1.1 Data-driven Policy and Management

Using mathematical models and data to solve policy and management problems is not new. Policy modeling, or the use of quantitative modeling to support the public policy making process was developed in the 70s. Data science is a renewed face of policy modeling that enriches the traditional perspective that used mostly statistics as a policy ana;ysis tool to incorporate a broader breadth of modeling and analytical techniques that come primarily from Computer Science and includes Artificial Intelligence and other machine learning algorithms. Additions to the toolset have been necessary to manage the large amounts of data that we produce every day in our interactions with people, organizations and electronic devices such as cell phones, tablets and personal computers. Further, advances in modeling and visualization techniques have made it possible to gain new insights into the importance of including groups of policy makers and other stakeholders in using models, data and other technical tools to analyze problems and policy alternatives. In this way, beyond recognition of the importance of modeling and data, contemporary data science promotes stakeholder involvement as well as interdisciplinary approaches to policy analysis.

According to the OECD: “Open Government Data (OGD) is a philosophy- and increasingly a set of policies – that promotes transparency, accountability and value creation by making government data available to all.”

 

This renewed interest in Data Science has been promoted, at least partially, by the increased amount of data available. New imperatives for more open governments, including increased transparency, accountability and engagement, are driving investments in initiatives designed to make government data increasingly “open”. To fully leverage these investments in open data, governments and other stakeholders are turning to data science as the means to gain expected public value from the newly available data. Our colleagues Marijn Janssen and Natalie Helbig (2018) describe this new trends in a very interesting paper published in Government Information Quarterly. A major take-away from the paper is that using data and models in the process of policy development provides a sandbox or laboratory where the data scientist and othe stakeholers learn about the policy problem and alternative solutions. There are many different approaches to Data Science, but the Data Science tradition at Rockefeller College relies on four main pillars: analytical methods, data, technology tools that facilitate the use of methods and the engagement of stakeholders (see Puron-Cid et al., 2016 and Zhang et al., 2016). The sections below on analytical methods and techniques, data and open data, information technologies and stakeholder involvement outline some of the latest thinking about the building blocks, or pillars, of Policy Informatics and the chapters themselves provide the reader with new insights into their use.

Analytical methods and techniques

A model, in general, is a conceptual representation of a problem, and it helps policy makers and other stakeholders structure the inquiry process. In many cases policy analysis requires the use of different modeling tools and techniques depending on the problem at hand, and in fact, using different modeling techniques with the same problem may lead to different policy options, each of them with advantages and limitations. In this way, using different modeling approaches to the same problem offers different perspectives about the problem and helps the analyst to learn a specific aspect of the problem at hand. Researchers and practitioners in the fields of economics, mathematics, operational research and systems analysis, among other disciplines, have, over the years, developed many types of qualitative and quantitative models, and we will talk in more detail about some of these perspectives in the introductory chapter to the models section of the book. 

Data and Open Data

While making government data open and available creates a host of new opportunities, using these data for data science and policy analysis presents many challenges. First, data needs to be integrated from multiple, disparate sources and in different formats. Beyond the technical challenge of integrating these data, data scientists must understand both the context where data was created and the context where data is going to be applied, which are frequently different. In other words, using data that has been captured, managed and used as part of a government business process in data science often requires that data to be reorganized and often cleansed and consolidated in ways that make it relevant or usable to that modeling effort. Lack of recognition of these challenges may lead to biased or simply wrong conclusions.

In the United States, Federal Government Agencies are required to produce open datasets by the Foundations for Evidence-Based Policymaking Act of 2018. Most of these datasets are available through the repository data.gov.

Information Technologies

The use of information technologies in the context of public policy analysis is a critical component of Data Science. Data repositories are a clear example of how technologies can be used to share and made data widely available. Developments in database technologies allow for integration of disparate datasets in data warehouses and data lakes (more of this on chapter 7), and faster computers allows for increased processing capabilities. In addition, software applications that made available visualization tools as well as mathematical and statistical applications to be used in a more intuitive way open the possibility for the application of these techniques by multidisciplinary teams. Today policy makers, analysts, interest groups, and even citizens have access to mathematical and statistical software packages and tools to process data. Some are highly sophisticated instruments and others are user-friendly applications; these quantitative software tools provide quasi-intuitive methods to analyze and describe data and to infer associations among the critical indicators of the related public policy problem.

Stakeholder involvement

Finally, stakeholder involvement in the construction and analysis of models to explore policy is the final pillar of Data Science. The particular type of stakeholder engagement found to facilitate policy modeling finds its roots in decision conferences and groups decision support systems. Experiences in running this type of facilitated meetings show that there are two main components in the process: dividing the facilitation task in different roles, and organizing the conversations around small facilitated tasks or scripts (Richardson et al. 2015). In other words, engaging stakeholders in developing models for policy analysis is better done with a team that includes expert facilitators, modeling and domain experts. Moreover, facilitation is more effective when using specific activities to elicit knowledge from the domain experts as well as exploring models with computational tools.

 

Further Readings

  • Janssen, M., & Helbig, N. (2018). Innovating and changing the policy-cycle: Policy-makers be prepared! Government Information Quarterly, 35(4), S99–S105.
  • Puron-Cid, G., J. R. Gil-Garcia and L. F. Luna-Reyes (2016) “Opportunities and Challenges of Policy Informatics: Tackling Complex Problems Combining Open Data, Technology and Analytical Tools.” International Journal of Public Administration in the Digital Age (IJPADA). 3(2):66-85
  • Richardson, G. P., D. F. Andersen and L. F. Luna-Reyes (2015) “Joining Minds:  Group System Dynamics Modeling to Create Public Value,” in Bryson, J. M., Crosby, B. C. And Bloomberg, L. (Eds.), Public Value and Public Administration, Georgetown University Press, Washington, DC, pp. 53-67
  • Zhang, J., L. F. Luna-Reyes and T. A. Pardo (2016) “Information, Policy, and Sustainability: The Role of Information Technology in the Age of Big Data and Open Government” in J. Zhang, L. F. Luna-Reyes, T. A. Pardo and D. S. Sayogo (Eds.), Information, Models, and Sustainability: Policy Informatics in the Age of Big Data and Open Government, Public Administration and Information Technology (PAIT) Series, Springer, pp. 1-19

Attribution

By Luis F. Luna-Reyes, and licensed under  CC BY-NC-SA 4.0.

License

Icon for the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

Data Analytics for Public Policy and Management Copyright © 2022 by Luis F. Luna-Reyes, Erika G. Martin and Mikhail Ivonchyk is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, except where otherwise noted.

Share This Book