Global

The data that UNHCR collect on specific needs in its population database (proGres Version 4) are generally based on the observations of registration or protection staff who are using the UNHCR Registration Handbook, 2003, specific Needs Codes &amp; definitions to identify and classify these elements. These data are recorded under specific data fields, but there are no central monitoring tools to verify the reliability of the data collected. Most of these data are recorded under “free text” in case management notes which are not available to other V4 modules (refugee status determination, resettlement etc.) nor otherwise connected within PRIMES. UNHCR and its partner agencies collect and record the same information on specific needs from the registered persons without any systematic way to organize, analyse, or monitor the reliability of these data. Analysis of these data is done on a case-by-case basis. Furthermore, UNHCR global data systems do not have minimum Age, Gender, Diversity (AGD) typologies to allow for country, regional, global, and temporal aggregation, and data comparability. The over 30 million asylum seekers and refugees across the globe that UNHCR reported under its care possess diverse experiences and needs, that no single tool, dashboard, technology, or institution can fully capture. A centralised, integrated system could amalgamate insights from various tools and databases through a set of shared indicators, thereby enhancing UNHCR's ability to effectively monitor these diverse needs. Its absence has profound implications on the overall efficacy of monitoring and evaluating UNHCR's operations on multiple levels. Take the example of average processing time for first instance refugee status determination decisions, which needs to meet a three-month standard after the interview based on the procedural standards. Various data points in UNHCR's various databases would need to contribute to its aggregation to calculate the resultant aggregate figures that this indicator will generate. We need to compute the ratio of the total processing days for all cases that received a first instance decision by a country of asylum during the reporting period, to the total count of cases for which a first-instance final decision was taken upon decision review: •	Numerator: Count of processing days for all cases that received a first instance decision by a country of asylum during the reporting period•	Denominator: Total count of cases for which a first-instance final decision was taken upon decision review •	Unit: Days •	Type: Integer Multiple proGres log entries, like the date of decision issuance or record updates, need to be used to aggregate this average time metric to assess UNHCR's compliance with its own three-month standard for issuing decisions after interviews. This example is merely one indicator in a specific area of UNHCR's operational work listed in multiple procedural guidelines. These indicators are distinct, but ideally, they should be identified to enable relational connection and interpretation between the different data sets relating to different areas of UNHCR's operational work. At the operational level, lack of consistent data collection methods across different countries and lack of interoperability between systems affects the quality of assistance provided to refugees and displaced persons. This disconnection prevents detailed analysis using attributes such as nationality, sex, marital status, school grade, age, and specific needs like 'elderly' or 'child at risk.' The impact of this data deficiency extends beyond inaccurate record-keeping; it undermines service provision itself. For example, without reliable data, it becomes challenging to monitor whether people recorded with medical needs are actually prioritised for RSD interviews and subsequently referred to the appropriate healthcare partners. The absence of a clear, real-time understanding of the conditions and needs of the populations it serves can lead to aggravating their vulnerability. At the strategic planning level, integrating these systems would not only streamline data interpretation but also enable complex analyses, like designing targeted aid programs based on year of arrival or evaluating local integration potential across different refugee groups in various host countries.

On Specific Needs

UNHCR does not have an integrated system to know what data to use, how to source it, how to assemble it, or how to assess the validity of the data it collects from and about the populations affected by forced displacement. This “data problem,” as it's internally referred to within the organisation, has emerged over the years due to the absence of coordination and planning within and across UNHCR’s data and information hierarchy, which divides into three layers with data roles and responsibilities: Headquarters (HQs), regions, and countries. In 2019, I began my final role with UNHCR at the Global Data Service (GDS), the cross-organisational service responsible for overseeing all of UNHCR's data-related affairs, including digital identity, registration, and data management, across its expansive global network. As my terms of reference specified, I was tasked with conducting a business analysis to consolidate data needs, pinpoint data origin and destination across systems, and aggregate data from country operations to the HQ, ensuring accurate counts of registered asylum seekers and refugees with UNHCR. By that time, UNHCR had been grappling with this data problem for quite some time, with costly projects that did not address or solve the root causes. I contributed to the last two such projects before leaving the organisation in mid-2021. The latest one, Project Orion, followed its numerous predecessors with long-term undertakings, often far from completion or, in some cases, never completed at all. The name Orion, representing a distinct constellation of stars, aptly described the status of UNHCR's data. Beneath the surface, the data were housed in myriad databases and systems that were not interoperable; instead, they were meant to be just like the stars aligned in Orion. The challenge was integrating these data, a task highlighted in project reports using technical jargon like 'back-end', 'front-end desk office', and 'interoperability'. Although UNHCR publicly championed its 2019 Data Transformation Strategy, it internally grappled with a panic to tackle the data problem due to the absence of a centralized system for precise historical and present data analytics to guide its protection activities. Reliable data, real-time statistics, and balance coherence with flexibility would have meant that UNHCR had a data collection system with rigid workflows and forms to be used globally in every operation, to coherently collect data. In the absence of such an integrated system, data quality checks are confusing to implement because of the long list of both legacy and up-to-date data entry codes included in the design of multiple UNHCR databases and systems used in parallel. Initiatives and projects around UNHCR data collection focus on back-end integration of all these data sources, without a quality assurance process and automated quality check on the content of registered applicants’ UNHCR case files. This has implications for the accuracy of data UNHCR provides to the public, as well as the donors for planning, and the impact of services, which are generated from a separate database called COMPASS. Sources of potential error and bias across data collection, data processing, and quality assurance processes, across all 35+ data sources that UNHCR uses, generate errors in data across various levels: Data collection:•	Variation in detailed data categories collected locally.•	Different data collection modes.•	Inconsistent use of residual categories: “Null”•	Staff understanding of the value of data varied.•	Different staff in different operations may collect data differently.•	Same data collected multiple times e.g., by different partner organisations or UNHCR systems; little is known about the reliability of different entries. Data processing:•	Mapping detailed data categories to aggregated data as harmonised categories is not necessarily consistent between different regions and operations, leading to potential inconsistencies.•	Various systems may use different algorithms or criteria for data cleaning, risking the loss or misinterpretation of vital information.•	Absence of standardised processing procedures can lead to discrepancies when comparing data from different sources or time periods.•	Manual data entry and transfer between systems introduce risks of human error, duplication, or omission.•	Legacy systems might not be equipped to handle newer data types or formats, leading to potential compatibility issues or loss of information. Quality assurance:•	Lack of standardised guidelines on quality control led to discrepancies in data validation processes across different operations.•	Feedback loops for correcting errors or inconsistencies at country level might not be in place or may be inefficient, delaying the resolution of issues.•	The absence of automated quality checks means that errors can persist in the database until manually detected and corrected.•	Staff training on data quality assurance might be inconsistent and continue to lead to varied levels of scrutiny and potential oversights.

UNHCR's Data Problem

Jordan

Egypt

Syrian Arab Republic

I was working as the regional registration and identity management officer for the Middle East and North Africa Region (MENA) when I was tasked to work out the profile of persons from Syria registered with the organization in 2018 together with a project team. The necessary conditions had not been in place in Syria for a safe and dignified return of refugees. The overall approach of UNHCR regarding forcibly displaced persons from Syria was therefore guided by a comprehensive protection and solutions strategy. Some landmark global developments had affected UNHCR’s operations in general and the financial situation in 2018. UNHCR had convened a high-level meeting in Geneva on 30 March 2016 on 'legal pathways' for Syrian refugees and was playing a key role in planning a High-Level Plenary of the General Assembly on large movements of refugees and migrants in September. Both of these events expanded global mechanisms for responsibility sharing. A few months after that, UNHCR engaged in global discussions at the World Humanitarian Summit in 2016 rallying international support for the organization’s work and developing new approaches for taking it forward. Syria emergency being a global strategic priority, there was an immense pressure on UNHCR and the staff working in data like myself to take advantage of new opportunities to strengthen collective outcomes in humanitarian operations through innovation and new financing models and defend the foundations of international protection. UNHCR had to show, particularly to development actors, international financial institutions, and the private sector, that it was prepared for rapid and effective emergency response from Syria or for potential returns to the country. We needed to identify obstacles, conditions, and actions necessary for asylum seekers and refugees’ return to Syria with “data” and address them in budgetary planning. The effort to bring the data together on Syrian refugees as a single authoritative source, linking different datasets from different databases across the region, was a highly complex and impossible task to achieve. The records in UNHCR’s registration and population database proGres (the most widely used corporate data collection tool out of the 35 databases that were yet to become interoperable) were either incomplete, missing, or unreliable. The series of different versions of proGres population database, which was at last upgraded to V4 in 2021 with 300 million dollars for design and roll-out, did not have the technical functions to diagnose the limitations, take corrective measure on the data, quantify, and monitor the reliability of data generated over time. We could not quantify the size of the problem and undertake a comparison of the demographical factors e.g.; age, gender, dependency, specific needs, family size, place of origin, time of arrival, time of registration between refugee and returnee population and assess the correlations or causation between refugees who did not return to Syria and returnee groups. We could not generate the simple baseline figures on how many new-born babies, separated families, unaccompanied minors, or persons with serious medical conditions from Syria had registered with the organization in the MENA region, or how many persons had been resettled to a third country or returned to Syria. While trying to diagnose the limitations of data for the Syrian refugees, for example, we identified 14,546 adults registered as children at risk, 2,634 children as being in detention since 2013 onwards, thousands of persons with inactive refugee identity cards in the MENA region, and 1,700 cases waiting for UNHCR to issue the decision on refugee status determination since 2005 in one country only. But we coordinated the undertaking of the survey with Syrian refugees across the UNHCR offices in the MENA region with Kobo Toolbox and successfully published the results of the 5th Regional Survey on Refugee Return Perceptions and Intentions in March 2019.

Preparing for Returns to Syria without Data

I started my work as a refugee status determination caseworker  with UNHCR in 2005. At times, the encounters with colleagues surrounding the RSD casework became much more topical than the legal meaning of the testimonies, bordering the shift from common judgements about the applicants lying all the time or not understanding the comments to denunciatory contents. In another UNHCR operation, a reviewing officer, was famous for his blunt sense of humor, jokingly calling a woman “slut” for having committed adultery following her escape from her home country, for example. The head of RSD unit recommending that male applicants must give a detailed account of their first sexual encounter and “the blood stain on their underwear to prove their gayness”. A trans applicant, protesting UNHCR running around in front of the office premises taking their clothes off, exclaiming in despair and anger for having waited for their RSD decision result more than enough and the indifference they faced in UNHCR towards their claim of being raped by the local police, while it made the staff watching the scene smile indoors. 

LGBTQI+ as a Specific Need

Q. Are there any corporate tools that produce data to monitor the compliance of policy and standards with the actions the organization under a specific protection mandate responsibility producing data or activity?This is a big question. There are a set of tools. COMPASS is the programme tool launched in 2021. ProGres is the main example I can give as population registry. Then you have other regional tools. Q. I believe the data in COMPASS is updated by the staff, and not directly connected to proGres or other databases. How does COMPASS help to measure the gap between policy intent and actual impact of services on the populations?That is right, there are things being cooked, but not complete yet. It should be one of the ultimate objectives of Orion. To be honest, it might not really say where the gap is, it would connect the tools and databases. Ultimately, it will fulfil the official version of things, and not everything. It would require a sort of a collective vision on what is done at country or level at the moment, scripting queries to get the data out of multiple databases to identify people waiting with inactive refugee identity documentation, waiting for refugee status determination interviews etc. Openly sharing with the donors, or externals would be helpful, but we cannot. At least internally we can change things to acknowledge we have problems and what they are, then consider if and how it can be fixed. Sometimes doing the right thing will hurt more than doing the wrong thing. It often happens for example when there is a number of individuals who should be inactivated in the system, but while your budget is not enough to support the real population, so how do you explain that to the donors? It may trigger the dramatic consequences on the persons of concern. So, sometimes it is with the good intention, not to go public I mean. Pros and cons would tell you what the benefit for the public to know would be. But of course, UNHCR never goes public when something goes wrong. Fixing something may also cost more than keeping it e.g. proGres V4. It would cost more to fix it, so we go with it, but at least we can document for everyone to know what we are dealing with and avoid duplicating the same problem in the future while leaving things as they are. Q. Where do the priorities of the donors and the people UNHCR serves sit in that dynamic?Prioritizing donors create some harm sometimes. Take earmarked funding as an example. So, the donors would say I want that money to be used only for population with X nationality, specifically. Your country has populations from Y nationality an Z and everything else in the alphabet, then you find a way to hide the problem. It does not make sense to build a school for that one specific population. It is not with ill-intentions but to serve the best interests of under-funded population groups. When we budget, it is an issue of impact. So when you plan with your budget to send to HQs, which may not be accurate in its decisions around the benefits of the activities, then you end up building or thinking of a project not to lose the money, as unspent budget would mean you have less money the next year. The structure of UNHCR is not helping with the problem either. In a way, you end up building a special bond with the people you serve for, in that specific country or region, thinking it is my people, they are my responsibility, and I need to do more, and help them more. So for frontline responders some time the challenge is about balancing the fund with the needs and for some of them If this comes with underreporting and playing with numbers, they will end up doing it.&nbsp; This has an impact on the people. For HQs, it may be pure numbers, but for the one overseeing that population, it is real people. You don’t think about them as 30 cases, you think of them as 30 people as a caseworker. I don’t know if it makes sense.&nbsp;Q. Very interesting what you say. Going back to donors’ interest in more and better data, especially since the World Humanitarian Summit in 2016, what is your take on siloed metrics to monitor impact, if you could reflect on some real life situations that you witnessed, how would you exemplify erasures, simplifications, errors in data?&nbsp;The key issue word here is siloed. People design programs in isolation, so data remained isolated, and monitoring of the service remain isolated. In most cases to my knowledge at least. We do not look at our tools and programs to work together for the populations we serve. So, if I am monitoring Caritas that provides health services, they monitor the impact vis-a-vis that partner only. One operation has more than 10 partners providing the same services. UNHCR receiving report on how many people received child protection intervention, sex and gender-based violence support, psychosocial intervention etc. How do you monitor the impact of the integrated services? An SGBV survivor may need health care and psychosocial support. We look at the chunk of interventions from an isolated perspective and procedure. SGBV would use the inter-agency tool for SGBV, child protection would be using CPIMS, mental health will use whatever, but these tools don’t talk to each other so how can we measure the impact from the point of view of the persons of concerns? Advocacy on SGBV for example. Is there any way to measure the impact of the SGBV sessions on the number of reported cases? Having a wholistic protection strategy could be a good idea, but it is difficult to implement. It is not easy to do. It is a data gap for sure, it is the data gap in the humanitarian sector, it goes beyond PRIMES.

From Population Data to Compass

Inefficient data management can significantly impair the capacity to offer the intertnational refugee protection. If UNHCR and partner agencies do not have accurate, up-to-date information about the asylum seeker and refugees they serve, including vulnerabilities and protection needs, there could be critical delays or oversights in providing necessary support. This underscores the importance of robust data systems that can track and respond to the needs of refugees to uphold their right to protection effectively. Questions:&nbsp; What are the sources of error and bias in the process of collecting information on persons with specific needs and reporting funding requirements across all UNHCR databases and systems including the data stored in UNHCR's partner agencies? What are the sources of error and bias in the process of collecting information on and reporting funding requirements to assist persons with specific needs across all UNHCR databases and systems including the partner agencies?

Lost in Metrics

This document provides guidance on various aspects of registration and data management in the context of refugee protection. This handbook serves as a practical tool for UNHCR staff, governments, and other stakeholders involved in refugee registration and data management.Here's a brief overview of the key topics covered in the handbook:

Access to Registration: Guidelines for equitable access to registration, ensuring non-discrimination and the right to seek asylum.
Process of Registration: Procedures for identifying and registering refugees, including locations and responsible entities.
Fraud Prevention and Management: Measures to detect and prevent fraud while maintaining registration system integrity.
Data to be Gathered and Recorded: Information collection guidelines for protection, assistance, and planning.
Documentation: Guidance on issuing legal documents enabling access to services, rights, and protection.
Updating and Verification: Procedures for keeping records current to reflect changing circumstances.
Data Protection & Confidentiality: Measures to safeguard personal data and privacy.
Partnerships: Emphasis on collaboration with stakeholders, including governments and NGOs, for effective registration processes.

UNHCR Handbook for Registration: Procedures and Standards for Registration, Population Data Management and Documentation

The Refugee Coordination Model (RCM) delineates the principles and responsibilities guiding UNHCR's coordination of refugee responses. This model emphasizes preparedness, protection strategy, resource mobilization, sectoral setup, information management, and coordination forums as key elements. It reiterates UNHCR's role in supporting states' primary responsibility for refugee protection, and its mandate to provide international protection, assistance, and solutions for refugees. The RCM acknowledges that cooperation among a wide array of stakeholders, including development actors, is essential for comprehensive and sustainable refugee responses. It is adaptable to various refugee situations and seeks to foster resilience in both refugees and host communities. The document also outlines UNHCR's coordination in mixed situations, emphasizing collaboration with other UN agencies and partners.The Refugee Coordination Model (RCM) was formalized in 2013 and further updated in 2019. This model has been developed and revised to provide a framework for leading and coordinating refugee operations, ensuring predictability and clarity in the roles and responsibilities of UNHCR and its partners in responding to refugee situations.

Refugee Coordination Model

Diversity characteristics include ethnicity, nationality, religion, linguistic minority, social status, economic status, disability, LGBTQI. Aside from economic status and social status (on a case-by-case basis) and LGBTQI, these diversity characteristics are important as they underpin international refugee protection and are the basis for refugee status e.g. social group.


UNHCR Policy on Age, Gender, and Diversity Accountability

I am registered with UNHCR.  I want to access services and have specific needs e.g. unaccompanied minor, medical condition, disability. How  can I seek help?

The right to protection for refugees and asylum seekers is a fundamental aspect of international law and human rights principles, ensuring that individuals who have fled their country due to a well-founded fear of persecution are safeguarded from further harm, exploitation, and discrimination in their host country. This right is enshrined in various international instruments, such as the 1951 Refugee Convention and its 1967 Protocol, and mandates that states and international bodies like UNHCR take all necessary measures to provide a secure environment for refugees.

Effective protection includes a range of measures such as legal assistance, access to fair and efficient asylum procedures, physical safety from violence, and the provision of essential services like healthcare and education. It also encompasses safeguarding individuals' rights and dignity, ensuring that they are not forcibly returned to a country where they face threats, and that they have opportunities for durable solutions such as resettlement, local integration, or voluntary repatriation under safe conditions.

Right to Protection

Since 2017, surveys have been conducted bi- annually to ensure the centrality of refugee voices, intentions, and concerns, better understand past, current and future dynamics related to return and provide an evidence base to support planning and programming. So far, over 13,000 Syrian refugees have been reached through a structured dialogue on their intentions. While such surveys can provide valuable insights and are a useful tool to collect important data and conclude on some findings, they have limitations,
including the fact that the reflection of refugees’ intentions represents a given point in time only and may subsequently change depending on a wide range of factors.

The appeal for donations that UNHCR updates every three to six months or immediately following a significant crisis on its global website, classifying its data in relation to different legal definitions. Using a standardised template, changes are made to datasets and images to reflect the current emergency or the circumstances of forcibly displaced populations, typically presenting a focused, consistent perspective on how UNHCR does what it does to alleviate their suffering.


<ol><li>Inclusive programming built on 2021 Roundtable recommendations.</li><li>Capacity enhancement for staff and partners on LGBTIQ+ issues.</li><li>Over 40% of operations actively addressing LGBTIQ+ needs.</li><li>Consultative stocktaking to assess progress and identify areas for improvement, notably in data collection.</li><li>Extensive training initiatives for staff in multiple regions.</li><li>Development of partnerships with LGBTIQ+ organizations for better service delivery.</li><li>Advocacy and awareness campaigns to uphold LGBTIQ+ rights.</li><li>Provision of tailored services, including health care and emergency assistance for LGBTIQ+ refugees.</li></ol>

Welcome to REFUGEE DATA MINDER tutorial!

Tags