How can data and evidence be more effectively used to guide integration policy? This chapter examines how monitoring and evaluation are used in practice to guide integration policy. It focusses on how evidence feeds into governance processes and how it is used to inform decision making.
5. Evidence-based decision making
Copy link to 5. Evidence-based decision makingAbstract
A central element of integration governance is the ability to identify what works in practice and use that knowledge to improve policy over time. Monitoring systems and regular evaluation help governments assess outcomes, identify gaps and adapt implementation when results fall short.
This chapter examines how OECD countries collect, organise and use data to monitor integration outcomes. It looks at how different types of evidence, including administrative data and surveys capturing migrants’ experiences, are used in governance and decision-making processes, how they inform policy and implementation, and where important gaps remain.
Practices differ substantially between countries. Some countries have well-established monitoring and evaluation systems, whereas others rely on more fragmented arrangements or have limited analytical capacity. Monitoring systems are common in OECD countries, though their scope and level of integration vary considerably. Evaluation practices are more uneven, and evidence does not always feed systematically into policy decisions. This creates an important governance challenge: ensuring that monitoring and evaluation not only produce information but also support policy learning and adaptation over time.
5.1. Monitoring frameworks. Administrative and survey data
Copy link to 5.1. Monitoring frameworks. Administrative and survey dataAdministrative registers, such as employment, education or social security records, form the backbone of integration monitoring in most OECD countries. They provide consistent and detailed information on employment, education, language learning and participation in integration programmes, and in some cases on benefits or housing support. These systems enable systematic data collection with national coverage and allow outcomes to be tracked over time.
In most OECD countries, responsibility for monitoring integration outcomes is anchored in ministries responsible for migration or interior affairs, often in co‑operation with employment, education or social-affairs authorities. In countries where monitoring indicators are linked to budgeting or performance management, finance ministries may also play a role.
5.1.1. Administrative data
Most OECD countries report having established monitoring systems based on administrative data and formal reporting frameworks, although they vary in scope and how far data are integrated between systems and institutions. A majority of responding countries report having monitoring frameworks for integration outcomes (27 countries), while a smaller number report no such framework (seven countries).
Across OECD countries, monitoring systems tend to focus more strongly on participation and short-term administrative outcomes than on longer-term integration trajectories. Administrative data are generally strongest in areas such as language‑course participation, employment registration or programme completion, whereas outcomes related to sustained employment, social inclusion or civic participation are tracked less systematically.
In several countries, administrative data are embedded directly in performance frameworks or budget processes. In the Nordic countries, where municipalities are responsible for much of integration service delivery, monitoring systems are designed to aggregate local administrative data into national frameworks. In Denmark and Norway, administrative indicators on participation in language courses and use of integration measures are collected systematically at municipal level and reported through national frameworks. This allows for comparison of outcomes between municipalities.
In Germany, Austria and the Netherlands, monitoring obligations are formally anchored in legislation governing migration or integration. This creates stable, long‑term reporting obligations. In Germany, integration and orientation courses, language provision and related measures are regulated in federal law. This legislation defines who must report, what must be reported, and how frequently. Länder and municipalities then complement this with their own monitoring frameworks. In the Netherlands, the Civic Integration Act establishes the monitoring and reporting duties for municipalities and contracted providers involved in the civic‑integration system. Because these requirements are defined in law, they apply consistently across regions. This helps ensure that monitoring remains standardised and comparable between regions and provides continuity across political cycles.
Differences in monitoring systems often reflect broader institutional arrangements, including how integration responsibilities are distributed between levels of government, the role of contracted service providers, and whether monitoring focusses primarily on migration management or longer-term integration outcomes. Examples from other OECD regions illustrate these different approaches. In New Zealand, administrative data are complemented by the Integrated Data Infrastructure (IDI), a linked data environment covering the whole population. While the IDI is not migration‑specific, it allows outcomes to be analysed by migration characteristics such as visa category, time since arrival or residence status, which can support longitudinal analysis of integration outcomes across policy areas. In Japan and Korea, monitoring frameworks are more closely linked to migration status categories rather than broader integration outcomes, although targeted initiatives exist for specific groups such as marriage migrants.
5.1.2. Survey-based evidence
Despite their strengths, administrative datasets do not capture all dimensions of integration. Administrative monitoring systems are effective for tracking participation in programmes and socio‑economic outcomes, but they are less well suited to capturing how migrants experience integration in practice. These limitations create important governance blind spots in areas such as discrimination, institutional trust, social participation or perceived belonging, which are difficult to capture. Several of these aspects are not reflected in administrative indicators, including understanding institutional procedures, navigating services or feeling welcome (OECD, 2023[1]). To reduce such blind spots, many OECD countries complement administrative statistics with periodic surveys or thematic studies. These instruments make it possible to measure subjective outcomes, attitudes and experiences not reflected in administrative systems. France’s Longitudinal Survey on the Integration of First-Time Arrivals (ELIPA), for example, complements administrative monitoring systems by tracking integration trajectories over time (see Box 5.1).
Box 5.1. Longitudinal monitoring of integration trajectories in France: The ELIPA survey
Copy link to Box 5.1. Longitudinal monitoring of integration trajectories in France: The ELIPA surveyIn France, longitudinal survey tools complement administrative monitoring by providing information on longer-term integration trajectories and experiences that are not visible through administrative data alone.
Alongside administrative reporting linked to the Contrat d’intégration républicaine (CIR), France has developed the Enquête longitudinale sur l’intégration des primo‑arrivants (ELIPA), a longitudinal survey following first-time arrivals through different stages of the integration process.
ELIPA complements administrative monitoring by collecting information on dimensions of integration that are more difficult to capture through administrative registers alone. These include migrants’ experiences with public services, language acquisition, housing conditions, employment trajectories, social participation and perceptions of discrimination. By following individuals over several years, the survey also provides information on how integration outcomes evolve over time rather than only at a single point of contact with public institutions.
The survey forms part of a broader monitoring framework linking the Ministry of the Interior, OFII and other public actors involved in integration policy. It supports evaluation of integration measures and contributes evidence for policy development and programme adjustment. The second ELIPA survey (ELIPA 2), launched in 2019, expanded the use of longitudinal monitoring within France’s integration system.
Source: OECD questionnaire on integration governance (2025).
In most OECD countries, labour force, education and living conditions surveys provide harmonised data that allow comparison between countries and over time. Other examples include settlement outcome surveys in Canada, Australia and New Zealand, which capture aspects such as belonging, perceived discrimination, well-being and access to services. Switzerland combines administrative data with survey-based indicators capturing subjective dimensions such as sense of belonging, perceived discrimination and social participation. These indicators are integrated into the national monitoring system and used alongside socio‑economic measures, reflecting broader efforts in some OECD countries to integrate also measures of social inclusion. Some countries also commission targeted studies on specific groups, such as youth with migrant parents or beneficiaries of protection.
Survey-based evidence remains unevenly developed. This reflects, in part, their resource‑intensive nature, as regular surveys require sustained funding, technical capacity and repeated fieldwork that may not be prioritised. In countries with small migration flows, the use of dedicated surveys may be limited. In fact, only a minority of countries run regular, dedicated surveys on migrant integration. Where these data are missing, important aspects of integration, such as whether people feel they belong, understand how to navigate institutions, or experience discrimination, are often not visible in policy discussions. Where such data exist, they are important for broadening analysis and for informing policies that respond more effectively to barriers that are otherwise difficult to detect through administrative systems alone.
5.1.3. Integrated monitoring tools – dashboards and barometers
Several countries have developed integrated monitoring systems that combine indicators from administrative data, surveys and programme reporting into a policy-oriented format. These tools are typically managed by ministries or specialised integration agencies, often in co‑operation with statistical authorities and local governments. They mark a shift from periodic reporting towards continuous monitoring that can support operational follow-up and co‑ordination. Data are used not only to track outcomes but also to support day-to-day implementation and co‑ordination decisions across levels of government.
In Denmark, the national government manages the Integration Barometer, which tracks nationally defined objectives in areas such as employment, education, language proficiency, civic participation and equal treatment. It is used by municipalities to benchmark performance and adjust local strategies and is publicly available. In Norway, the Directorate of Integration and Diversity (IMDi) manages a barometer integrating administrative data with selected survey and programme indicators. It enables counties and municipalities to monitor progress under the Introduction Programme and related measures. In France, SI-AGIR dashboards used by central authorities and contracted service providers provide near real-time monitoring of support intensity, duration and outcomes for beneficiaries of international protection.
Dashboard-based monitoring is in some cases used to inform concrete adjustments. In Denmark, Integration Barometer results are used in dialogue between central government and municipalities and inform the calibration of performance‑linked financial incentives within the integration system.
Despite these developments, integrated monitoring tools remain relatively uncommon in OECD countries. In many countries, data are collected separately by policy area such as employment, education or housing. As a result, the same information may be collected multiple times, while important links between policy areas remain difficult to identify. This fragmentation can make it harder to co‑ordinate action between institutions, monitor longer-term integration trajectories or build a coherent picture of overall progress.
Countries using integrated dashboards and barometers report that combining data within a single system can support co‑ordination, improve understanding of implementation and make gaps between policy areas more visible. However, these tools also have limitations. They depend on the quality and comparability of underlying data, which can vary by sector and level of government. They also reflect a broader limitation of monitoring systems: they tend to prioritise what can be measured. Indicators tend to focus on outcomes that are easier to quantify, such as employment or programme participation, whereas broader dimensions of integration, including social inclusion, discrimination, access to services or lived experience, are sometimes captured in surveys but generally less consistently.
When monitoring systems rely heavily on standardised indicators, they may create incentives to prioritise outcomes that are easier to measure in the short term, such as rapid labour-market entry or programme completion. In integration policy, this can make it more difficult to capture longer-term processes such as language acquisition, social participation or gradual labour-market progression. Monitoring indicators also often focus on final outcomes rather than intermediate stages of integration. As a result, measures that improve skills, work readiness or understanding of institutions may appear less effective when only immediate outcomes are considered.
5.1.4. International benchmarking
The choice of benchmark also matters. Comparing outcomes to native‑born populations can highlight gaps but may not reflect differences in starting points or barriers faced by migrants. International comparisons can be a valuable tool for strengthening monitoring systems where indicators are defined consistently and interpreted carefully. Where data are harmonised between countries and linked to comparable policy objectives, cross‑country benchmarks can help identify gaps, contextualise performance and highlight areas for improvement. Regional statistical co‑operation provides one way to support this (see Box 5.2).
Box 5.2. Nordic statistics: A harmonised regional approach to integration monitoring
Copy link to Box 5.2. Nordic statistics: A harmonised regional approach to integration monitoringThe Nordic countries have developed a shared transnational monitoring system for integration. The Nordic Statistics Database, which is updated annually, provides harmonised indicators on population, labour-market participation, education, income and housing for both native‑born and foreign-born residents.
The integration and migration statistics in the database are the result of a co‑operation project involving the Nordic Statistical Institutes. This project has been ongoing since 2015 with the aim to harmonise immigration and integration statistics among the Nordic countries.
Key features include:
Consistent definitions enabling cross-country comparison.
Integration of multiple administrative registers, creating longitudinal datasets.
Alignment with national monitoring frameworks, ensuring coherence between national and regional reporting systems.
Accessible data, supporting researchers, municipalities and civil society in planning.
Regional statistical co‑operation can support more coherent and comparative monitoring of migrant integration. EU‑wide frameworks such as Eurostat provide harmonised integration statistics for EU Member States, while regional co‑operation can enable more timely and granular analysis by aligning national registers and reporting cycles. This type of harmonised statistical co‑operation also supports broader comparative analysis of integration outcomes. OECD comparative work, including the Settling In series, similarly relies on internationally comparable indicators to analyse migrant integration outcomes across countries and over time.
Source: OECD questionnaire on integration governance (2025); Nordic Statistics Database, https://nordicstatistics.org.
International frameworks increasingly support co‑ordination and comparability in integration governance. While monitoring systems are developed nationally, international co‑operation helps governments compare approaches, develop common indicators and exchange operational information on implementation and outcomes.
At EU level, the EU Action Plan on Integration and Inclusion (2021‑2027) encourages EU Member States to strengthen monitoring systems and improve data collection, including the use of disaggregated data. EU funding instruments such as the Asylum, Migration and Integration Fund (AMIF) and the European Social Fund Plus (ESF+) also require reporting on measurable outputs and outcomes, reinforcing the link between funding and monitoring. A mid-term review of the Action Plan published in 2025 highlighted continued governance challenges, including co‑ordination across levels of government and the need to strengthen partnerships with local actors and civil society organisations (European Commission, 2025[2]).
The European Integration Network (EIN) is the main EU-level forum for exchange between national authorities responsible for migrant integration. It provides a structured setting for Member States to discuss policy developments, implementation challenges and governance approaches in areas such as employment, education, health, housing and social inclusion. Through meetings, peer exchange and thematic discussions, the Network contributes to greater comparability in how integration policies are monitored and organised across Member States.
The European Migration Network (EMN), also established by the European Commission, supports comparative analysis and information exchange on migration, asylum and integration policies. Through studies, policy reports and ad hoc queries, the EMN helps governments to compare policy approaches and share practical experience across EU Member States.
At OECD level, comparative data collections, peer reviews and flagship publications such as Settling In, developed jointly with the European Commission (see Box 5.3) support benchmarking and cross-country learning. The OECD’s Continuous Reporting System on Migration, known by its French acronym SOPEMI, provides a long-standing mechanism for collecting and comparing information on migration trends and policy developments across OECD countries through a network of national correspondents. The annual OECD International Migration Outlook is based on this system.
These international frameworks support the exchange of information, development of comparable indicators and access to cross-country evidence on integration outcomes and governance systems. They can also encourage greater consistency in reporting practices and provide governments with comparative information on how integration governance is organised and monitored.
Box 5.3. Indicators of Immigrant Integration: Settling In
Copy link to Box 5.3. <em>Indicators of Immigrant Integration: Settling In</em>The joint OECD-European Commission publication Indicators of Immigrant Integration – Settling In provides a comprehensive comparison of the integration outcomes of immigrants and their children in OECD, EU and selected other countries. The report compiles harmonised indicators based on comparable administrative and survey data and includes 83 indicators covering three broad areas: labour market and skills; living conditions; and civic engagement and social integration. These indicators span areas such as employment, education, income, housing, health, civic participation and social inclusion, while also providing detailed information on the characteristics of immigrant populations and households.
5.2. Disaggregated data and differential outcomes
Copy link to 5.2. Disaggregated data and differential outcomesDifferences in integration outcomes between women and men are well documented in OECD countries, but the treatment in monitoring and evaluation frameworks remains limited and uneven. Differences in integration outcomes depend in part on their statistical visibility, that is, how well they are captured in data. This means that monitoring systems shape not only what governments know about integration outcomes, but also which groups and barriers become visible in policymaking. For example, if differences in access to employment measures or childcare are not systematically tracked, they may remain under-recognised in integration policies, even where they have a direct impact on labour-market participation.
Most countries collect sex-disaggregated data in key policy fields such as employment services, language programmes, health and social services. However, these differences are seldom analysed systematically or translated into policy adjustments. The main challenge often lies less in data availability than in the systematic use of evidence in policymaking.
Some targeted approaches exist. France integrates gender considerations into its national evaluation framework and finances measures targeting barriers faced by migrant women, such as childcare constraints and access to health services. These barriers can affect participation in language training and employment programmes and are not always visible in standard monitoring indicators. In the Nordic countries, sex-disaggregated indicators are routinely published in labour, education and family policy areas, and countries increasingly combine these with other variables such as age, educational background and migration background. In contrast, Korea and Japan primarily organise monitoring by visa category rather than by sex, although targeted initiatives exist for migrants arriving through marriage channels, who are predominantly women.
Many countries still face challenges in analysing how integration outcomes differ across groups with different characteristics and support needs. A smaller number of countries systematically combine sex with other variables such as age, educational background, disability, migration channel or protection status. Outcomes may differ substantially between groups, including young women with low literacy, older migrants with limited schooling and beneficiaries of protection with trauma histories. More detailed and connected monitoring systems can help governments identify cumulative disadvantages and design more targeted measures. This often requires linking data from different policy areas and ensuring that different administrative systems can be combined. In many countries, this remains a technical and institutional challenge. More flexible monitoring systems may also be needed to respond to changing migration patterns and emerging forms of vulnerabilities.
5.3. Evaluation
Copy link to 5.3. EvaluationIn contrast to the widespread use of monitoring, evaluation practices remain more fragmented and unevenly institutionalised across countries. Monitoring systems track participation, outputs and short-term outcomes through regular indicators and reporting. Evaluation goes further by assessing whether policies and programmes are achieving their intended results, why some measures work better than others, and how implementation affects outcomes over time. In integration policy, this is particularly important because outcomes often depend on co‑ordination across multiple policy areas and levels of government.
Although many countries conduct evaluations of specific integration measures or programmes, fewer report having a comprehensive or recurring evaluation framework covering integration policy as a whole. Many evaluations are commissioned on an ad hoc basis, particularly in connection with externally funded programmes, especially those supported by EU funds, rather than being embedded in a systematic national evaluation cycle. As a result, countries differ in how evaluation is organised, ranging from regular, legally anchored evaluation cycles to more project-based approaches. Evaluation also tends to be less developed in policy areas characterised by shared responsibilities and long-term outcomes, such as integration, where results depend on multiple actors and policy areas.
Some countries have institutionalised evaluation requirements in legislation, strategies or programme contracts. Finland and Norway, for example, conduct regular evaluations linked to statutory programme cycles. In other countries, including Germany, Austria, France and the Netherlands, evaluation is embedded in accreditation and quality-control systems for service providers, combined with performance indicators and reporting obligations. Canada and Australia complement monitoring systems with beneficiary surveys, third-party evaluations and outcome‑based assessments of settlement programmes. These examples highlight the variety of mechanisms countries use to integrate evaluation into governance systems, including legal requirements, administrative procedures and programme‑based assessments.
In contrast, several countries, particularly those with more decentralised or project-based approaches, rely more heavily on ad hoc evaluations. This makes it more difficult to assess long-term results, compare performance between regions, or ensure continuity in learning over time. In many cases, evaluations are conducted by external actors such as research institutes or consultants, which can strengthen analytical quality if findings are integrated into policy processes.
5.4. How evidence informs policy
Copy link to 5.4. How evidence informs policyThe central governance question is how monitoring and evaluation findings are translated into policy decisions. The availability of data does not automatically lead to evidence‑informed policymaking, particularly in complex policy areas involving multiple actors and levels of government.
Only a limited number of countries report having formalised processes through which monitoring data or evaluation findings are systematically used to inform adjustments to integration policy or funding. Where such feedback mechanisms exist, they are most often embedded in broader budgetary or performance‑management processes, such as annual programme reviews, budget hearings or action-plan updates. In many countries, by contrast, monitoring and evaluation primarily serve reporting and accountability purposes, with limited evidence of structured mechanisms to translate findings into policy or funding changes.
Many countries still struggle to translate monitoring results into policy adjustments. This reflects both institutional and political constraints, including limited analytical capacity and competing policy priorities.
Some countries have nonetheless developed more structured feedback mechanisms linking monitoring and evaluation to operational or budgetary decisions. Some governments link budget allocations to performance indicators or require periodic strategy reviews based on monitoring results. In Denmark, Integration Barometer results feed directly into municipal funding decisions. Municipalities receive performance subsidies tied to successful employment and education outcomes tracked by the barometer, creating a legally established link between monitoring data and resource allocation. In countries using digital monitoring tools, such as AGIR in France or the Introduction Programme in Norway, real-time data can support operational adjustments, for example by adapting guidance to service providers or reallocating support measures. In addition, some countries with long-standing experience of large‑scale displacement, such as Türkiye, report using monitoring and evaluation findings to adjust programmes and budgets on a regular basis, illustrating how evidence can support operational management in large‑scale integration systems. In Estonia, findings from the 2023 Integration Monitoring Survey identifying gaps in language proficiency among adults led to additional national and EU resources being allocated to expand free Estonian language courses and digital learning opportunities.
The use of evidence nonetheless remains uneven. Countries report limited uptake of findings on discrimination, social exclusion or gender disparities. Although these dimensions are widely recognised as important, they are less consistently reflected in policy adjustments, partly because they are more difficult to measure and less directly linked to existing performance frameworks. This is consistent with findings from the OECD and European Migration Network, which indicate that although Member States have developed a wide range of indicators and monitoring tools, these often remain fragmented by policy area and do not capture the full range of integration outcomes (EMN; OECD, 2024[7]).
OECD country reviews also suggest that the extent to which evaluation findings are systematically linked to policy development varies across countries. In systems where monitoring and evaluation are embedded in legal or strategic review cycles, such as Norway and Finland, feedback mechanisms tend to be more structured. In other contexts, monitoring may be well developed in specific policy areas, while evaluation remains more project-based or decentralised, as in Sweden and Iceland.
Several factors help explain these differences. Monitoring and evaluation can be resource‑intensive and are not always prioritised alongside service delivery. In some systems, limited analytical capacity or fragmented data infrastructures constrain the ability to generate and use evidence effectively. Institutional arrangements may also play a role, particularly where responsibilities for data collection, analysis and policy design are distributed between multiple actors without clear co‑ordination mechanisms. In addition, some outcomes of integration, such as social inclusion or experiences of discrimination, are inherently more difficult to measure, which can limit the scope of monitoring systems. As a result, governments may have limited information on which integration measures are most effective, particularly for groups facing complex barriers.
5.5. Key takeaways
Copy link to 5.5. Key takeawaysAdministrative data form the basis of integration monitoring in almost all OECD countries, though the extent to which this is complemented by surveys, dashboards or longitudinal studies varies. Evaluation practices are more uneven, ranging from regular, legally anchored cycles to ad hoc, project-based assessments. Despite this variation, a few considerations stand out:
Countries that combine administrative data with regular surveys and qualitative evidence are better placed to track outcomes such as belonging, discrimination and lived experience.
While monitoring systems are widely established, evaluation remains less consistently embedded. In many countries, evaluations are carried out at programme level and are not systematically linked to policy reform. Countries that require evaluation findings to feed into strategy reviews or budget cycles may more easily identify effective measures and reallocate resources accordingly.
Disaggregated data shape which groups and barriers become visible in policymaking and which remain invisible. Where differences between women and men in access to employment measures, childcare or health services are not tracked, policies risk being designed around an incomplete picture of about half the migrant population.
Integrated monitoring tools such as dashboards and barometers can bring together data across policy areas. While still relatively uncommon, countries that have developed such tools report stronger co‑ordination between institutions and clearer accountability for outcomes. However, these tools are only as good as the data that feed into them. Expanding their use requires investment in common indicators, data-sharing arrangements and attention to data quality.
For many OECD countries, the main challenge is no longer necessarily data availability but its use. Monitoring and evaluation often generate substantial information that does not consistently reach policy or budget decisions. Closing this gap between what is known and what is acted on remains one of the central governance challenges for integration policy.
References
[7] EMN; OECD (2024), Monitoring the integration of third-country nationals European Migration Network-OECD Joint Inform.
[2] European Commission (2025), Mid-term review of the Commission Action Plan on Integration and Inclusion 2021-2027.
[1] OECD (2023), Introduction Measures for Newly-Arrived Migrants, Making Integration Work, OECD Publishing, Paris, https://doi.org/10.1787/5aeddbfe-en.
[3] OECD (2012), Settling In: OECD Indicators of Immigrant Integration 2012, OECD Publishing, Paris, https://doi.org/10.1787/9789264171534-en.
[6] OECD/European Commission (2023), Indicators of Immigrant Integration 2023: Settling In, OECD Publishing, Paris, https://doi.org/10.1787/1d5020a6-en.
[5] OECD/European Union (2018), Settling In 2018: Indicators of Immigrant Integration, OECD Publishing, Paris/European Union, Brussels, https://doi.org/10.1787/9789264307216-en.
[4] OECD/European Union (2015), Indicators of Immigrant Integration 2015: Settling In, OECD Publishing, Paris/European Union, Brussels, https://doi.org/10.1787/9789264234024-en.