Summary and Key findings
The Somalia Resilience Program (SomReP) is intended to enhance the resilience of vulnerable households and communities in Southern Somalia against cyclical shocks and stressors. SomReP is a consortium of seven international non-governmental organisations (INGOs). The program’s activities focus on improving livelihoods and increasing adaptive capacities of communities and households in Somalia.
On behalf of SomReP, Forcier Consulting is conducting Third Party Monitoring (TPM) of the program in two districts in South Central Somalia: Baidoa, Bay Region and Afgooye, Lower Shabelle Region. This document represents the initial main findings of the end line of the TPM. The end line report will report trends in program results after the program’s completion. Further, the status of the indicators will be compared to the baseline and midline values.
The data collection for this midline study was collected in August 2019. All data was collected, cleaned, and analysed by Forcier Consulting. Analyses compare midline and baseline data to end line results to identify trends among the results. This report follows the structure of the midline and baseline assessments, which was conducted by Forcier Consulting in 2017 and 2018.
A total of 1,590 program beneficiaries were surveyed, including 58% female and 42% male respondents.
Respondents were sampled in urban/peri-urban, pastoral, agro-pastoral, and IDP livelihood zones. The vast majority of respondents were household heads, with an average age of 37 years. A large majority of participants had attended madrasa only (70%), and the average household size was 4.3 members.
Overall, the findings are positive for many program areas, however conditions varied over the period of the study; there was some recovery from the very severe 2016/2017 drought in 2018, but deterioration into 2019.
These poor conditions created challenges for program beneficiaries. As a result, the analyses in the end line study show some mixed results. One limitation of the study is that the negative effects of the drought cannot be differentiated from the positive effects of the program.
The key findings from this report can be summarised as follows:
Livelihoods and Food Security:
As in the baseline and midline studies, the most common livelihood activity in the end line was agricultural – 65% of respondents were agricultural labourers and 22% reported selling crops. In general, Afgooye is more agricultural than Baidoa, where construction labour (26%) and trade (17%) were more common. This means that a large majority of respondents are in livelihoods that are vulnerable to climate change.
There was a large increase in the percentage of respondents receiving some form of assistance from an aid program in the end line (84%) from the midline (62%). Approximately 30% of respondents had received training, which is an increase from the midline (24%) and a very large increase from the baseline (2%).
Satisfaction with the training was very high, 81% of those who had taken training found it to be very helpful. This was confirmed in the qualitative data, in which FGD participants and interviewees reported very high satisfaction with the quality and subject matter of the training.
The ability to give help to others and to obtain help from others remained fairly stable since the midline. A majority of respondents were either very likely or reasonably likely to give support (67%) and receive support (80%). The number of people that respondents felt they could turn to for help had also increased at the end line (2.7 on average), and the percentage of people who felt they had no one to turn to had decreased (8%). Many focus group and interview participants felt that they could turn to SomReP for support and that project activities had built social capital, improving support within the community.
Unfortunately, the use of unprotected water sources has not changed significantly over the three waves of the study and remains common (30% of respondents at the end line). The percentage of people using a protected water source remained consistent from the midline to end line (9% at midline, 10% at the end line). Water sources were also separated into sustainable sources (e.g. boreholes, hand pump wells, public taps, and collected rainwater) and unsustainable water sources (e.g. unprotected surface water, wells, or springs), the former being more reliable during drought. There was significant regression in the use of sustainable water sources. The percentage of respondents using a sustainable water source at the end line (64%) was roughly equivalent to midline (66%).
Household expenditures had decreased significantly since the midline. However, given that rising food prices was named as a shock experienced by many respondents, it seems that decreased expenditures are a result of decreased incomes, not decreased costs. The most common expenditures were food (n = 1,520), water (n = 1,315), and health (n = 1,007). In the preceding month, households had reported spending 46.52 USD on food.
However, income diversification had increased since the midline and, as at the midline, there was no significant difference between men and women in average income diversification.
Food security: Indicators of food security reveal some worrying decline from the midline. Positively, the percentage of respondents who reported acceptable scores on the Food Consumption Scale (FCS) increased from the midline (93%), and the average FCS score had increased from 66 to 75. FCS scores among IDP households were significantly lower than in other livelihood zones, IDP households had higher scores than at the midline.
However, the percentage of households Household Hunger Scale behaviours (no food, sleep hungry, go the whole day without eating) had increased from the midline. Likewise, more households are using coping strategies from the Coping Strategies Index from the midline. While food consumption has improved, household behaviours suggest there may be growing food insecurity.
While it is not possible to establish causality for the deteriorating indicators in sustainable water use, household expenditures, and food security with the quantitative data, the qualitative data suggests that most respondents blame increasing drought conditions between the midline and end line. In focus groups and interviews, participants noted that either water sources had dried up or communities wanted boreholes for more reliable water sources. Worsening weather conditions not only affect water availability but also may be part of the explanation for lower expenditures and increased food security coping strategies. In largely agricultural areas, reduced rainfall and increased temperatures lead to lower-income from crops and may contribute to worsening food security.
The finding that while food consumption is improved, coping strategies are increasing is surprising. This may suggest that while consumption is currently improved, households are anticipating future shortfalls and are preparing by saving. It could also suggest that coping strategies are effectively preventing worsening in food consumption. Without a non-participant comparison group, it is impossible to attribute the decline to project activities or lack thereof; for example, the non-participant population in the area may be experiencing significantly worse conditions and the project is providing buffering for participants.
Social Safety Nets
Risk transfer/sharing: Respondents report receiving more assistance with shocks and hazards as compared to the midline (85% as compared to 62%). Although receipt of cash assistance had decreased from the midline to the end line, it remained the second most common type of assistance received by households. In addition, there has been a large increase in the receipt of agricultural inputs which is now the most common type of assistance. VSLA participation has increased to 38% (from 21% at the midline), and the vast majority of those participating had received assistance from a savings scheme. However, perceptions of this assistance had declined, with fewer participants expressing satisfaction.
Contingency resources: This section will address the overall objective to change the percentage of households using new contingency resources. Respondents reported that their access to and knowledge of contingency resources available to them had improved since the baseline and midline. The percentage of respondents who knew of no resources had decreased from 42% at the baseline, to 25% at the midline, and 18% at the end line. The most common resource named is food reserves (40%), followed by financial savings (28%), and seed reserves (25%).
Shocks, hazards, and vulnerabilities: Respondents were asked what hazards they currently face; there was a very large increase in those reporting drought, which is now the most commonly reported hazard (nearly 80%). In addition, there was a decrease in the percentage of people reporting no hazards since the baseline. However, flash floods had decreased as a concern from the midline. The effects of those hazards were reported to be less severe at the end line, as compared to the midline. In contrast, respondents reported that shocks had a greater effect on their lives, as compared to the midline. In the midline, the most common shock among respondents was sickness or health expenditures (45%); however, that had changed to drought at the end line (53%). Many respondents also reported suffering rising food prices (45%), unemployment in the family (45%), and a sickness or health expenses in the family (31%). This section will help to understand whether households are more resilient to cyclical shocks and stressors, one of the program’s objectives.
Natural Resource Management
There was an increase over the three waves in the knowledge of an NRM/Rangeland committee in the community (27% at the end line). In addition, there was an improvement in the perceived functioning of the NRM committees. The average amount of land under improved technology also increased to approximately 100 ha on average.
Rehabilitation of agricultural water sources was a theme that arose in multiple FGDs and interviews. The Balgure Early Warning/Early Action committee spoke about the improved resilience as a result of rehabilitated canals.
The Lafoole NRM committee stated that their efforts had led to improved water access, less waste of water, and less livestock spoiling water.
In addition, committees conducted sensitization that has led to improved natural resources management. For example, the Lafoole Early Warning/Early Action Committee spoke about the early warnings allowing communities to protect water sources and collect rainwater. In addition, the community established rules for damaging resources, such as overharvesting firewood, and selected fines for violations.
Local Governance Capacity Building
Community-based early warning systems: Knowledge of CbEWS saw a very large increase over the three waves of the study; 30% of the respondents knew of CbEWS in the community at the end line, as compared to 4% at midline and 10% at baseline. What is more, most respondents who knew of the early warning system found it highly functional. On average, 4.5 were considered functional by respondents at the end line. Person-to-person communication was the most common form of early warning system (57%), the phone (41%) and radio (32%) were also common.
Community initiatives facilitated to access support from sub-national and national institutions and authorities: During the baseline, 12.2% said initiatives with the aim to access support from sub-national and national institutions and authorities to respond to and cope with the recurrent shocks and stressors that exist in the community existed, this decreased at the midline (5%), but had increased to 23% at the end line.
Women and marginalized groups involved in local planning and decision-making: At the end line, respondents’ involvement in local decision making rebounded to baseline levels (9%), after decreasing to 6% at the midline. In contrast to the midline, involvement was comparable across genders. Approximately 9% of males and 10% of females reported involvement in local decision-making. At the midline, 8% of males participated, while only 4% of females participated. The same dynamic was true of the gender of the household head, with 10% of maleheaded households (8% at midline) and 9% of female-headed households (4% at midline) participated.