Use of Secondary Data in Scientific Research

The use of secondary data has gained central importance in applied public policy research with the expansion of open data portals, the standardization of data governance principles, and the decreasing cost of analytical tools. For managers and researchers, working with existing datasets enables faster responses to complex problems, scalability of analyses, and reproducibility — provided that methodological and legal best practices are followed. International initiatives such as the FAIR Principles (Findability, Accessibility, Interoperability, and Reusability) guide the governance and reuse of digital assets, enhancing the discoverability, accessibility, and reusability of data for scientific and policy purposes (WILKINSON et al., 2016).

In parallel, the Open Data Charter has consolidated the principles of open data (open by default; timeliness and completeness; accessibility and usability; comparability and interoperability; data for better governance and engagement; and for inclusive development), reinforcing the use and reuse of government data as key inputs for the policy cycle — formulation, implementation, monitoring, and evaluation. At the regional level, Latin America and the Caribbean (LAC) have expanded national catalogs (e.g., dados.gov.br, datos.gob.mx, datos.gov.co, datos.gob.cl, datos.gob.ar, catalogodatos.gub.uy) and strengthened national statistical systems (IBGE, INEGI, DANE, INE/Chile, INDEC, INEI/Peru), in addition to regional and global repositories such as CEPALstat and World Bank Open Data.

This article, written in an analytical style, presents four sections: (i) what secondary data are; (ii) the pros and cons of their use; and (iii) examples of public data in LAC, with references and citations following the ABNT standard and verified online sources (excluding Wikipedia).

What Are Secondary Data

Secondary data are information originally collected for a specific purpose (statistical, administrative, regulatory, or operational) and later reused to answer new research questions. The methodological literature defines them as data produced by other researchers or institutions and analyzed at a later time, either in aggregated form (indicators) or as microdata (anonymized unit records) (JOHNSTON, 2014; WICKHAM, 2019).

In public policy research, the four most common sources of secondary data are:
(a) Official statistics based on censuses and sample surveys (e.g., Censo Demográfico and PNAD Contínua in Brazil; ENOE in Mexico; GEIH in Colombia; CASEN in Chile);
(b) Administrative records generated by government operations (health, education, social assistance, justice, public safety, budget, and public finance);
(c) Management and monitoring data (accountability systems, open government portals, and program evaluation systems); and
(d) International datasets comparable across countries (World Bank, ECLAC, IMF). These data are typically published on national open data portals and on the websites of statistical agencies, as well as in regional and global repositories.

The distinction between primary and secondary data is conceptual and operational: primary data are collected directly by the researcher for a specific research question, whereas secondary data were collected by third parties under predefined designs and definitions. In both cases, compliance with quality and ethical principles is essential. The UN Fundamental Principles of Official Statistics and the FAIR Data Agenda provide robust frameworks for governance, documentation, and responsible reuse.

Advantages and Limitations of Using Secondary Data

Advantages

• Speed and cost-efficiency: secondary data reduce research time and cost by avoiding the need to collect information from scratch — a critical advantage in contexts of budget constraints and urgent policy needs (JOHNSTON, 2014; QUALTRICS, 2021).
• Scale and comparability: they allow for long-term time series analysis, cross-national comparisons, and quasi-experimental evaluations when policy shocks occur. Official portals in LAC and international repositories (World Bank Open Data; CEPALstat) expand access to harmonized series.
• Reproducibility and transparency: open datasets with clear metadata and explicit licenses facilitate replication and public scrutiny. The FAIR Principles and the Open Data Charter define criteria to make data findable, accessible, interoperable, reusable, open by default, and comparable (WILKINSON et al., 2016; ODC, 2015).
• Thematic breadth: administrative and statistical records cover multiple policy areas (health, education, social protection, public finance, safety), enabling integration and linkage among datasets (respecting data protection frameworks) for systemic analysis. Frameworks such as the UN NQAF provide guidance for quality management across data sources.

Limitations and Risks

• Alignment between research question and original design: as datasets were designed for other purposes, conceptual mismatches (variables and classifications), operational differences, or methodological shifts may affect internal validity and comparability. The Total Survey Error (TSE) framework highlights coverage, sampling, nonresponse, measurement, and processing errors that—even in official data—must be understood and mitigated (GROVES; BIEMER).
• Variable quality in administrative data: administrative records may suffer from undercoverage, inconsistencies, operational biases, or linkage errors. Recent UN/UNECE guidelines recommend quality assessment, documentation, auditing, and governance for administrative data (UNECE, 2021; UNSD/UN-NQAF, 2019–2025).
• Access, privacy, and ethics: reuse must comply with legal frameworks such as the Access to Information Law (Lei 12.527/2011) and the Open Data Policy (Decreto 8.777/2016) in Brazil, and with data protection laws (e.g., LGPD in Brazil; Law 1581/2012 in Colombia). Microdata are often anonymized or made available under safeguards. Administrative data tend to reflect the logic of service provision, possibly underrepresenting certain groups. The literature highlights selection, coverage, and linkage biases, recommending diagnostic and mitigation strategies (HARRON et al., 2017; KATZ et al., 2022).
• Institutional capacity and maintenance: official portals and data series require sustained budgets, governance, and updating routines. International experiences show that without continuous investment, the quality of key statistics deteriorates — a relevant warning for data governance in the region.

Mitigation Strategies

To maximize benefits and minimize limitations, it is recommended to:
(i) adopt a clear methodological plan, with a theory of change and well-defined questions;
(ii) verify metadata and dataset versions;
(iii) apply quality frameworks (UN NQAF, TSE, OECD Data Quality Framework) and administrative data checklists;
(iv) conduct sensitivity analyses for definitional or classification changes; and
(v) ensure legal compliance (LGPD, habeas data, open data decrees) with ethical risk assessment and anonymization where applicable.

Examples of Public Data in Latin America

Below is an overview of open and official data sources in the region, with examples of datasets, bulletins, and microdata relevant for policy design, implementation, monitoring, and evaluation. Emphasis is given to public access and documentation, which are essential for rigorous reuse.

Brazil

• Brazilian Open Data Portal (dados.gov.br): central repository for federal and subnational datasets with documented APIs. Useful for data mapping, thematic search, and ETL integration.
• IBGE – PNAD Contínua: quarterly and annual series on labor, education, and income; essential for evaluating employment and inclusion policies.
• DATASUS / TABNET: aggregated and microdata on health (morbidity, mortality, primary care, etc.), with dictionaries and tabulation tools.
• Transparency and Public Finance Portals (Portal da Transparência; SICONFI; Tesouro Transparente): data on expenditure, revenue, contracts, and intergovernmental transfers, with public APIs.
• Legal framework: Access to Information Law (Law 12.527/2011), Open Data Policy (Decree 8.777/2016), and General Data Protection Law (LGPD, Law 13.709/2018).

Mexico, Colombia, Chile, Argentina, Peru, and Uruguay sections follow similarly (INEGI, DANE, INE, INDEC, INEI, AGESIC, etc.), describing their open data systems, legal frameworks, and applied research examples such as quasi-experimental evaluations, linkage of household surveys with fiscal data, and assessments of transparency or inequality.

Regional and Multilateral Repositories

• CEPALstat: regional indicators for SDG monitoring and comparative historical series.
• World Bank Open Data and BOOST: macroeconomic, sectoral, and fiscal data for policy evaluation and budget transparency.
• IMF Data: high-frequency macroeconomic and financial series for economic monitoring.

Best Practices for Research Using Secondary Data

• Define research question and analytical design: specify hypotheses and indicators before dataset selection; consult metadata and technical notes to understand methodological changes.
• Traceability and documentation: record dataset versions, filters, and transformations; prioritize stable, well-documented sources with APIs.
• Quality and comparability: apply multidimensional quality criteria (relevance, accuracy, timeliness, coherence, comparability, completeness).
• Ethics and data protection: ensure compliance with LGPD (Brazil) and analogous data protection laws; prefer anonymized microdata and evaluate reidentification risk.
• Transparency and reproducibility: whenever possible, publish code and data dictionaries; cite dataset versions and official URLs; describe limitations and biases.

Final Considerations

Secondary data are not a “shortcut” but a structured field guided by principles and quality standards. When combined with sound documentation, quality assessment, and legal compliance, they enhance the capacity of researchers and policymakers in Latin America to design, implement, and evaluate public interventions with agility, scope, and transparency. The sources listed — national portals, statistical agencies, and multilateral repositories — provide a solid foundation. The challenge ahead lies in strengthening analytical capacities and data governance within public and academic institutions to ensure that expanded access translates into quality, rights protection, and public impact.