Introduction. With the rise of information technologies, a genuine culture of evaluation has become within reach. In this sense, data has become the most valuable resource for solving the most urgent problems in our societies. Only through its correct use is it possible to generate useful knowledge and to improve decision-making, as well as to increase efficiency and transparency at public institutions. Far from requiring great efforts and investments in data collection, these same agents have already the largest pool of resources for evaluation: the administrative data.
Why administrative data. Public institutions have administrative databases of great breadth and depth, sustained over time and complementary between different departments, as repositories of extensive records on participants in public programs, taxpayers, recipients of social assistance, affiliations to Social Security, contracts signed, or unemployed people registers. Taking advantage of their potential use would represent an important qualitative leap in the evaluation of public policies, since it would allow to overcome the classic limitations that evaluation faces in terms of the obtention of sufficient data.
Difficulties and challenges. However, there are still many obstacles that prevent the extensive use of administrative data in research. They are, in many countries, difficult to access for the research public, being usually of exclusive internal use for the institution responsible for their management. Additionally, the lack of communication between the different administrations sometimes entails a low compatibility between data records that would be of interest to analyze jointly. Lastly, problems derived from anonymity, failures in the collection and treatment of information or inconsistencies over time prevent a full use of its potential to boost the evaluation of public policies.
With all the previous challenges and opportunities on the table, some recommendations are provided for the expansion and improvement of the use of administrative data in the evaluation of public policies:
Unleash the potential of already existing databases. Government agencies have access to a large amount of data that, after the initial expenditure in acquiring and processing it, is rarely used for research or decision-making. Making existing administrative databases available to the research public (always respecting anonymity) is only the first step in an ambitious evaluation culture. Researcher access to administrative databases should be based on transparent rules instead of ad-hoc decisions. With the appropriate guarantees, the flow of information that the institutions receive could become a continuous source of learning and knowledge with which to improve the way in which citizens and administrations interact. Indeed, researcher access to administrative databases could ensure reproducibility, which could improve the validity and quality of available evidence.
Compromise with anonymity. The right of individuals to remain anonymous must be adequately guaranteed, establishing a cooperation commitment between the administrations and the research community. A carefully designed legal framework can ensure that a fluid information exchange relationship is not incompatible with the responsibility towards the citizenry, the main beneficiary of this bond.
Improve the complementarity between public databases. The data is already there, but it is sometimes hard to make use of it in a joint and global manner. When it comes to the management of public information, each institution goes its own way, like in the Tower of Babel, hindering, for example, the merging of tax collection records with data on social assistance beneficiaries. Quite the opposite, the different administrations must speak the same language in terms of coding, anonymization and data processing, hence allowing the compatibility between diverse databases. Collaborating, rather than working alone, is therefore essential.
Improve complementarity with external databases. Surveys, questionnaires or data records from external institutions or private companies do also provide a complementary and necessary perspective, beyond administrative databases. Whenever the capacity of the public sector proves to be insufficient, the collaboration with external actors opens new ways for research to explore. This way, to foster compatibility between the public data and the aforementioned sources is to swim in favor of an ambitious and advanced evaluation culture.
Collect data from long periods of time. The more data available, the better. Ensuring the consistency and sustainability over time of the administrative data collection is crucial. Likewise, newly collected information shall not replace that previously obtained. It is in the long term where the prevailing challenges are best appreciated, and it is in this period that the ambition of evaluation must lie.
Incorporate Big Data and AI techniques. Applying the new technologies and the use of artificial intelligence for the management of large databases could represent a significant qualitative leap: they could help minimize error, automate data collection, achieve a remarkable level of detail and improve the efficiency of administrations and public institutions, reducing bureaucracy and freeing job counselors from that responsibility.
Digitization and computerization of data processing. Beyond the new techniques, the correct digitization of the processes of collection, treatment and storage of public databases is of the greatest urgency. There are enough resources and the time is right to take a step forward and put the public sector at the forefront of efficient and responsible computerization.
Further reading:
Brugarolas, P., Gorjón, L. y de la Rica, S. (2020). El uso de datos administrativos para la investigación. El caso de Dinamarca: un ejemplo de buenas prácticas. Informe ISEAK 2020/3. (Only available in Spanish).
Available at: https://iseak.eu/documentos/el-uso-de-datos-administrativos-para-la-investigacion-el-caso-de- dinamarca-un-ejemplo-de-buenas-practicas/
Csillag, M., Krekó, J. y Scharle, A. (2021). Counterfactual evaluation of youth employment policies. Methodological guide. Youth Employment PartnerSHIP. 1/2021.
Available at: http://yepartnership.ibs.org.pl/p/counterfactual-evaluation-of-youth-employment-policies
Crato, N., Paruolo, P., (2019). Data-Driven Policy Impact Evaluation. How Access to Microdata is Transforming Policy Design, Springer.
Available at: https://link.springer.com/book/10.1007/978-3-319-78461-8
Gorjón, L., Lizarraga, I. y de la Rica, S. (2021). La evaluación como proceso de aprendizaje. Una necesidad del presente,una inversión de futuro. Revista ICE. (Only available in Spanish). Forthcoming.
