Needs assessments in the Central African Republic

Understanding humanitarian needs is key to responding to humanitarian crises efficiently. Yet in many humanitarian crises, obtaining an accurate picture of humanitarian needs has been a challenge. This has been particularly true in the Central African Republic (CAR), where humanitarian access and resources have been limited. In June 2014, the Assessment Capacities Project (ACAPS) published an analytical report on humanitarian needs assessments in CAR. The Monitoring Needs Assessments report is available in English and French on the ACAPS website at http://acaps.org.  The report aimed to strengthen the humanitarian community’s understanding of and response to the CAR crisis by:

• Analysing the current situation in terms of humanitarian needs assessments.
• Identifying gaps, limitations and lessons.
• Making recommendations on how to improve the humanitarian community’s approach to needs assessments in CAR.

To do this, ACAPS collected and analysed humanitarian needs assessments conducted since December 2013 by humanitarian organisations working in CAR. This article presents the operational context which motivated this exercise, and details its methodology and limitations. It then provides key findings and recommendations from the report.

Rapid changes in the context

The latest crisis in CAR must be viewed against the background of decades of political instability, widespread poverty, weak and fragile state institutions and direct or indirect political and military interference from regional and international players. Relative stability following the signing of a peace agreement in Gabon in 2009 prompted a move to reorient international assistance from relief to recovery programmes, despite the fact that substantial parts of the country were under the constant threat of attacks by armed groups and gangs. The March 2013 coup by the Seleka and the subsequent deterioration of the humanitarian situation altered this approach dramatically. Following a surge of violence in Bangui and Bossangoa in the first week of December that left over 1,000 people dead, the UN declared CAR a Level 3 Emergency.

The deterioration of the humanitarian situation and the sudden increase in needs led to a return to emergency programming and an increase in the number of humanitarian staff in the capital, followed by a gradual increase of staff in the field. Humanitarian coordination mechanisms were also strengthened. However, ensuring an accurate understanding of needs remained a challenge. Some humanitarian agencies continued to implement programmes aimed at addressing the chronic problems that had existed before the conflict began at the end of 2012, while in the same prefecture or even sometimes in the same locations other organisations established emergency programmes for newly displaced people.

Needs assessments: a situation overview

The declaration of a Level 3 Emergency in December 2013 triggered the Multi-Sector/Cluster Initial Rapid Assessment, a coordinated process that led to the publication of the MIRA report in January 2014. The report included both a secondary data review per sector and the findings from a primary data collection and analysis process.

In February and July 2014, ACAPS published Disaster Needs Analysis reports which provided detailed reviews of humanitarian needs and gaps. Analyse des Besoins de Crise, February 2014 and July 2014. The executive summary of the February report is also available in English. See http://acaps.org/  Numerous needs assessments were conducted throughout 2013 by NGOs. Repositories for humanitarian needs assessment reports were created by the Humanitarian Response platform (humanitarianresponse.info) and OCHA (http://sdr. ocharowca.info/SearchDocument.php). Despite these initiatives, major challenges persisted concerning the collection, processing and sharing of information on humanitarian needs.

Needs assessments: gaps and challenges

ACAPS identified three main obstacles to building a solid, shared understanding of the crisis at a sectoral and crosssectoral level. The first concerned partial geographic, thematic and time coverage. Access constraints, such as insecurity and poor roads, mean that primary data collection covers only part of CAR’s territory and population. The onset of the rainy season in May has also hampered access to affected populations outside Bangui. In some prefectures, such as Vakaga in the north-east, hardly any humanitarian agencies were present.

There are also problems around reporting and sharing assessments. Not all needs assessment reports are publicly available, sometimes due to the sensitivity of the information they contain, and within information platforms it is sometimes difficult to locate relevant data. At the same time, given the absence of a systematic monitoring system for planned, ongoing and completed needs assessments, it has been difficult to understand humanitarian needs and gaps per geographic location, time period, sector and population group. Communication gaps between and even within humanitarian organisations, the high turnover of humanitarian staff and insufficient knowledge management have led to a loss of existing knowledge.

Finally, there are concerns around the reliability and quality of information. Even within publicly available assessments, the quality and reliability of information has been inconsistent. Evaluations have been conducted on an ad hoc basis and others have been iterative, making it difficult to determine whether the data is relevant and current. Due to the variety of methodologies used to assess humanitarian needs, comparability between data sets, and therefore the analysis of humanitarian needs, is limited.

The Monitoring Needs Assessments process

To address these gaps, ACAPs launch the Monitoring Needs Assessments process to identify and analyse the key characteristics of needs assessments carried out by humanitarian organisations, with a view to highlighting information gaps and future information needs and informing decision-making about future assessment strategies.

Working closely with humanitarian organisations, ACAPS collected 83 assessment reports during May and June 2014, and indexed them in a database. ACAPS then analysed the needs assessment meta-data, presented the limitations of the study, and validated it through consultations with analysts and practitioners to help ensure its reliability and usefulness.

The study confirmed that there were major information gaps on humanitarian needs in remote inaccessible regions, especially in the north-east. It also challenged some commonly held views about humanitarian needs assessments in CAR. For example, despite a strong focus on IDPs in humanitarian appeals and responses, there turned out to be almost no data on the needs of IDPs outside Bangui. Some of the key findings are described below, and are further discussed in the recommendations.

Geographic coverage of assessments

Key results

• Nearly 40% of the assessments covered the Bangui sub-prefecture.
• Prefectures in the north-east and east of the country, as well as Sangha-Mbaéré in the south-west, were the least assessed.

Related observations

• For the least assessed prefectures, information on the humanitarian situation almost exclusively comes from individual field staff, often informally. It is thus highly sensitive to changes in the situation and staff rotations.
• Due to this limited coverage, it is not possible to compare the severity of humanitarian situations between different regions, making evidence-based prioritisation between them at the national level nearly impossible.

Frequency and timing of assessments

Key results

• Nine of the 83 assessments were conducted monthly, weekly or daily.
• Nearly 90% of assessments were ad hoc or event related. An assessment is event-related when it is systematically triggered by a particular event. This is, for example, the case of the Rapid Response Mechanism (RRM), which is triggered as soon as population movements are observed.

Related observations

• In rapidly evolving humanitarian crises, there is an increased need for regularly updated information on displacement and on humanitarian needs. However, monitoring and regular updating of humanitarian needs outside Bangui was absent (apart from the monitoring of organisations’ operations).
• As the crisis is dynamic, the lack of regular updates renders information rapidly obsolete, leading to a risk that strategic and programming decisions will be based on data and analyses that are inaccurate or no longer relevant.

Assessment methods

Key results

• Fifty-nine of the 83 assessments collected qualitative data, mainly through interviews with key informants and direct observation.

Related observations

• •As a general rule, during the course of a humanitarian crisis the accuracy and quality of information and analysis on humanitarian needs is expected to improve over time, beginning at the outset with rapid assessments at community level and progressing to more in-depth assessments at household and individual level. It is also expected to progress from the identification of needs to their quantification. However, in CAR, six months after a Level 3 crisis was declared, data is still collected mostly at community level, and is mostly qualitative.
• Although some assessment mechanisms are in place, a variety of methods are used for assessing the same needs (qualitative and quantitative; at community and household level), limiting the comparability of data.

Sectoral and geographical coverage

Key results

• Education was the sector which was most extensively assessed at the national level by the cluster.

Related observations

• The geographical coverage of sectoral evaluations is generally insufficient. For every sector apart from education, it is difficult to establish which geographical location should be prioritised.
• The Education cluster collected data through phone interviews with key informants, which was made possible thanks to the cluster’s knowledge of and engagement with local stakeholders. This was not replicated in other clusters, partly due to sector-specific requirements and difficulties in securing a consensus around one given approach.

Coverage by affected population group

Key results

• In Bangui, the majority of evaluations focused on displaced people. • Outside Bangui, no evaluation focused on a particular population category.

Related observations

• The needs of resident and host populations in Bangui, as well as those of IDPs in host families, were rarely if ever assessed.

Lessons and recommendations

The analysis showed that some assessment practices have led to positive outcomes in terms of coverage and data quality, but some difficulties still need to be addressed. In Bangui, security and access conditions are better than in the rest of the country, enabling the deployment of assessment teams. Even so, the transition to conducting systematic and quantitative assessments at household level has been slow. The obstacles to the collection of household-level data still need to be identified and addressed, but could be linked to a lack of awareness and expertise among humanitarian organisations and coordination mechanisms.

In the Monitoring Needs Assessments report, ACAPS highlighted the key challenges identified during the analysis, and suggested a range of concrete measures that could be taken to address these challenges. As an overarching priority for improving needs assessments, ACAPS recommended that actors work towards a more coordinated approach to humanitarian needs assessments and promote better information-sharing, building on existing coordination platforms in CAR such as the clusters and the Comité de Coordination des ONG Internationales (CCO). Actions to be taken include supporting the harmonisation of assessment tools and their validation by all partners; establishing a task force to implement and monitor this; capitalising on actors’ feedback and on good practice to improve assessment methodologies; and monitoring needs assessments in order to measure progress. Humanitarian actors also need to develop and agree a clear data analysis process, an important step in assessing and interpreting primary data. Measures to be taken include joint analysis meetings and the production, endorsement and use of key documents such as the Monitoring Needs Assessments report and secondary data reviews by humanitarian actors.

Technical recommendations were also made, although some are partly dependent on the implementation of the general recommendations made above, as well as on the availability of adequate resources and expertise. Despite the strengthening of humanitarian teams in Bangui, assessment expertise is still seriously lacking. At operational level, achieving more comprehensive geographical, thematic and group-specific coverage is very important. Consistency of assessment terminologies between all organisations (including assessment vocabulary and the names of official administrative divisions) is needed. A tool to assess and monitor the needs of IDPs outside Bangui should be developed, implemented and regularly updated (this was reportedly under way at the time of writing); quantitative assessments should be conducted at individual level in IDP sites in Bangui.

CAR is a complex context, and successful implementation of these recommendations will depend on the capacity of coordination platforms to facilitate such processes, and on the participation of a wide range of humanitarian actors. While humanitarian actors in the field clearly recognised the usefulness of such measures and were willing to engage in such processes further, a number of obstacles to uptake were also identified. Actors and coordination structures are stretched, and it may be difficult to justify spending more time and resources on improving assessments in a context where coordination has already been strengthened. High staff turnover means that raising awareness on why assessments should be improved needs to be a constant process, something for which few resources are available. Actors are working to different priorities and timeframes, which sometimes means that their interest in the topic is limited.

Lola Wilhelm is Information Analyst at the Assessment Capacities Project (ACAPS).