Kiramu ESNFI Post-Distribution Monitoring (PDM) Report

Monitoring & Evaluation Assessment Report

Acknowledgment

I express my sincere gratitude to all individuals and institutions whose contributions made this Post-Distribution Monitoring (PDM) exercise and report possible.

First and foremost, my deepest appreciation goes to the returning internally displaced persons (IDPs) and host community households across Burqqa Sorruma, Lalistu Sommboo, and Caffee Gudina kebeles of Kiremu Woreda, who generously gave their time and shared their experiences during the household interviews. Their participation provided the evidence base for this report and will directly inform improvements in future humanitarian programming.

Special thanks go to the Ethiopia Humanitarian Fund (EHF), whose financial support made the broader life-saving integrated humanitarian assistance project—and this PDM—possible. Without EHF’s generous funding, neither the assistance nor this assessment would have been realized.

I extend my gratitude to the field teams, and trained enumerators—for their dedication in leading data collection, quality assurance, and analysis under challenging security and access conditions. Their professionalism ensured a clean, complete, and internally consistent dataset of 184 respondents.

I acknowledge the support of local government authorities, kebele administrations, community leaders, and representatives in the three accessible kebeles, whose facilitation and coordination enabled safe and effective community engagement and data collection. I am also grateful to the security actors and humanitarian coordination structures whose guidance informed the rapid security assessments that shaped the operational scope of this exercise.

I further recognize the contribution of the project’s implementing and partner staff across the shelter, Multipurpose Cash, Agriculture, WaSH, Health, and Nutrition sectors, whose integrated efforts underpin the assistance assessed in this report.

I acknowledge the beneficiaries and communities in Naccino and Gudina Jiregna kebeles, who could not be reached due to security constraints. Their exclusion from this assessment is a reminder of the access challenges facing humanitarian actors in Kiremu Woreda, and I remain committed to reaching them as conditions permit.

Any errors or omissions in this report remain my responsibility. The findings, interpretations, and recommendations presented herein are intended to strengthen accountability, learning, and the effectiveness of future Emergency Shelter and Non-Food Items (ESNFI) and cash-based interventions in displacement-affected contexts.

Executive Summary

This Post-Distribution Monitoring (PDM) exercise assessed the mixed Emergency Shelter and Non-Food Items (ESNFI) assistance— ESRK plus cash—delivered to returning IDPs in Kiremu Woreda, East Wollega Zone, Oromia. Funded by the Ethiopia Humanitarian Fund (EHF), the project targeted 500 households across five kebeles, but security constraints rendered Naccino and Gudina Jiregna inaccessible, limiting this PDM to three kebeles (Burqqa Sorruma, Lalistu Sommboo, and Caffee Gudina). Using a quantitative cross-sectional design, 184 returnee respondents (all household heads or representatives) were surveyed via KoboCollect at a 95% confidence level and 5% margin of error. Data passed full quality assurance checks, and analysis was descriptive.

Key findings. The intervention was procedurally strong and well-received. Receipt (100% for most items) and utilization (98.9–100%) of shelter NFI were near-universal, quality ratings were overwhelmingly positive (98.9–100% “good”), and cash satisfaction was universal (100%). Awareness, consultation, and timeliness were near-universal, and 95%+ satisfaction was recorded for content, quality, usefulness, and distribution. However, the findings also point to several areas for improvement. On sufficiency, the uniform 12-piece CGI allocation was insufficient for 26.1% of recipients, and the fixed 2 kg nail allocations for 8.2–13.6%, leaving roughly one in four households unable to complete shelter repairs. On protection and accountability, 5.4% reported that vulnerable households were excluded, 0.5% reported a lack of dignity or privacy, and 1.6% reported distribution-related household conflict, while no respondent reported payment solicitation. Distribution also imposed heavy physical burdens (heavy kit 57.1%, long distance 55.4%, no shades 41.3%, overcrowding 24.5%). While information sharing was near-universal (99.5%), only 58.2% knew complaint channels and 77.2% used the CFRM, with reliance on interpersonal rather than confidential channels. The returnee population is demographically strained (mean household size 6.28; 60.6% under 18; 54.3% female-headed; 35.9% elderly-headed), limiting self-recovery.

Conclusion. The intervention achieved its delivery objective with strong procedural performance but fell short of fully meeting shelter needs due to uniform kit design, residual targeting gaps, and an underutilized accountability framework. Key recommendations. (1) Shift to needs-based kit allocation using rapid shelter damage assessments; (2) strengthen inclusive targeting for elderly-, female-, and disability-affected households; (3) sustain and strengthen staff conduct and dignity training with accountability mechanisms; (4) improve distribution site conditions (proximity, shade, seating, latrines, phased distribution, transport support); (5) strengthen CFRM awareness and confidential reporting options; (6) integrate shelter with livelihoods and social protection; (7) build remote monitoring contingency plans for access-constrained areas.

Lessons learned. Distributing uniform kit packages is simple but often insufficient; high satisfaction can mask assistance–need mismatch; dignity and staff conduct matter as much as materials; physical logistics can undermine access; accountability requires more than information sharing; exclusion of vulnerable households persists; security constraints demand adaptive monitoring; household demographics shape recovery; and integrated programming is essential for sustainable recovery.

Chapter-I: Introduction

1.1 Background and Context

Kiremu Woreda, situated in the East Wollega Zone of the Oromia region, has undergone considerable displacement pressures as a result of conflict that have disrupted the lives and livelihoods of returning IDPs and their host communities. To address the most critical humanitarian needs emerging from these displacements, a life-saving integrated humanitarian assistance project was formulated to assist returning IDPs and affected host community members in Kiremu Woreda.

The sectors encompassed in this project comprise shelter, Multipurpose Cash, Agriculture, WaSH, Health, and Nutrition. The project is directed at five kebeles: Burqqa Sorruma (B/Sorruma), Lalistu Sommboo (L/Sommboo), Naccino, Caffee Gudina (C/Gudina), and Gudina Jiregna (G/Jiregna). For shelter response the total planned beneficiary caseload stands at 500 households, delivered via an Emergency Shelter and Non-Food Items (ESNFI) intervention modality that merges Emergency Shelter and Reconstruction Kits (ESRK) with cash assistance.

Nevertheless, implementation realities on the ground differed from the original plan. A rapid security assessment carried out prior to deployment pinpointed access constraints in two of the five targeted kebeles. Consequently, only three kebeles could be reached: Burqqa Sorruma, Lalistu Sommboo, and Caffee Gudina. Naccino and Gudina Jiregna were not accessed owing to security concerns. This post-distribution monitoring (PDM) exercise thus covers only the three accessible kebeles and the beneficiaries who received assistance there.

The PDM is intended to verify whether the assistance provided — mixed ESNFI consisting of ESRK items and cash — reached the intended recipients, was utilized as intended, and contributed to the shelter and basic needs of returning IDPs and vulnerable host community members in the accessible areas.

1.2 Purpose

The purpose of this post-distribution monitoring exercise is to systematically assess the delivery, utilization, sufficiency, quality, and satisfaction outcomes of the mixed ESNFI (ESRK plus cash) assistance provided to returning IDPs and affected host community households in the three accessible kebeles of Kiremu Woreda.

Specifically, the PDM seeks to verify that assistance reached the intended beneficiaries, to measure the extent to which the shelter and NFI items and cash transfers were used as intended and met household needs, to capture beneficiary perceptions of quality, sufficiency, timeliness, and distribution process, to assess the functionality of accountability to affected populations and complaint and feedback response mechanisms, and to generate evidence-based recommendations for corrective action and future programming in a context shaped by both displacement and access constraints.

1.3 Objective

General Objective

To assess the outcomes, effectiveness, and accountability of the life-saving integrated humanitarian assistance — mixed ESNFI (ESRK plus cash) — delivered to returning IDPs and affected host community households in the accessible kebeles of Kiremu Woreda, East Wollega Zone, Oromia region.

Specific Objective

Specific Objectives (SOs)

1. SO1: Receipt, Quality, & Utilization
   To verify that assisted households across the three accessible kebeles received the cash and ESRK shelter and NFI items — including nails, CGI sheets, locks, and hinges — and to assess how these were utilized, whether they were sufficient, and their quality.

2. SO2: Beneficiary Satisfaction
   To measure the level of satisfaction among beneficiaries with the cash component and the ESRK package in terms of content, quantity, quality, usefulness, the distribution process, and the behavior of staff delivering the assistance.

3. SO3: AAP & CFRM Functionality
   To evaluate how well information sharing, awareness of complaint channels, use of complaint mechanisms, and responsiveness functioned under the AAP and CFRM framework during this distribution process.

4. SO4: Process Challenges & Recommendations
   To identify process-related challenges, including logistical constraints, dignity concerns, exclusion of vulnerable households, and payment solicitation, and to generate actionable recommendations for future ESNFI and cash-based interventions in Kiremu Woreda.

1.4 Scope

This post-distribution monitoring exercise covers the mixed ESNFI (ESRK plus cash) intervention implemented in Kiremu Woreda, East Wollega Zone, Oromia region, Ethiopia. Geographically, the PDM is limited to the three kebeles that were accessible at the time of assessment — Burqqa Sorruma, Lalistu Sommboo, and Caffee Gudina — following a rapid security assessment that rendered Naccino and Gudina Jiregna inaccessible. The findings therefore do not represent the experiences of beneficiaries in the two unaccessed kebeles, and any kebele-level conclusions are restricted to the three surveyed locations.

Thematically, the PDM spans the full continuum of assistance outcomes — from receipt and non-receipt of shelter and NFI items, through utilization and sufficiency of ESRK kits, to multidimensional satisfaction with cash, content, quantity, quality, usefulness, distribution, and staff conduct, as well as AAP and CFRM functionality.

It further captures respondent and household profiles, including geographic distribution, displacement and residential status, sex and marital status, household headship, household size, and age- and sex-disaggregated composition, alongside process indicators such as orientation attendance, information sharing, needs consultation, timeliness, distribution challenges, targeting accuracy, payment solicitation, disability support, and household conflict.

The total beneficiary population for the project is 500 households. The PDM sample comprises 184 respondents drawn from the three accessible kebeles, all of whom are returnees who were displaced but have returned to their original area. The exercise is a post-distribution assessment and does not include baseline or endline comparison data, nor does it include non-beneficiary or non-returnee comparison groups. Findings are therefore descriptive and interpretive rather than causally attributive, and they should be read within the limitations imposed by the returnee-only composition of the sample, the absence of the two inaccessible kebeles, and the security-driven access constraints that shaped the assessment.

Chapter-II: Methodology

2.1 PDM Design and Approach

The Post-Distribution Monitoring (PDM) exercise employed a quantitative cross-sectional design to assess the outcomes and implementation experience of the mixed Emergency Shelter and Non-Food Item (ESNFI) assistance, comprising Emergency Shelter Repair Kits (ESRK) and cash assistance, provided to returning internally displaced persons (IDPs) and affected households in Kiremu Woreda. The assessment was conducted after completion of the assistance distribution and therefore provides a retrospective, descriptive assessment of beneficiaries’ experiences and reported outcomes rather than a causal evaluation of the intervention.

The PDM was designed to assess several key dimensions of the assistance, including receipt and utilization of the distributed ESRK items and cash assistance; perceived sufficiency and quality of assistance; beneficiary satisfaction; distribution processes and staff behavior; and Accountability to Affected Populations (AAP), including access to and functionality of complaint and feedback mechanisms. The assessment relied primarily on beneficiary-reported information collected through structured household interviews.

At the time of the assessment, access was constrained by the prevailing security situation. A rapid security assessment identified Naccino and Gudina Jiregna as inaccessible; consequently, data collection was limited to the three kebeles that remained accessible: Burqqa Sorruma, Lalistu Sommboo, and Caffee Gudina. The PDM therefore reflects the experiences of beneficiaries in the accessible locations and should be interpreted within this operational and access context.

2.2 Sampling and Data Collection

The PDM sample was determined using standard statistical parameters of a 95% confidence level and 5% margin of error and was distributed proportionally across the three accessible kebeles. These kebeles represented 350 beneficiaries out of the shelter project’s total planned caseload of 500 households across the five kebeles. A total of 184 respondents were included in the PDM.

Of the 184 respondents, 118 (64.1%) were from Lalistu Sommboo, 60 (32.6%) from Caffee Gudina, and 6 (3.3%) from Burqaa Soorumaa. All respondents (184; 100%) were returnees, reflecting the beneficiary composition of the assessed caseload. No non-displaced or currently displaced comparison group was included; therefore, the findings describe the experiences of returnee beneficiary households and should not be interpreted as a comparison between different displacement-status groups.

The 184 respondents represented the beneficiary population reached in the three accessible kebeles. At household level, 183 respondents (99.5%) were household heads, while 1 respondent (0.5%) was an adult household representative who was not the household head. The predominance of household-head respondents provided direct household-level reporting from the primary household representatives for nearly all interviews.

Data collection was conducted in person at household level across the three accessible kebeles, subject to the security and access conditions prevailing during the assessment period. The PDM process was led by the MEAL team, with oversight from the MEAL Officer, and implemented with the support of trained enumerators and relevant field staff.

2.3 Data Collection Tools and Procedures

Data were collected using a structured quantitative household PDM questionnaire programmed for electronic data capture through KoboCollect on the EU Server. The tool was designed to capture the main dimensions of the intervention and beneficiary experience, including geographic identification; respondent and household demographic characteristics; displacement and residential status; receipt and utilization of ESRK items and cash assistance; perceived sufficiency and quality of assistance; beneficiary satisfaction with the assistance, including ESRK content, quantity, quality, usefulness, and cash assistance; distribution processes; staff behavior; and AAP, complaint, and feedback indicators.

Prior to field deployment, enumerators were oriented on the questionnaire, data-collection procedures, the project context, and the objectives of the PDM. The orientation was intended to promote a common understanding of the questionnaire and consistent administration of the assessment across the accessible kebeles.

Household interviews were conducted with the household head where available. Where the household head was unavailable, an adult household representative was interviewed. Responses were recorded electronically through KoboCollect, allowing standardized administration, direct digital submission, and subsequent data quality checking and analysis by the MEAL team.

The use of an electronic, structured questionnaire also supported consistency in data capture and reduced the need for manual data entry. The collected information was subsequently subjected to data quality checking and quantitative analysis to generate the PDM findings presented in the following sections.

2.4 Data Quality Assurance

A structured and sequential data quality assurance (DQA) process was applied to the final PDM data-set before statistical analysis. The purpose of the validation process was to establish the integrity and analytically readiness of the data-set by assessing record completeness, uniqueness of observations, identification consistency, numerical plausibility, and internal consistency of household-level variables. These checks were applied to the full set of 184 records, with the specific validation procedures and their respective purposes summarized in Table 1.

Table 1: Summary of Data Quality Procedures Followed
Data Quality Check	Procedure Followed	Purpose
Dataset confirmation	Verified the number of records and variables in the final dataset.	To confirm that the correct dataset was loaded for analysis.
Data structure assessment	Reviewed variable names and data types to ensure appropriate data recognition.	To identify potential data-type or structure problems.
Missing data assessment	Calculated missing observations and percentages for each variable.	To assess data completeness and identify unexpected missingness.
Duplicate record check	Checked interview identification codes (pdm_id, ubi) for repeated records.	To ensure that each interview was represented only once.
Interview ID verification	Checked for missing and unique interview identification codes.	To ensure that each record could be uniquely identified.
Numerical value assessment	Reviewed numerical variables for potentially implausible values.	To identify potential data-entry or measurement errors.
Household composition consistency	Compared reported household size with the sum of household members across age/sex groups.	To assess internal consistency of household composition.

Numerical plausibility checks were conducted for key continuous and count variables to identify missing, out-of-range, or potentially erroneous values that could arise from data-entry errors, respondent reporting errors, or inconsistencies during data collection. The checks found complete data across all assessed numerical variables, with valid values recorded for all 184 observations and no missing entries.

Household head age ranged from 17 to 90 years, with a median of 48 years, while reported household size ranged from 1 to 13 members, with a median of 6 members. The recorded age- and sex-disaggregated household member counts also fell within plausible ranges. Although the minimum household-head age was 17 years and one household consisted of a single member, neither observation was considered inherently implausible within the context of the assessed returnee population. Importantly, no values were identified that required exclusion or correction on the basis of numerical implausibility, and no influential or unexplained outliers were detected.

Table 2: Numerical Data Plausibility Checks
Variable	Valid (n)	Minimum	Median	Maximum
Household head age (years)	184	17	48	90
Reported household size	184	1	6	13
Children under 5 years old	184	0	0	3
Children 5 – 17 years old	184	0	3	7
Adults 18 – 59 years old	184	0	2	7
Older persons 60+ years old	184	0	0	2

A further internal consistency check was undertaken to validate household composition data. For each record, the reported total household size was compared against the sum of household members recorded across the respective age- and sex-disaggregated categories. All 184 records passed this validation check, with zero discrepancies between the reported household size and the reconstructed household composition. This exact agreement provides strong evidence of internal coherence in the household roster data and indicates that the age- and sex-disaggregated composition was recorded consistently with the reported household totals.

Table 4: Household Composition Internal Consistency Check
Records Assessed	Records with Mismatch	Consistency Status
184	0	Consistent

Taken together, the DQA process identified no duplicate records, no unidentifiable observations, no missing values in the assessed numerical variables, no implausible numerical values, and no internal inconsistencies in household composition. Consequently, no observations were removed or excluded on data-quality grounds. The complete sample of 184 respondents was retained for analysis, providing a consistent and internally validated data-set for subsequent PDM analysis.

2.5 Data Management and Analysis

Data were collected electronically and underwent cleaning and validation in accordance with the procedures outlined in Section 2.4 prior to analysis. Statistical analysis was conducted using R, an updated version of which is cited on the report’s citation page.

The analysis was primarily descriptive in nature. Categorical and binary variables were summarized using frequencies and proportions, while continuous and count variables were summarized using minimum, median, maximum, and, where appropriate, mean values.

Cross-tabulations were employed to examine key indicators across kebeles and relevant respondent or household characteristics. Multi-response indicators, including distribution challenges and CFRM channel use, were analyzed by enumerating individual response options and calculating their respective prevalence. Unless otherwise stated, percentages were calculated based on the 184 respondents, and no weighting was applied.

2.6 Ethical, Protection and Confidentiality Considerations

The PDM followed humanitarian principles and data protection standards. Participation was voluntary, and respondents were informed of the assessment purpose and their right to decline or withdraw without consequence. Informed consent was obtained before each interview. Given the sensitivity of issues related to displacement, staff conduct, payment solicitation, exclusion, and household conflict, interviews were conducted with attention to privacy, dignity, and respondent safety.

Data were recorded using unique identification codes, with personally identifiable information excluded from the analytical data-set. Access to raw data was restricted to authorized PDM personnel, and sensitive findings were analyzed and reported only in aggregate. Respondents were also informed of available complaint and feedback channels.

2.7 Methodological Limitations

The findings should be interpreted within several important limitations. The PDM covered only three of the five targeted kebeles, as Naccino and Gudina Jiregna were inaccessible because of security constraints; therefore, findings do not represent beneficiaries in the two inaccessible locations.

The sample comprised 184 returnee respondents (100%), with no currently displaced, non-displaced, or non-beneficiary comparison group, limiting comparative and causal interpretation. Kebele representation was uneven, particularly in Burqaa Soorumaa, which contributed only 6 respondents (3.3%); its findings should therefore be interpreted cautiously.

Most indicators were based on self-reported information and may therefore be affected by recall, social desirability, or courtesy bias, particularly for satisfaction and accountability measures. The assessment was also conducted post-distribution without base line, end line, or comparison data and therefore cannot establish causal effects attributable to the intervention.

Finally, the sample of 184 represents only part of the 500-household project caseload, and no weighting was applied. Security conditions may also have influenced respondents’ willingness to disclose sensitive concerns, particularly regarding staff conduct, payment solicitation, exclusion, and household conflict.

Chapter-III: Key Findings

3.1 Respondent and Household Profile

Respondent Profile

The survey captured a total sample of 184 respondents across three operational kebeles. Lalistuu Somboo accounted for the largest share at 118 respondents (64.1%), followed by Caffee Guddinaa with 60 respondents (32.6%), and Burqaa Soorumaa with only 6 respondents (3.3%). This distribution reflects the concentration of returnee populations in Lalistuu Somboo, which serves as the primary settlement hub, while Burqaa Soorumaa’s minimal representation limits the generalizability of kebele-specific conclusions for that location.

All 184 respondents (100.0%) share a single residential status: returnees who were displaced but have returned to their original area. This universal returnee composition means the study captures the post-return phase exclusively, with no non-displaced or currently displaced comparison groups.

Respondent headship status is near-universal: 183 respondents (99.5%) are themselves the household head, while only 1 respondent (0.5%) is not. This is a strong data quality indicator, as head-respondents report their own characteristics directly, reducing proxy reporting bias. The single non-head respondent is almost certainly the widowed female identified above.

The household head age distribution,reveals a population concentrated in the prime adult and older working-age years, with the largest single cohort being those aged 35–39 (32 heads, 17.4%), followed by the 45–49 and 50–54 brackets, each accounting for 22 heads (12.0%). Together, heads aged 35–54 constitute 102 individuals (55.4%), indicating that the majority of returnee households are led by middle-aged adults. A notable secondary concentration appears in the older age brackets, with 66 heads (35.9%) aged 55 and above — including 16 (8.7%) in each of the 55–59 and 60–64 brackets, 10 (5.4%) aged 65–69, 11 (6.0%) aged 70–74, and a combined 13 heads (7.1%) aged 75–89 — pointing to a substantial elderly-led household segment that may face heightened vulnerability and reduced capacity for self-recovery. At the younger end, headship is limited: only 1 head (0.5%) falls in the 15–19 bracket, 7 (3.8%) in 20–24, 8 (4.3%) in 25–29, and 14 (7.6%) in 30–34, reflecting the near-absence of adolescent and young-adult household leadership.

The household head sex profile shows that 100 household heads (54.3%) are female, while 84 (45.7%) are male, making female headship the majority status within returnee households. This likely reflects conflict-induced household restructuring, whereby displacement-related death, disappearance, or separation of male members has left women to assume headship.

Household head marital status indicates that 146 heads (79.3%) are married, 23 (12.5%) are widowed, 12 (6.5%) are single, and 3 (1.6%) are divorced. While married heads dominate, the combined 38 non-married heads (20.7%) represent a demographically significant minority. The 23 widowed heads (12.5%) likely reflect displacement-related mortality or separation, and their intersection with the female-majority headship profile confirms that widowhood and female headship are closely linked in this population.

Household profile

The 184 surveyed households comprise a total population of 1,156 individuals, yielding a mean household size of 6.28 persons and a median of 6. The minimum household size is 1 and the maximum is 13, indicating considerable variation in household composition across the returnee population. The mean of 6.28 exceeds the median of 6, suggesting a slight rightward skew driven by a small number of unusually large households at the upper end of the distribution, while the presence of a single-person household at the lower extreme points to at least one solitary returnee living without family support. The spread from 1 to 13 members reflects the demographic heterogeneity produced by displacement and return, where some households have been consolidated through the absorption of extended kin while others have been reduced through loss, separation, or the inability of members to return.

The population distribution by age cohort and sex reveals a markedly youthful structure. Children under 5 years account for 149 individuals, or 12.9% of the total population, while the 5–17 year cohort is by far the largest single group at 551 individuals, representing 47.7% of the population. Together, these two youngest cohorts constitute 700 individuals, or 60.6% of the total, meaning that nearly two-thirds of the returnee population is under 18 years of age. The 18–49 year cohort comprises 376 individuals (32.5%), and those aged 50 and above number 80 (6.9%). This age pyramid, heavily weighted toward childhood and adolescence with a comparatively narrow adult base, has significant implications for dependency ratios and household labor capacity. A returnee population in which dependents under 18 outnumber working-age adults by a substantial margin places considerable pressure on the 18–49 cohort, which must simultaneously provide care, generate income, and manage the reconstruction of livelihoods and shelter.

The sex distribution across the total population is relatively balanced, with 607 males (52.5%) and 549 females (47.5%), producing a sex ratio of approximately 111 males per 100 females. This overall balance, however, conceals variation across age cohorts. Among children under 5, the distribution is nearly equal, with 76 females and 73 males. In the 5–17 cohort, males outnumber females more noticeably at 306 versus 245, a gap of 61 individuals. The 18–49 cohort is the most balanced of the adult groups, with 190 males and 186 females. In the 50+ cohort, females slightly outnumber males at 42 versus 38. The relative balance in the working-age cohort, alongside the female majority in the oldest cohort, is consistent with the household headship profile, in which women constitute 54.3% of heads, many of them widowed. The deficit of females in the 5–17 cohort relative to males may warrant further investigation, as it could reflect differential return patterns, mortality, or displacement-related separation affecting girls disproportionately.

When the 1,156 household members are set against the 184 household heads, the ratio of dependents and non-head members to heads stands at roughly 5.3 to 1, underscoring the weight borne by household heads in a population where children and adolescents form the demographic majority. Combined with the earlier finding that 99.5% of respondents are household heads, the household profile confirms that the survey captured the primary decision-makers of unusually young and large households, the majority of which are headed by women. The convergence of a youthful population structure, a mean household size above six, and female-majority headship defines the returnee household as a unit under substantial demographic and economic strain, and it establishes the baseline conditions against which service access, livelihood recovery, and reintegration outcomes should be assessed.

3.2 Distribution Process

The distribution process was assessed across all 184 respondents, and the findings reveal a broadly well-executed operation with near-universal awareness, consultation, and timeliness, alongside a set of persistent logistical and dignity-related challenges affecting substantial minorities of recipients.

Communication

Awareness and communication were near-universal. All 184 respondents (100.0%) attended the orientation, all 184 (100.0%) judged the orientation sufficient, all 184 (100.0%) reported being informed about distribution details, and 182 (98.9%) reported being informed about the items to be distributed, with only 2 respondents (1.1%) indicating they were not. The dominant communication channel was public meetings, cited by 117 respondents (63.6%), followed by organization staff at 37 (20.1%), community leaders or representatives at 12 (6.5%), government officials at 9 (4.9%), and phone calls at 9 (4.9%). The heavy reliance on public meetings indicates a community-wide dissemination approach rather than individualized notification, which accounts for the near-universal awareness but may disadvantage households unable to attend such gatherings.

Consultation and timeliness

Consultation and timeliness were strongly positive. A total of 175 respondents (95.1%) reported that their needs had been consulted extensively, 2 (1.1%) reported moderate consultation, and 7 (3.8%) reported no consultation at all. On timeliness, 157 respondents (85.3%) described the assistance as very timely and 27 (14.7%) as timely, while all 184 respondents (100.0%) confirmed that the distribution itself occurred on time. The combination of universal orientation attendance, near-universal information provision, extensive needs consultation, and fully on-time distribution points to a process that was procedurally well organized from the recipient perspective.

Targeting, integrity, and social impact

All 184 respondents (100.0%) confirmed that the selection criteria were met, and no respondent reported being asked to pay for the assistance (184; 100.0%). On exclusion, 174 respondents (94.6%) reported that vulnerable households were not excluded, while 10 (5.4%) indicated that vulnerable households were excluded, a finding that warrants attention given its implications for targeting accuracy. Special support for persons with disabilities was nearly universal, with 176 respondents (95.7%) reporting it was provided and 8 (4.3%) reporting it was not. Finally, household conflict was reported by 3 respondents (1.6%), while 181 (98.4%) reported no such conflict, indicating that conflict in connection with the distribution process was rare.

Challenges

Against this positive picture, the data on distribution challenges show that logistical and dignity-related problems were widespread. The most frequently reported single challenge was the absence of shades, reported by 43 respondents (23.4%) as a standalone constraint. Long distance to the distribution point was reported by 17 respondents (9.2%) alone, and a heavy kit by 14 respondents (7.6%) alone. More striking is the prevalence of multiple overlapping challenges, as many respondents reported combinations of two or more constraints. The single most common combined challenge was heavy kit together with overcrowding and long distance, reported by 22 respondents (12.0%). Heavy kit combined with long distance was reported by 14 respondents (7.6%), and long distance combined with a heavy kit by a further 9 respondents (4.9%). Numerous smaller combinations appear throughout the distribution, reflecting a highly varied burden of logistical hardship across the recipient population.

3.3 NFI Kits Receipt, Quality, and Utilization

Receipt

Receipt of shelter hardware was near-universal. All 184 respondents (100.0%) received nail no. 8, nail no. 10, roofing nails, CGI sheets, external locks, internal door locks, internal window locks, door T-hinges, and window T-hinges. Only nail no. 12 recorded a single non-receipt (183; 99.5%). This uniformity indicates a well-controlled distribution with no systematic shortfalls, and the lone gap (0.5%) represents an isolated discrepancy rather than a patterned failure.

Utilization

Utilization was almost complete across all items. All 184 respondents (100.0%) used roofing nails, CGI sheets, internal locks, internal window locks, door T-hinges, and window T-hinges. Nail no. 8 and nail no. 12 were each used by 183 respondents (99.5%), and nail no. 10 by 182 (98.9%), with the few non-use cases attributed to loss rather than lack of need or unsuitability.

Sufficiency

Sufficiency was strong for hardware but markedly weaker for CGI sheets. Internal locks (1; 0.5%), internal window locks (2; 1.1%), and window T-hinges (1; 0.5%) recorded negligible shortfalls, while door T-hinges were sufficient for all 184 respondents (100.0%). Nail sufficiency clustered between 86.4% and 91.8%: nail no. 12 (15 insufficient; 8.2%), nail no. 10 (23; 12.5%), roofing nails (24; 13.0%), and nail no. 8 (25; 13.6%). The uniform 2 kg allocation across all nail types suggests the shortfall stems from fixed quantities rather than any specific item being under-supplied. CGI sheets recorded the largest gap in the kit: 48 respondents (26.1%) found the 12-piece allocation insufficient, reflecting variable demand tied to shelter size and damage extent.

Quality

Quality ratings were overwhelmingly positive. All 184 respondents (100.0%) rated CGI sheets, internal locks, internal window locks, door T-hinges, and window T-hinges as good quality. Nail no. 8, nail no. 10, and nail no. 12 were each rated good by 182 respondents (98.9%), and roofing nails by 183 (99.5%), leaving dissatisfaction at 1.1% or below for all nail types.

Syntheses

Taken together, the kit was reliably received, almost universally used, and highly rated for quality. The principal limitations are quantity-related: the fixed 2 kg nail allocation and the 12-piece CGI allocation left roughly one in eight and one in four recipients respectively unable to complete their shelter work, pointing to the need for larger or needs-based allocations.

3.4 Satisfaction

Cash satisfaction

Cash satisfaction was universal: all 184 respondents (100.0%) reported being satisfied with the cash component of the assistance, making it the strongest-performing dimension in the assessment.

Distribution satisfaction

Distribution satisfaction recorded 176 respondents (95.7%) very satisfied and 8 (4.3%) fairly satisfied , with no respondents reporting being unsatisfied. This indicates broad contentment with how the distribution process was organized and executed, making it one of the strongest-performing dimensions in the assessment.

Staff behavior

Staff conduct and behavior is the weakest-performing satisfaction dimension in the assessment and the only indicator to record any outright dissatisfaction. A total of 99 respondents (53.8%) reported being fairly satisfied, 83 (45.1%) very satisfied, and 2 (1.1%) unsatisfied. Combining the positive categories, 182 respondents (98.9%) expressed some level of satisfaction with staff behavior, while 2 respondents (1.1%) fell into the unsatisfied category.

Content, Quantity, and Quality satisfaction

Content satisfaction was similarly high, with 180 respondents (97.8%) reporting being very satisfied and 4 (2.2%) fairly satisfied. Quantity satisfaction recorded 167 respondents (90.8%) very satisfied and 17 (9.2%) fairly satisfied, making it the lowest-rated dimension alongside staff behavior. Quality satisfaction stood at 177 respondents (96.2%) very satisfied and 7 (3.8%) fairly satisfied.

Overall Assistance and Usefulness satisfaction

The overall assessment of assistance reinforces the positive picture. A total of 172 respondents (93.5%) stated “Yes, completely” when asked whether they were satisfied with the assistance, 9 (4.9%) responded “Mostly yes,” and 3 (1.6%) responded “Don’t know.” No respondent selected a negative option, meaning that outright dissatisfaction with the assistance as a whole was effectively zero. Usefulness satisfaction reached 175 respondents (95.1%) very satisfied and 9 (4.9%) fairly satisfied.

Synthesis

Satisfaction across all assistance dimensions was consistently high, with no indicator recording more than negligible outright dissatisfaction. Cash satisfaction was universal (100.0%), and very-satisfied ratings exceeded 95% for kit content (97.8%), item quality (96.2%), distribution process (95.7%), and item usefulness (95.1%). Quantity satisfaction was the lowest of the kit-related dimensions (90.8% very satisfied; 9.2% fairly satisfied), while overall assistance satisfaction remained high (93.5% completely satisfied; 4.9% mostly satisfied; 1.6% uncertain; 0.0% dissatisfied).Staff conduct and behavior was the sole dimension to record any outright dissatisfaction, though this was marginal at 1.1%, and satisfaction remained overwhelmingly positive, with 98.9% of respondents expressing satisfaction — 53.8% fairly satisfied and 45.1% very satisfied.

3.5 AAP and CFRM

Information Sharing and Community Engagement

Information sharing was highly effective across the 184 respondents. A total of 183 respondents (99.5%) reported being completely well informed, while 1 (0.5%) reported being mostly well informed; no respondent reported being poorly informed. This near-universal result is consistent with the earlier findings that respondents attended orientation and received information on distribution details, indicating strong performance in communicating key assistance-related information to beneficiaries.

However, awareness of complaint channels was comparatively lower. While 107 respondents (58.2%) reported knowing the complaint channel completely, 2 (1.1%) reported not really knowing it, 1 (0.5%) reported not knowing it at all, and 1 (0.5%) reported mostly knowing it, with the remaining respondents not providing a clear response. Thus, despite very high general information access, knowledge of how to lodge a complaint remained a notable gap.

Complaint and Feedback Response Mechanism (CFRM)

Perceived availability of the CFRM was high, with 174 respondents (94.6%) confirming that a mechanism was available, compared with 9 (4.9%) who reported that none was available and 1 (0.5%) who did not know. Actual use was lower: 142 respondents (77.2%) reported using the FIDO CFRM, 41 (22.3%) had not used it, and 1 (0.5%) declined to respond. The difference between perceived availability and actual use indicates that awareness or availability did not consistently translate into utilization, although non-use may also reflect households having no complaint or feedback to submit.

CFRM use was strongly concentrated in community-based channels. The community help desk was the most frequently used channel, followed by kebele leaders and mobile phones. Based on the binary indicators, the community help desk was used by 111 respondents (60.3%), kebele leaders by 80 (43.5%), and mobile phones by 43 (23.4%). In contrast, the free hotline was reported by only 2 respondents (1.1%) and the suggestion box by 1 (0.5%). This pattern indicates that beneficiaries relied predominantly on accessible, interpersonal channels rather than formal remote mechanisms, with potential implications for the accessibility and confidentiality of more sensitive complaints.

Response performance was strongly positive. 175 respondents (95.1%) reported that complaints were completely responded to, while 7 (3.8%) reported mostly adequate response and 2 (1.1%) did not know. No respondent reported that a complaint remained unanswered. Similarly, 178 respondents (96.7%) reported that their views were completely considered and 6 (3.3%) reported that they were mostly considered, indicating strong perceived responsiveness among respondents who engaged with the mechanism.

Participation, Responsiveness and Accountability

The overall findings indicate strong perceived responsiveness and accountability in the assessed households. Almost all respondents reported that their views were considered, with 96.7% reporting complete consideration and 3.3% mostly considered, while no respondent reported that complaints were left unanswered. These results suggest that, among beneficiaries who interacted with the accountability system, feedback was generally acknowledged and acted upon.

At the same time, the difference between the near-universal perception of being informed (99.5%), the high reported availability of a CFRM (94.6%), and the lower actual CFRM utilization (77.2%) highlights an important participation gap. Strengthening awareness of specific complaint pathways and maintaining diverse, accessible channels—particularly options that can support confidential reporting—would help ensure that beneficiaries can participate and raise concerns through mechanisms suited to different needs.

Chapter-IV:Interpretation, Conclusion and Recommendations

4.1 Interpretation and Discussion

The PDM findings describe a procedurally strong intervention that delivered high-quality shelter materials and cash to a vulnerable returnee population, but one constrained by standardized kit design, targeting gaps, staff conduct issues, and underutilized accountability mechanisms.

Household vulnerability. Returnee households are under real demographic strain. The average household has 6.28 members, and 60.6% of the population is under 18. Most households are headed by women (54.3%), 12.5% of heads are widowed, and 35.9% of heads are aged 55 or older. This means the working-age group (32.5% of the population) must support more than twice its number in dependents while also rebuilding homes and livelihoods. The lower number of girls compared to boys in the 5–17 age group (245 vs. 306) also warrants protection follow-up.

Delivery and utilization. Almost every household received the full shelter kit. All 184 respondents received nine of the ten tracked items, and only one person did not receive nail no. 12. The materials were also put to use: most items were used by everyone who received them, and the few exceptions were due to loss, not because the items were unsuitable. Distribution was therefore well controlled, with no systematic gaps. However, receiving and using an item is not the same as having enough of it. Door hinges, window hinges, and locks were sufficient for nearly everyone, but nails fell short for roughly one in eight households (8.2% to 13.6%, depending on nail type), and CGI sheets fell short for about one in four (48 respondents; 26.1%). In other words, many households received and used the kit but still could not finish their shelter because some NFI’s quantities were too small.

Satisfaction. Satisfaction was high across the board, and almost no one expressed outright dissatisfaction. Cash assistance received the strongest response, with all 184 respondents satisfied. More than 95% were very satisfied with kit content (97.8%), item quality (96.2%), the distribution process (95.7%), and item usefulness (95.1%). Quantity received the lowest rating of the kit items (90.8% very satisfied). Overall, 172 respondents (93.5%) were completely satisfied with the assistance, 9 (4.9%) mostly satisfied, and 3 (1.6%) unsure; no one was dissatisfied. Staff conduct was the only area where anyone expressed dissatisfaction, and even then it was just 2 respondents (1.1%), with 98.9% expressing satisfaction. The important caveat is that high satisfaction does not necessarily mean needs were fully met — people may express gratitude for what they received even when it was not enough.

Distribution process. The process itself was well run. Everyone attended the orientation, found it useful, and received information about the distribution; 182 respondents (98.9%) also knew which items they would receive. Most people heard about the distribution through public meetings (63.6%), followed by organization staff (20.1%), community leaders (6.5%), government officials (4.9%), and phone calls (4.9%) — meaning information was shared community-wide rather than individually. Consultation and timeliness were strong: 95.1% said their needs were consulted extensively, 85.3% said assistance was very timely, and all 184 confirmed distribution happened on time. Targeting and integrity were also strong: everyone confirmed the selection criteria were met, and no one was asked to pay. However, 10 respondents (5.4%) said vulnerable households were left out, showing that formal compliance did not guarantee accurate inclusion. Support for persons with disabilities was nearly universal (95.7%), and distribution-related household conflict was rare (1.6%). The main problems were physical, not procedural: heavy kit (57.1%), long distance (55.4%), lack of shade (41.3%), and overcrowding (24.5%). Sanitation and dignity concerns were uncommon — 16 respondents (8.7%) noted missing latrines, and only 1 (0.5%) reported a lack of dignity or privacy. Importantly, these physical barriers often came together: the most common combinations were heavy kit with overcrowding and long distance (12.0%), and heavy kit with long distance (7.6%). For many recipients, the hardship was cumulative, not isolated.

AAP and CFRM. Nearly everyone felt well informed (99.5%), and most knew a complaints mechanism existed (94.6%). However, only 58.2% knew exactly how to lodge a complaint, and actual use of the mechanism (77.2%) was lower than its availability. People preferred face-to-face channels — community help desk (60.3%) and kebele leaders (43.5%) — while the hotline (1.1%) and suggestion box (0.5%) were barely used. Among those who did complain, responses were strong: 95.1% said complaints were fully addressed, 96.7% said their views were fully considered, and no complaint went unanswered. Still, the gap between awareness and use, and the reliance on visible, in-person channels, raises a question about whether sensitive concerns — such as staff conduct, requests for payment, or exclusion — were ever reported at all.

Synthesis. The intervention successfully delivered quality assistance to the intended recipients under difficult conditions, but it did not fully meet shelter needs. High satisfaction should be read alongside the fact that many households still lacked enough materials; a well-run process should be read alongside the physical burdens people faced; and a functioning complaints mechanism should be read alongside the fact that many people did not know how to use it.

4.2 Conclusion

The Kiremu ESNFI intervention largely achieved its core objective: delivering shelter materials and cash to returning IDPs in three assessed kebeles. Receipt and utilization were near-universal, material quality was highly rated, cash satisfaction was universal, and the distribution process was procedurally sound with near-universal awareness, consultation, and timeliness. Targeting and payment integrity were also strong, with all respondents confirming selection criteria were met and no one reporting requests for payment.

However, several limitations constrain the intervention’s overall effectiveness. The standardized kit was reliably delivered but not always adequate: 26.1% of recipients found the 12-piece CGI allocation insufficient, and roughly 8–14% reported insufficient nails, leaving a substantial minority unable to complete shelter reconstruction. Targeting was formally compliant but not fully accurate, with 5.4% reporting that vulnerable households were excluded. The physical conditions of distribution imposed significant and often cumulative burdens — heavy kit (57.1%), long distance (55.4%), lack of shade (41.3%), and overcrowding (24.5%) — affecting half or more of recipients. The accountability framework functioned well for those who used it, but only 58.2% knew how to lodge a complaint, and reliance on interpersonal channels over confidential options raises concerns about whether sensitive issues were ever reported. These constraints occurred against a backdrop of acute demographic vulnerability: large households (mean 6.28), a majority of members under 18 (60.6%), and female-majority (54.3%), elderly-headed (35.9%), and widowed (12.5%) household leadership, all of which limit self-recovery capacity.

In sum, the intervention delivered quality assistance with strong procedural performance, but did not fully meet shelter needs due to quantity limitations, standardized kit design, residual targeting gaps, and the physical burden of distribution. These findings provide a clear, evidence-based foundation for corrective action in future programming.

4.3 Recommendations

Immediate (short-term):

1. Revise kit composition to needs-based allocation. Although a damage assessment tool was used for this distribution, future distribution should adopt needs-based allocation. Rapid shelter damage assessments should be conducted by ESNFI technical personnel using the cluster-shared assessment tool to classify households as minor, moderate, or severe, and ESRK kits should then be allocated according to each household’s assessed need.

2. Strengthen targeting to reduce exclusion. For future similar interventions, address the reported exclusion of vulnerable households by revising verification processes to systematically include elderly-headed, female-headed, widowed, and disability-affected households. Community targeting committees should be trained on practical inclusive criteria, and before verification post-targeting audits should be conducted to confirm that vulnerable households were not missed.

3. Address staff conduct and dignity. Although outright dissatisfaction with staff was marginal (1.1%) and reports of undignified treatment minimal (0.5%), these findings do not remove the need for continued vigilance. Training on humanitarian principles, protection, and respectful communication is already delivered on a routine basis and was provided to staff engaged in this distribution; the priority is therefore to sustain and strengthen this practice rather than introduce it. Signed codes of conduct with disciplinary procedures should continue to be required for all distribution staff, confidential misconduct reporting with non-retaliation guarantees should remain in place and be reinforced, and distribution-site supervision should be maintained and, where possible, intensified.

4. Improve distribution site conditions. Relocate points closer to settlements where security permits; provide shade, seating, and latrines; phase or stagger distribution with time slots for vulnerable groups; and provide transport assistance for heavy kits, addressing the four dominant barriers (heavy kit 57.1%, long distance 55.4%, lack of shade 41.3%, overcrowding 24.5%).

5. Strengthen CFRM awareness and accessibility. Close the gap between general information access (99.5% well informed) and complaint channel awareness (58.2%) through targeted awareness-raising; provide information in multiple formats and languages; strengthen interpersonal channels while promoting confidential options (phone call, suggestion box) for sensitive complaints; and adopt proactive outreach with confidentiality assurances.

Medium-term:

1. Integrate shelter with livelihoods and social protection. Combine shelter assistance with cash-for-work, vocational training, and agricultural inputs; prioritize elderly- and female-headed households for labour assistance; link to government social protection schemes.

2. Strengthen needs assessment and beneficiary consultation. Strengthen household-level shelter damage and item-level needs verification; give beneficiaries choice in kit composition where possible.

3. Address gender and protection dimensions. Conduct gender analysis; strengthen GBV, child protection, and disability inclusion; deploy female staff to distribution sites and help desks; investigate the deficit of girls in the 5–17 cohort (245 vs. 306 males).

4. Improve data collection and learning. Include baseline data(can be blended with verification assessment); build remote monitoring contingency plans (phone surveys, third-party monitoring) for access-constrained areas; document and share lessons with clusters and partners.

Policy and advocacy:

1. Advocate for increased and flexible shelter funding. Push donors and the shelter cluster for funding levels that reflect actual needs and support needs-based kit standards rather than fixed kits.

2. Strengthen coordination with government and security actors. Advocate for humanitarian access to security-intense areas such as Naccino and Gudina Jiregna; develop contingency plans for remote delivery; strengthen coordination with government social protection and DRM structures for continuity.

4.4 Lessons Learned

1. Distributing uniform kit packages is simple but often insufficient. Fixed allocations (12 CGI sheets, 2 kg nails) left 26.1% and 8–14% of recipients unable to complete repairs. Adopt needs-based approaches.

2. High satisfaction can mask assistance–need mismatch. Near-universal satisfaction and zero overall dissatisfaction coexisted with 26.1% reporting insufficient CGI sheets, indicating that satisfaction alone does not capture adequacy. Use item-level sufficiency and outcome measures alongside satisfaction.

3. Dignity and staff conduct matter as much as materials. Staff conduct was the only dimension to register outright dissatisfaction (1.1%), and 0.5% reported a lack of dignity or privacy. Treat dignity as a core protection outcome with mandatory training and accountability.

4. Physical logistics can undermine dignity and access. Heavy kit (57.1%), long distance (55.4%), lack of shade (41.3%), and overcrowding (24.5%) affected half or more of recipients, frequently in combination. Plan for accessibility, shade, seating, latrines, and phased distribution whenever possible.

5. Accountability requires more than information sharing. 99.5% were well informed, but only 58.2% knew complaint channels, and interpersonal channels dominated while the hotline (1.1%) and suggestion box (0.5%) went almost unused. Invest in targeted awareness and confidential, accessible channels.

6. Exclusion of vulnerable households persists. 5.4% reported exclusion despite formal targeting compliance (100.0% confirmed criteria were met). Strengthen community verification, post-targeting checks, and outreach to isolated households.

7. Security constraints require adaptive monitoring. Inaccessibility of two kebeles limited coverage to 184/218. Build remote monitoring (phone surveys, third-party, community focal points) into programme design from the outset.

8. Household demographics shape recovery. Large households (mean 6.28), 60.6% under 18, 54.3% female-headed, and 35.9% elderly-headed limit self-recovery. Integrate demographic analysis into targeting and complementary support.

9. Assistance has unintended social impacts. Although distribution-related household conflict was rare (1.6%), its presence still warrants conflict sensitivity, monitoring of social tensions, and adaptive distribution design.

10. Integrated programming is essential for sustainable recovery. Shelter alone cannot address multi-dimensional vulnerability. Strengthen the link of shelter response to livelihoods, social protection, and basic services for holistic recovery.