to robust tools and technologies, and (4) finding ways to connect and balance policy informatics competencies with core curriculum requirements.

Finding real cases, using real data, and dealing with realistic levels of complexity

Effective use of data and computational tools for problem solving demands attention to situations, assumptions and dynamics that reveal the complexity of the problem and the suitability of different interventions. Speakers urged that analytical projects and assignments should use real world situations, not made up problems, and apply existing available data, not artificially constructed data sets. The open government data movement provides an opportunity to do this as large numbers of datasets are being cataloged and released for public use. Most of these data resources are rich in content, but they have also limitations and flaws that students must learn to address in their analyses. Cases can come from any level of government and any policy domain, as long as they are reasonably good representations of the interests and conflicts that are at play. These kinds of activities and resources prepare students for the likely issues they will face when they take positions in government and will give them a realistic dress rehearsal for the work they are training to do and the challenges it will inevitably hold. For faculty, the challenge is to teach the principles and tools of analysis without relying on simplified cases or sanitized data that give students a false comfort of a straightforward application that leads to a “best” answer.

Imparting a broad appreciation for the role of data in public policy and management

While some students will want to develop strong technical expertise, all should be able to discern what types of analysis and sources are appropriate in various contexts. Every public affairs graduate should be a discriminating consumer of data, a critical audience for data analysis, and a trusted steward of data resources. Graduates who go into government positions will inevitably be in some way responsible for the quality and management of data in their own agencies. Workshop participants noted that policy informatics has its own policy components, including ethics and legal frameworks, or as one workshop participant put it, “We need policies for data, not just data for policies.” Participants began a discussion of the importance of identifying the threshold knowledge and skills that all students should acquire in their degree programs, emphasizing the need to connect and integrate these newer demands with existing core competencies and traditional classes.

Acquiring robust tools and technologies

Students who want a specialization in policy informatics face another challenge in the cost and accessibility of robust tools. The challenge is three fold: first, tools (or licenses) that can handle large, realistic data sets with a good selection of features can be costly. Second, the tools and techniques for using them are constantly changing, thus requiring the ability to upgrade and branch out to different packages or features. Third, very few public affairs faculty have the knowledge or skills to teach about or with these tools, or room in the teaching schedule to devote whole courses to the topic. Open source programs for classroom use such as R or Quantum GIS (QGIS) can provide good teaching tools, but even these may be beyond the reach of individual students or departmental resources. Team teaching, cross-listing with statistics, math, or business courses, and cross-disciplinary courses are all possible ways to deal with this challenge. It is also helpful to remember that many government agencies also lack the resources for high-end tools, so affordable options in the classroom may well be the right choice for long term usefulness in practice.

Advocating for changes in curriculum

Demand for students with data-intensive skills is on the rise in many fields.17 Public affairs graduates are competing with students from other disciplines for these positions. Given the multi-faceted perspective that policy informatics imparts, graduates with policy informatics training can act as change agents or boundary spanners in their agencies. They are likely to be better communicators about and savvy consumers of data and evidence. Their higher levels of data and technology literacy, grounded in the public affairs context, can allow them to play leadership and facilitation roles that demonstrate the special value of hiring managers and analysts with a public affairs degree.

16Dawes, Sharon S. "Stewardship and usefulness: Policy principles for information-based transparency." Government Information Quarterly 27, no. 4 (2010): 377-383.
17Manyika, James, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles Roxburgh, and Angela H. Byers. "Big data: The next frontier for innovation, competition, and productivity." (2011).