Mozambique HHS: CCRF Case Study Analysis

library(tidyverse)
library(readxl)
library(janitor)
library(here)
library(scales)
library(knitr)
library(forcats)

knitr::opts_chunk$set(
  message = FALSE,
  warning = FALSE,
  fig.width = 11,
  fig.height = 7,
  dpi = 300
)

theme_set(theme_minimal(base_size = 12))

1. Purpose

This analysis supports the Mozambique case study for the proposed manuscript “Operationalizing the Effectiveness of Fisheries Other Effective Area-based Conservation Measures: A Composite Coastal Resilience Framework”.

The aim is to use the Mozambique household survey data to show how the Composite Coastal Resilience Framework (CCRF) can diagnose relative strengths and weaknesses across resilience domains. Here, we analyse two HHS-based core components:

1. Capacity for Collective Action
2. Sustainable Livelihoods

The analysis is descriptive. The results can show patterns across places and years and can be connected to the package of initiatives implemented in each site, but they should not be interpreted as causal impacts without a stronger evaluation design.

# # Data paths
# 
# This R Markdown assumes the files are stored in `data/raw/` inside the R project.
# The path chunk below uses the same direct `here("data", "raw", ...)` structure you were already using.

# Put the input files in data/raw/ inside this R project, or edit the paths below.
hhs_file <- here("data", "raw", "all_hhs_moz.csv")

# Optional context files, used mainly for checking the project context.
# The core HHS analysis below uses the HHS file and the cleaned crosswalk created in the Rmd.
annex_file <- here(
  "data", "raw",
  "ANNEX III_BAF_EbA_Rare_Site Specific Information_2022-06-20(1).xlsx"
)

beneficiary_file <- here(
  "data", "raw",
  "Beneficiaries and overlaps(1).xlsx"
)



# Load data
hhs_raw <- read_csv(hhs_file, show_col_types = FALSE) %>%
  clean_names()

# The two Excel files are used mainly to define the program context and project-site crosswalk.
# The HHS analysis itself uses `all_hhs_moz.csv`.


# Inspect Excel sheets only if the optional files are present.
# This prevents the Rmd from failing when you only have the HHS CSV available.

if (file.exists(annex_file)) {
  annex_sheets <- excel_sheets(annex_file)
  annex_sheets
} else {
  annex_sheets <- character(0)
  message("Optional annex Excel file not found; continuing with HHS-only analysis.")
}

if (file.exists(beneficiary_file)) {
  beneficiary_sheets <- excel_sheets(beneficiary_file)
  beneficiary_sheets
} else {
  beneficiary_sheets <- character(0)
  message("Optional beneficiary Excel file not found; continuing with HHS-only analysis.")
}


# These objects are useful for checking the original context files.
# The actual analysis below uses a cleaned project-site crosswalk because the Excel files contain merged cells and notes.
# They are optional: if the files are not present, the Rmd still runs.

if (file.exists(annex_file)) {
  livelihoods_raw <- read_excel(annex_file, sheet = "3. Livelihoods") %>%
    clean_names()
} else {
  livelihoods_raw <- NULL
}

if (file.exists(beneficiary_file)) {
  beneficiaries_raw <- read_excel(beneficiary_file, sheet = "Target Reached Individuals") %>%
    clean_names()
} else {
  beneficiaries_raw <- NULL
}


# Helper functions
clean_place <- function(x) {
  x %>%
    as.character() %>%
    iconv(from = "UTF-8", to = "ASCII//TRANSLIT") %>%
    str_to_lower() %>%
    str_replace_all("\\s+", " ") %>%
    str_squish() %>%
    str_replace_all("\\s*-\\s*", "-")
}

mean_na <- function(x) {
  if (all(is.na(x))) {
    return(NA_real_)
  }

  mean(x, na.rm = TRUE)
}

median_na <- function(x) {
  if (all(is.na(x))) {
    return(NA_real_)
  }

  median(x, na.rm = TRUE)
}

p25_na <- function(x) {
  if (all(is.na(x))) {
    return(NA_real_)
  }

  as.numeric(quantile(x, 0.25, na.rm = TRUE))
}

p75_na <- function(x) {
  if (all(is.na(x))) {
    return(NA_real_)
  }

  as.numeric(quantile(x, 0.75, na.rm = TRUE))
}

2. Project-site context and crosswalk

Important harmonization decisions:

• Memba and Memba-sede in the HHS were treated as Memba Sede.
• Insular and Ilha Insular were treated as Ilha Insular.
• Mahilene and Mahelene were treated as Mahelene.
• Namige-sede / Namalungo in the livelihoods sheet is ambiguous because one row combines Namige-sede and Namalungo, while another row lists Namalungo separately.

program_context <- tribble(
  ~project_site,    ~district,              ~population_total, ~reached_individuals, ~program_maturity,            ~fmp_status,                                      ~main_livelihood_package,                                                                 ~site_specific_notes,
  "Sanculo",        "Ilha de Mocambique",    38195,             2022,                 "Former/older Rare site",    "FMP approved district/provincial; national review", "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition; aquaculture", "Aquaculture active in Sanculo; two tanks completed and one group started fish farming",
  "Ilha Insular",   "Ilha de Mocambique",     9062,              728,                 "Former/older Rare site",    "FMP approved district/provincial; national review", "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "Full behavior adoption campaign implemented in Ilha district",
  "Quissanga",      "Ilha de Mocambique",    18553,              775,                 "Former/older Rare site",    "FMP approved district/provincial; national review", "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "Aquaculture group reportedly resettled and reallocated to agriculture",
  "Memba Sede",     "Memba",                 22572,             2026,                 "Former/older Rare site",    "FMP approved district/provincial; national review", "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "Watch tower handed over; mangrove and seagrass restoration relevant; mid-term HHS delayed due to security",
  "Baixo Pinda",    "Memba",                  7122,              998,                 "Former/older Rare site",    "FMP approved district/provincial; national review", "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "CCP room completed but handover delayed by security; mangrove/seagrass restoration relevant",
  "Namige Sede",    "Mogincual",             65890,             1735,                 "Newer expansion site",      "FMP under development / approval process",          "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition; apiculture", "Mangrove restoration and hydrological restoration relevant",
  "Namalungo",      "Mogincual",              5000,             1793,                 "Newer expansion site",      "FMP under development / approval process",          "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition; apiculture", "CCP room under construction; beekeeping kits replaced after protests/cyclone losses",
  "Meculuvelane",   "Mogincual",              5117,              856,                 "Newer expansion site",      "FMP under development / approval process",          "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition; apiculture", "CCP room under construction; disaster-risk committees revitalized",
  "Quissimajulo",   "Nacala Porto",           9411,             1892,                 "Newer expansion site",      "FMP under development / approval process",          "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "CCVA results validated with Quissimajulo and Mahelene; CCP room under construction",
  "Mahelene",       "Nacala Porto",           4118,             1832,                 "Newer expansion site",      "FMP under development / approval process",          "Microenterprises; regenerative agriculture; TVET; markets; poultry; nutrition",             "CCVA results validated; land/permits secured for CCP room; some cooperative governance concerns noted"
) %>%
  mutate(
    project_site_clean = clean_place(project_site),
    district_clean = clean_place(district)
  )

program_context %>%
  select(project_site, district, population_total, reached_individuals, program_maturity, fmp_status, site_specific_notes) %>%
  kable(caption = "Project-site context")

Project-site context

project_site	district	population_total	reached_individuals	program_maturity	fmp_status	site_specific_notes
Sanculo	Ilha de Mocambique	38195	2022	Former/older Rare site	FMP approved district/provincial; national review	Aquaculture active in Sanculo; two tanks completed and one group started fish farming
Ilha Insular	Ilha de Mocambique	9062	728	Former/older Rare site	FMP approved district/provincial; national review	Full behavior adoption campaign implemented in Ilha district
Quissanga	Ilha de Mocambique	18553	775	Former/older Rare site	FMP approved district/provincial; national review	Aquaculture group reportedly resettled and reallocated to agriculture
Memba Sede	Memba	22572	2026	Former/older Rare site	FMP approved district/provincial; national review	Watch tower handed over; mangrove and seagrass restoration relevant; mid-term HHS delayed due to security
Baixo Pinda	Memba	7122	998	Former/older Rare site	FMP approved district/provincial; national review	CCP room completed but handover delayed by security; mangrove/seagrass restoration relevant
Namige Sede	Mogincual	65890	1735	Newer expansion site	FMP under development / approval process	Mangrove restoration and hydrological restoration relevant
Namalungo	Mogincual	5000	1793	Newer expansion site	FMP under development / approval process	CCP room under construction; beekeeping kits replaced after protests/cyclone losses
Meculuvelane	Mogincual	5117	856	Newer expansion site	FMP under development / approval process	CCP room under construction; disaster-risk committees revitalized
Quissimajulo	Nacala Porto	9411	1892	Newer expansion site	FMP under development / approval process	CCVA results validated with Quissimajulo and Mahelene; CCP room under construction
Mahelene	Nacala Porto	4118	1832	Newer expansion site	FMP under development / approval process	CCVA results validated; land/permits secured for CCP room; some cooperative governance concerns noted

hhs <- hhs_raw %>%
  mutate(
    year = as.integer(year),
    site_name = as.character(g1_community),
    community_clean = clean_place(g1_community),
    municipality_clean = clean_place(g1_municipality),
    province_clean = clean_place(g1_province),
    project_site = case_when(
      community_clean %in% c("sanculo") ~ "Sanculo",
      community_clean %in% c("ilha insular", "insular") ~ "Ilha Insular",
      community_clean %in% c("quissanga") ~ "Quissanga",
      community_clean %in% c("memba", "memba-sede", "memba sede") ~ "Memba Sede",
      community_clean %in% c("baixo pinda", "baixo  pinda") ~ "Baixo Pinda",
      community_clean %in% c("namige sede", "namige-sede") ~ "Namige Sede",
      community_clean %in% c("namalungo") ~ "Namalungo",
      community_clean %in% c("meculuvelane") ~ "Meculuvelane",
      community_clean %in% c("quissimajulo") ~ "Quissimajulo",
      community_clean %in% c("mahelene", "mahilene") ~ "Mahelene",
      TRUE ~ NA_character_
    ),
    is_project_site = !is.na(project_site)
  ) %>%
  left_join(program_context, by = "project_site")

3. General coverage analysis: complete Mozambique HHS

Overall coverage

coverage_summary <- tibble(
  records = nrow(hhs),
  years = paste(sort(unique(hhs$year)), collapse = ", "),
  n_years = n_distinct(hhs$year),
  n_provinces = n_distinct(hhs$g1_province),
  n_municipalities = n_distinct(hhs$g1_municipality),
  n_communities = n_distinct(hhs$g1_community),
  project_site_records = sum(hhs$is_project_site),
  non_project_site_records = sum(!hhs$is_project_site),
  pct_records_in_project_sites = project_site_records / records * 100
)

coverage_summary %>%
  mutate(across(where(is.numeric), ~round(.x, 1))) %>%
  kable(caption = "Overall HHS coverage")

Overall HHS coverage

records	years	n_years	n_provinces	n_municipalities	n_communities	project_site_records	non_project_site_records	pct_records_in_project_sites
7297	2019, 2021, 2023, 2024, 2025, 2026	6	4	9	27	3894	3403	53.4

Records by year

records_by_year <- hhs %>%
  count(year, name = "n_hhs") %>%
  arrange(year)

records_by_year %>%
  kable(caption = "HHS records by survey year")

HHS records by survey year

year	n_hhs
2019	1460
2021	2493
2023	313
2024	711
2025	1865
2026	455

ggplot(records_by_year, aes(x = factor(year), y = n_hhs)) +
  geom_col(width = 0.7) +
  geom_text(aes(label = comma(n_hhs)), vjust = -0.25, size = 3.5) +
  scale_y_continuous(labels = comma, expand = expansion(mult = c(0, 0.08))) +
  labs(
    title = "Mozambique HHS coverage by year",
    x = "Survey year",
    y = "Number of HHS records"
  )

Records by province and year

records_by_province_year <- hhs %>%
  count(g1_province, year, name = "n_hhs") %>%
  arrange(g1_province, year)

records_by_province_year %>%
  kable(caption = "HHS records by province and year")

HHS records by province and year

g1_province	year	n_hhs
Inhambane	2019	560
Inhambane	2021	1288
Inhambane	2025	600
Maputo	2019	165
Maputo	2021	152
Nampula	2019	534
Nampula	2021	1024
Nampula	2023	313
Nampula	2024	711
Nampula	2025	1265
Nampula	2026	455
Sofala	2019	201
Sofala	2021	29

ggplot(records_by_province_year, aes(x = factor(year), y = n_hhs, fill = g1_province)) +
  geom_col(width = 0.75) +
  scale_y_continuous(labels = comma) +
  labs(
    title = "HHS coverage by province and year",
    x = "Survey year",
    y = "Number of HHS records",
    fill = "Province"
  ) +
  theme(legend.position = "bottom")

Records by municipality and year

records_by_municipality_year <- hhs %>%
  count(g1_province, g1_municipality, year, name = "n_hhs") %>%
  arrange(g1_province, g1_municipality, year)

records_by_municipality_year %>%
  kable(caption = "HHS records by municipality and year")

HHS records by municipality and year

g1_province	g1_municipality	year	n_hhs
Inhambane	Inharrime	2019	223
Inhambane	Inharrime	2021	228
Inhambane	Inhassoro	2019	197
Inhambane	Inhassoro	2021	923
Inhambane	Inhassoro	2025	600
Inhambane	Massinga	2019	140
Inhambane	Massinga	2021	137
Maputo	Matutuíne	2019	165
Maputo	Matutuíne	2021	152
Nampula	Ilha de Mocambique	2019	327
Nampula	Ilha de Mocambique	2021	322
Nampula	Ilha de Mocambique	2023	313
Nampula	Ilha de Mocambique	2025	957
Nampula	Memba	2019	207
Nampula	Memba	2021	702
Nampula	Memba	2024	145
Nampula	Memba	2026	274
Nampula	Mogincual	2024	306
Nampula	Mogincual	2025	308
Nampula	Nacala Porto	2024	260
Nampula	Nacala Porto	2026	181
Sofala	Dondo	2019	201
Sofala	Dondo	2021	29

ggplot(records_by_municipality_year,
       aes(x = factor(year), y = fct_reorder(g1_municipality, n_hhs, .fun = sum), fill = n_hhs)) +
  geom_tile(color = "white") +
  geom_text(aes(label = if_else(n_hhs == 0, "", as.character(n_hhs))), size = 3) +
  scale_fill_gradient(low = "grey95", high = "grey25", labels = comma) +
  labs(
    title = "HHS coverage by municipality and year",
    x = "Survey year",
    y = "Municipality",
    fill = "HHS records"
  )

Records by community / site

community_coverage <- hhs %>%
  count(g1_province, g1_municipality, g1_community, year, name = "n_hhs") %>%
  arrange(g1_province, g1_municipality, g1_community, year)

community_totals <- hhs %>%
  count(g1_province, g1_municipality, g1_community, name = "n_hhs") %>%
  arrange(desc(n_hhs))

community_totals %>%
  kable(caption = "Total HHS records by community / site")

Total HHS records by community / site

g1_province	g1_municipality	g1_community	n_hhs
Nampula	Ilha de Mocambique	Ilha Insular	1124
Inhambane	Inhassoro	Fequete	500
Inhambane	Inharrime	Zavora	451
Nampula	Ilha de Mocambique	Sanculo	430
Nampula	Memba	Memba	420
Nampula	Memba	Baixo Pinda	382
Nampula	Ilha de Mocambique	Quissanga	365
Inhambane	Inhassoro	Tsondzo	324
Nampula	Mogincual	Namige Sede	308
Inhambane	Inhassoro	Mucocuene	305
Inhambane	Massinga	Pomene	277
Nampula	Mogincual	Namalungo	263
Nampula	Nacala Porto	Mahelene	244
Inhambane	Inhassoro	Nhagondzo	223
Nampula	Memba	Serissa	207
Nampula	Nacala Porto	Quissimajulo	197
Inhambane	Inhassoro	Vuca	195
Maputo	Matutuíne	Santa Maria	192
Nampula	Memba	Simuco	180
Nampula	Memba	Memba-sede	139
Maputo	Matutuíne	Mabuluku	125
Sofala	Dondo	Sengo	123
Sofala	Dondo	Farol	107
Inhambane	Inhassoro	Petane	99
Inhambane	Inhassoro	Petane1	74
Nampula	Mogincual	Meculuvelane	22
Nampula	Mogincual	Maculuvelane	21

ggplot(community_totals,
       aes(x = n_hhs, y = fct_reorder(g1_community, n_hhs))) +
  geom_col(width = 0.7) +
  scale_x_continuous(labels = comma) +
  labs(
    title = "HHS coverage by community / site",
    x = "Number of HHS records",
    y = "Community / site"
  )

Community-year coverage heatmap

community_year_complete <- hhs %>%
  count(g1_community, year, name = "n_hhs") %>%
  complete(g1_community, year = sort(unique(hhs$year)), fill = list(n_hhs = 0))

ggplot(community_year_complete,
       aes(x = factor(year), y = fct_reorder(g1_community, n_hhs, .fun = sum), fill = n_hhs)) +
  geom_tile(color = "white") +
  geom_text(aes(label = if_else(n_hhs == 0, "", as.character(n_hhs))), size = 2.7) +
  scale_fill_gradient(low = "grey95", high = "grey25", labels = comma) +
  labs(
    title = "HHS coverage by community / site and year",
    subtitle = "Blank cells indicate no HHS records in that site-year",
    x = "Survey year",
    y = "Community / site",
    fill = "HHS records"
  )

4. Coverage analysis: BAF / Rare project sites

Project-site records by year

project_site_coverage <- hhs %>%
  filter(is_project_site) %>%
  count(project_site, district, program_maturity, year, name = "n_hhs") %>%
  complete(project_site, year = sort(unique(hhs$year)), fill = list(n_hhs = 0)) %>%
  group_by(project_site) %>%
  fill(district, program_maturity, .direction = "downup") %>%
  ungroup() %>%
  arrange(district, project_site, year)

project_site_coverage %>%
  kable(caption = "Project-site HHS records by year")

Project-site HHS records by year

project_site	year	district	program_maturity	n_hhs
Ilha Insular	2019	Ilha de Mocambique	Former/older Rare site	327
Ilha Insular	2021	Ilha de Mocambique	Former/older Rare site	125
Ilha Insular	2023	Ilha de Mocambique	Former/older Rare site	106
Ilha Insular	2024	Ilha de Mocambique	Former/older Rare site	0
Ilha Insular	2025	Ilha de Mocambique	Former/older Rare site	566
Ilha Insular	2026	Ilha de Mocambique	Former/older Rare site	0
Quissanga	2019	Ilha de Mocambique	Former/older Rare site	0
Quissanga	2021	Ilha de Mocambique	Former/older Rare site	91
Quissanga	2023	Ilha de Mocambique	Former/older Rare site	103
Quissanga	2024	Ilha de Mocambique	Former/older Rare site	0
Quissanga	2025	Ilha de Mocambique	Former/older Rare site	171
Quissanga	2026	Ilha de Mocambique	Former/older Rare site	0
Sanculo	2019	Ilha de Mocambique	Former/older Rare site	0
Sanculo	2021	Ilha de Mocambique	Former/older Rare site	106
Sanculo	2023	Ilha de Mocambique	Former/older Rare site	104
Sanculo	2024	Ilha de Mocambique	Former/older Rare site	0
Sanculo	2025	Ilha de Mocambique	Former/older Rare site	220
Sanculo	2026	Ilha de Mocambique	Former/older Rare site	0
Baixo Pinda	2019	Memba	Former/older Rare site	1
Baixo Pinda	2021	Memba	Former/older Rare site	101
Baixo Pinda	2023	Memba	Former/older Rare site	0
Baixo Pinda	2024	Memba	Former/older Rare site	145
Baixo Pinda	2025	Memba	Former/older Rare site	0
Baixo Pinda	2026	Memba	Former/older Rare site	135
Memba Sede	2019	Memba	Former/older Rare site	206
Memba Sede	2021	Memba	Former/older Rare site	214
Memba Sede	2023	Memba	Former/older Rare site	0
Memba Sede	2024	Memba	Former/older Rare site	0
Memba Sede	2025	Memba	Former/older Rare site	0
Memba Sede	2026	Memba	Former/older Rare site	139
Meculuvelane	2019	Mogincual	Newer expansion site	0
Meculuvelane	2021	Mogincual	Newer expansion site	0
Meculuvelane	2023	Mogincual	Newer expansion site	0
Meculuvelane	2024	Mogincual	Newer expansion site	0
Meculuvelane	2025	Mogincual	Newer expansion site	22
Meculuvelane	2026	Mogincual	Newer expansion site	0
Namalungo	2019	Mogincual	Newer expansion site	0
Namalungo	2021	Mogincual	Newer expansion site	0
Namalungo	2023	Mogincual	Newer expansion site	0
Namalungo	2024	Mogincual	Newer expansion site	134
Namalungo	2025	Mogincual	Newer expansion site	129
Namalungo	2026	Mogincual	Newer expansion site	0
Namige Sede	2019	Mogincual	Newer expansion site	0
Namige Sede	2021	Mogincual	Newer expansion site	0
Namige Sede	2023	Mogincual	Newer expansion site	0
Namige Sede	2024	Mogincual	Newer expansion site	151
Namige Sede	2025	Mogincual	Newer expansion site	157
Namige Sede	2026	Mogincual	Newer expansion site	0
Mahelene	2019	Nacala Porto	Newer expansion site	0
Mahelene	2021	Nacala Porto	Newer expansion site	0
Mahelene	2023	Nacala Porto	Newer expansion site	0
Mahelene	2024	Nacala Porto	Newer expansion site	134
Mahelene	2025	Nacala Porto	Newer expansion site	0
Mahelene	2026	Nacala Porto	Newer expansion site	110
Quissimajulo	2019	Nacala Porto	Newer expansion site	0
Quissimajulo	2021	Nacala Porto	Newer expansion site	0
Quissimajulo	2023	Nacala Porto	Newer expansion site	0
Quissimajulo	2024	Nacala Porto	Newer expansion site	126
Quissimajulo	2025	Nacala Porto	Newer expansion site	0
Quissimajulo	2026	Nacala Porto	Newer expansion site	71

ggplot(project_site_coverage,
       aes(x = factor(year), y = fct_reorder(project_site, n_hhs, .fun = sum), fill = n_hhs)) +
  geom_tile(color = "white") +
  geom_text(aes(label = if_else(n_hhs == 0, "", as.character(n_hhs))), size = 3) +
  scale_fill_gradient(low = "grey95", high = "grey25", labels = comma) +
  labs(
    title = "Project-site HHS coverage by year",
    subtitle = "This is the main sample-coverage check before comparing site-year scores",
    x = "Survey year",
    y = "Project site",
    fill = "HHS records"
  )

Project-site coverage summary

project_site_coverage_summary <- hhs %>%
  filter(is_project_site) %>%
  group_by(project_site, district, program_maturity) %>%
  summarise(
    n_hhs = n(),
    years_covered = paste(sort(unique(year)), collapse = ", "),
    n_years = n_distinct(year),
    first_year = min(year),
    last_year = max(year),
    .groups = "drop"
  ) %>%
  left_join(
    program_context %>%
      select(project_site, population_total, reached_individuals, fmp_status, site_specific_notes),
    by = "project_site"
  ) %>%
  arrange(district, project_site)

project_site_coverage_summary %>%
  kable(caption = "Project-site coverage summary")

Project-site coverage summary

project_site	district	program_maturity	n_hhs	years_covered	n_years	first_year	last_year	population_total	reached_individuals	fmp_status	site_specific_notes
Ilha Insular	Ilha de Mocambique	Former/older Rare site	1124	2019, 2021, 2023, 2025	4	2019	2025	9062	728	FMP approved district/provincial; national review	Full behavior adoption campaign implemented in Ilha district
Quissanga	Ilha de Mocambique	Former/older Rare site	365	2021, 2023, 2025	3	2021	2025	18553	775	FMP approved district/provincial; national review	Aquaculture group reportedly resettled and reallocated to agriculture
Sanculo	Ilha de Mocambique	Former/older Rare site	430	2021, 2023, 2025	3	2021	2025	38195	2022	FMP approved district/provincial; national review	Aquaculture active in Sanculo; two tanks completed and one group started fish farming
Baixo Pinda	Memba	Former/older Rare site	382	2019, 2021, 2024, 2026	4	2019	2026	7122	998	FMP approved district/provincial; national review	CCP room completed but handover delayed by security; mangrove/seagrass restoration relevant
Memba Sede	Memba	Former/older Rare site	559	2019, 2021, 2026	3	2019	2026	22572	2026	FMP approved district/provincial; national review	Watch tower handed over; mangrove and seagrass restoration relevant; mid-term HHS delayed due to security
Meculuvelane	Mogincual	Newer expansion site	22	2025	1	2025	2025	5117	856	FMP under development / approval process	CCP room under construction; disaster-risk committees revitalized
Namalungo	Mogincual	Newer expansion site	263	2024, 2025	2	2024	2025	5000	1793	FMP under development / approval process	CCP room under construction; beekeeping kits replaced after protests/cyclone losses
Namige Sede	Mogincual	Newer expansion site	308	2024, 2025	2	2024	2025	65890	1735	FMP under development / approval process	Mangrove restoration and hydrological restoration relevant
Mahelene	Nacala Porto	Newer expansion site	244	2024, 2026	2	2024	2026	4118	1832	FMP under development / approval process	CCVA results validated; land/permits secured for CCP room; some cooperative governance concerns noted
Quissimajulo	Nacala Porto	Newer expansion site	197	2024, 2026	2	2024	2026	9411	1892	FMP under development / approval process	CCVA results validated with Quissimajulo and Mahelene; CCP room under construction

Low sample-size warnings for project-site-year estimates

low_sample_project_site_years <- project_site_coverage %>%
  filter(n_hhs > 0, n_hhs < 30) %>%
  arrange(n_hhs)

low_sample_project_site_years %>%
  kable(caption = "Project-site-years with fewer than 30 HHS records")

Project-site-years with fewer than 30 HHS records

project_site	year	district	program_maturity	n_hhs
Baixo Pinda	2019	Memba	Former/older Rare site	1
Meculuvelane	2025	Mogincual	Newer expansion site	22

5. Impact Framework indicators

Indicator interpretation

The analysis focuses on two Impact Framework components that are relevant to the Composite Coastal Resilience Framework case study: Capacity for Collective Action and Sustainable Livelihoods.

Capacity for Collective Action reflects whether community members perceive that they are part of a fair, trusted, capable, and participatory local fisheries governance system. In the context of Fisheries OECMs, this component is important because durable conservation and fisheries-management outcomes depend not only on rules or management plans, but also on whether communities believe they can act collectively to manage resources.

The component is summarized using four indicators:

1. Social Equity in Fisheries Benefits: the share of fishery-dependent respondents who believe they benefit equally from the fishery as other members of the community.

2. Trust in Local Leadership: the share of respondents who trust local decision-makers or local authorities to make decisions that benefit the community over their own personal interests.

3. Collective Efficacy for Fisheries Management Score: the share of respondents who believe their community has the ability to manage the fishery effectively and maximize food and profits.

4. Empowerment & Participation in Management: the share of respondents who believe that local community participation in management will help maintain or improve fish catch.

Together, these indicators provide a practical measure of whether local social and governance conditions are supportive of collective fisheries management.

Sustainable Livelihoods reflects whether households have sufficient economic and food-security resilience to support long-term engagement in fisheries management and conservation. In the context of Fisheries OECMs, this component matters because households facing severe financial stress or food insecurity may have less capacity to comply with management rules, participate in governance, or invest in alternative livelihood strategies.

The component is summarized using two indicators:

1. Household can make ends meet: the share of respondents who report that their household can cover its needs fairly easily, easily, or very easily.

2. Food security: the share of respondents who report that they never worried about not having enough food for everyone in the household during the last 12 months.

For interpretation, the analysis also presents diagnostic negative indicators for financial strain and food worry. These are useful for communicating the livelihood vulnerability behind the Sustainable Livelihoods score.

agree_values <- c("Agree", "Strongly agree")

hhs_if <- hhs %>%
  mutate(
    # Capacity for Collective Action
    social_equity = case_when(
      g8_fishery_benefit_equal == "Yes" ~ 1,
      g8_fishery_benefit_equal == "No" ~ 0,
      g8_fishery_benefit_equal == "I don’t depend on or benefit from the fishery" ~ NA_real_,
      TRUE ~ NA_real_
    ),
    social_equity_sensitivity_nonbenefit_0 = case_when(
      g8_fishery_benefit_equal == "Yes" ~ 1,
      g8_fishery_benefit_equal %in% c("No", "I don’t depend on or benefit from the fishery") ~ 0,
      TRUE ~ NA_real_
    ),
    leadership_trust_local = case_when(
      g8_trust_local_decision %in% agree_values ~ 1,
      !is.na(g8_trust_local_decision) ~ 0,
      TRUE ~ NA_real_
    ),
    collective_efficacy = case_when(
      g8_my_community_ability %in% agree_values ~ 1,
      !is.na(g8_my_community_ability) ~ 0,
      TRUE ~ NA_real_
    ),
    empowerment_participation = case_when(
      g12_agreement_community_participation %in% agree_values ~ 1,
      !is.na(g12_agreement_community_participation) ~ 0,
      TRUE ~ NA_real_
    ),

    # Sustainable Livelihoods
    ends_meet_positive = case_when(
      g13_hh_ends_meet %in% c("Fairly easy", "Easy", "Very easy") ~ 1,
      !is.na(g13_hh_ends_meet) ~ 0,
      TRUE ~ NA_real_
    ),
    food_security_positive = case_when(
      g11_food_worry == "Never" ~ 1,
      g11_food_worry %in% c("Sometimes", "Often") ~ 0,
      TRUE ~ NA_real_
    ),

    # Diagnostic negative indicators
    financial_strain = case_when(
      g13_hh_ends_meet %in% c("With difficulty", "With great difficulty") ~ 1,
      g13_hh_ends_meet %in% c("Fairly easy", "Easy", "Very easy") ~ 0,
      TRUE ~ NA_real_
    ),
    food_worry_any = case_when(
      g11_food_worry %in% c("Sometimes", "Often") ~ 1,
      g11_food_worry == "Never" ~ 0,
      TRUE ~ NA_real_
    ),
    food_worry_often = case_when(
      g11_food_worry == "Often" ~ 1,
      g11_food_worry %in% c("Sometimes", "Never") ~ 0,
      TRUE ~ NA_real_
    )
  )

# Indicator metadata
core_indicator_vars <- c(
  "social_equity",
  "leadership_trust_local",
  "collective_efficacy",
  "empowerment_participation",
  "ends_meet_positive",
  "food_security_positive"
)

diagnostic_indicator_vars <- c(
  "financial_strain",
  "food_worry_any",
  "food_worry_often"
)

indicator_vars <- c(core_indicator_vars, diagnostic_indicator_vars)

indicator_meta <- tribble(
  ~indicator, ~domain, ~label, ~direction,
  "social_equity", "Capacity for Collective Action", "Social Equity in Fisheries Benefits", "positive_core",
  "leadership_trust_local", "Capacity for Collective Action", "Trust in Local Leadership", "positive_core",
  "collective_efficacy", "Capacity for Collective Action", "Collective Efficacy for Fisheries Management Score", "positive_core",
  "empowerment_participation", "Capacity for Collective Action", "Empowerment & Participation in Management", "positive_core",
  "ends_meet_positive", "Sustainable Livelihoods", "Household covers needs", "positive_core",
  "food_security_positive", "Sustainable Livelihoods", "Never worried about enough food", "positive_core",
  "financial_strain", "Sustainable Livelihoods", "Household covers needs with difficulty", "diagnostic_negative",
  "food_worry_any", "Sustainable Livelihoods", "Sometimes/often worried about food", "diagnostic_negative",
  "food_worry_often", "Sustainable Livelihoods", "Often worried about food", "diagnostic_negative"
)

Summary functions

Domain scores are calculated as the simple average of the positive indicator percentages within each domain. This means each indicator receives equal weight within its domain.

Error bars in the score plots show approximate 95% confidence intervals. For single indicators, the intervals are calculated from the respondent-level variation in the binary indicator. For composite domain scores, the intervals are approximated from the component-indicator standard errors under the same equal-weighting logic. Treat these as descriptive uncertainty intervals rather than formal causal inference intervals.

pct_percent <- function(x) {
  mean_na(x) * 100
}

se_percent <- function(x) {
  n_valid <- sum(!is.na(x))

  if (n_valid <= 1) {
    return(NA_real_)
  }

  sd(x, na.rm = TRUE) / sqrt(n_valid) * 100
}

ci_low_percent <- function(x) {
  pct <- pct_percent(x)
  se <- se_percent(x)

  if (is.na(pct) || is.na(se)) {
    return(NA_real_)
  }

  pmax(0, pct - 1.96 * se)
}

ci_high_percent <- function(x) {
  pct <- pct_percent(x)
  se <- se_percent(x)

  if (is.na(pct) || is.na(se)) {
    return(NA_real_)
  }

  pmin(100, pct + 1.96 * se)
}

summarise_indicators <- function(data, group_vars) {
  data %>%
    group_by(across(all_of(group_vars))) %>%
    summarise(
      n_hhs = n(),
      median_income = median_na(g13_hh_average_income),
      income_p25 = p25_na(g13_hh_average_income),
      income_p75 = p75_na(g13_hh_average_income),
      across(
        all_of(indicator_vars),
        list(
          n_valid = ~sum(!is.na(.x)),
          pct = ~pct_percent(.x),
          se = ~se_percent(.x),
          ci_low = ~ci_low_percent(.x),
          ci_high = ~ci_high_percent(.x)
        ),
        .names = "{.col}__{.fn}"
      ),
      .groups = "drop"
    ) %>%
    pivot_longer(
      cols = matches("__"),
      names_to = c("indicator", ".value"),
      names_sep = "__"
    ) %>%
    left_join(indicator_meta, by = "indicator")
}

make_domain_scores <- function(indicator_summary, group_vars) {
  indicator_summary %>%
    filter(direction == "positive_core") %>%
    group_by(across(all_of(group_vars)), domain) %>%
    summarise(
      n_hhs = first(n_hhs),
      domain_score = mean_na(pct),
      domain_se = {
        n_indicators <- sum(!is.na(pct))
        se_vals <- se[!is.na(se)]

        if (n_indicators == 0 || length(se_vals) == 0) {
          NA_real_
        } else {
          sqrt(sum(se_vals^2)) / n_indicators
        }
      },
      ci_low = pmax(0, domain_score - 1.96 * domain_se),
      ci_high = pmin(100, domain_score + 1.96 * domain_se),
      n_indicators_available = sum(!is.na(pct)),
      min_n_valid = min(n_valid, na.rm = TRUE),
      .groups = "drop"
    ) %>%
    mutate(
      min_n_valid = if_else(is.infinite(min_n_valid), NA_real_, as.numeric(min_n_valid))
    )
}

Missingness in selected Impact Framework variables

impact_vars_raw <- c(
  "g8_fishery_benefit_equal",
  "g8_trust_local_decision",
  "g8_my_community_ability",
  "g12_agreement_community_participation",
  "g11_food_worry",
  "g13_hh_ends_meet",
  "g13_hh_average_income"
)

missingness <- hhs_if %>%
  summarise(across(
    all_of(impact_vars_raw),
    list(
      n_missing = ~sum(is.na(.x)),
      pct_missing = ~mean(is.na(.x)) * 100
    ),
    .names = "{.col}__{.fn}"
  )) %>%
  pivot_longer(
    everything(),
    names_to = c("variable", ".value"),
    names_sep = "__"
  ) %>%
  arrange(desc(pct_missing))

missingness %>%
  mutate(pct_missing = round(pct_missing, 1)) %>%
  kable(caption = "Missingness in selected Impact Framework variables")

Missingness in selected Impact Framework variables

variable	n_missing	pct_missing
g12_agreement_community_participation	3004	41.2
g13_hh_average_income	420	5.8
g13_hh_ends_meet	124	1.7
g8_my_community_ability	100	1.4
g11_food_worry	91	1.2
g8_fishery_benefit_equal	90	1.2
g8_trust_local_decision	35	0.5

ggplot(missingness,
       aes(x = pct_missing, y = fct_reorder(variable, pct_missing))) +
  geom_col(width = 0.7) +
  scale_x_continuous(labels = label_percent(scale = 1), limits = c(0, 100)) +
  labs(
    title = "Missingness in selected Impact Framework variables",
    x = "Missing responses",
    y = NULL
  )

6. Whole-dataset CCRF analysis

This section uses the complete Mozambique HHS dataset, not only project sites.

Overall indicator scores

overall_indicators <- hhs_if %>%
  mutate(analysis_group = "All Mozambique HHS") %>%
  summarise_indicators(group_vars = "analysis_group")

overall_indicators %>%
  select(domain, label, direction, n_hhs, n_valid, pct, se, ci_low, ci_high) %>%
  mutate(across(c(pct, se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(domain, direction, desc(pct)) %>%
  kable(caption = "Overall indicator scores in the complete Mozambique HHS")

Overall indicator scores in the complete Mozambique HHS

domain	label	direction	n_hhs	n_valid	pct	se	ci_low	ci_high
Capacity for Collective Action	Empowerment & Participation in Management	positive_core	7297	4293	88.4	0.5	87.4	89.3
Capacity for Collective Action	Social Equity in Fisheries Benefits	positive_core	7297	6811	83.4	0.5	82.6	84.3
Capacity for Collective Action	Trust in Local Leadership	positive_core	7297	7262	80.6	0.5	79.6	81.5
Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	positive_core	7297	7197	71.3	0.5	70.2	72.3
Sustainable Livelihoods	Sometimes/often worried about food	diagnostic_negative	7297	7206	88.3	0.4	87.5	89.0
Sustainable Livelihoods	Household covers needs with difficulty	diagnostic_negative	7297	7173	73.5	0.5	72.4	74.5
Sustainable Livelihoods	Often worried about food	diagnostic_negative	7297	7206	22.0	0.5	21.0	22.9
Sustainable Livelihoods	Household covers needs	positive_core	7297	7173	26.5	0.5	25.5	27.6
Sustainable Livelihoods	Never worried about enough food	positive_core	7297	7206	11.7	0.4	11.0	12.5

overall_indicators %>%
  filter(direction == "positive_core") %>%
  ggplot(aes(x = fct_reorder(label, pct), y = pct)) +
  geom_col(width = 0.7) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.18) +
  coord_flip() +
  facet_wrap(~ domain, scales = "free_y") +
  scale_y_continuous(limits = c(0, 100), labels = label_percent(scale = 1)) +
  labs(
    title = "Overall positive indicator scores in the complete Mozambique HHS",
    subtitle = "Percent of valid responses coded as positive; error bars show approximate 95% CIs",
    x = NULL,
    y = "% positive"
  )

Overall domain scores

overall_domains <- overall_indicators %>%
  make_domain_scores(group_vars = "analysis_group")

overall_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  kable(caption = "Overall CCRF domain scores in the complete Mozambique HHS")

Overall CCRF domain scores in the complete Mozambique HHS

analysis_group	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
All Mozambique HHS	Capacity for Collective Action	7297	80.9	0.2	80.4	81.4	4	4293
All Mozambique HHS	Sustainable Livelihoods	7297	19.1	0.3	18.5	19.8	2	7173

ggplot(overall_domains, aes(x = fct_reorder(domain, domain_score), y = domain_score)) +
  geom_col(width = 0.65) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.16) +
  geom_text(aes(label = round(domain_score, 1)), vjust = -0.7, size = 4) +
  scale_y_continuous(limits = c(0, 100)) +
  labs(
    title = "Overall CCRF diagnostic scores in the complete Mozambique HHS",
    subtitle = "Error bars show approximate 95% CIs",
    x = NULL,
    y = "Mean domain score, 0–100"
  )

Whole-dataset analysis by year

Domain scores by year

year_indicators <- hhs_if %>%
  summarise_indicators(group_vars = "year")

year_domains <- year_indicators %>%
  make_domain_scores(group_vars = "year")

year_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(year, domain) %>%
  kable(caption = "Domain scores by year, complete Mozambique HHS")

Domain scores by year, complete Mozambique HHS

year	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
2019	Capacity for Collective Action	1460	81.6	0.6	80.5	82.7	4	757
2019	Sustainable Livelihoods	1460	26.2	0.8	24.6	27.8	2	1400
2021	Capacity for Collective Action	2493	85.6	0.4	84.8	86.3	4	1181
2021	Sustainable Livelihoods	2493	19.4	0.6	18.3	20.5	2	2425
2023	Capacity for Collective Action	313	95.8	0.6	94.7	96.9	4	261
2023	Sustainable Livelihoods	313	9.6	1.1	7.4	11.8	2	313
2024	Capacity for Collective Action	711	86.6	0.6	85.4	87.8	4	548
2024	Sustainable Livelihoods	711	21.7	1.1	19.6	23.9	2	711
2025	Capacity for Collective Action	1865	73.8	0.5	72.8	74.9	4	1250
2025	Sustainable Livelihoods	1865	13.4	0.5	12.4	14.5	2	1865
2026	Capacity for Collective Action	455	62.0	1.1	59.8	64.2	4	296
2026	Sustainable Livelihoods	455	21.9	1.3	19.4	24.4	2	455

ggplot(year_domains,
       aes(x = year, y = domain_score, group = domain, linetype = domain)) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.12, alpha = 0.7) +
  geom_line(linewidth = 0.8) +
  geom_point(size = 2.4) +
  scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, 25)) +
  scale_x_continuous(breaks = sort(unique(year_domains$year))) +
  labs(
    title = "CCRF domain scores by year: complete Mozambique HHS",
    subtitle = "Use cautiously: site composition varies across years; error bars show approximate 95% CIs",
    x = "Survey year",
    y = "Mean domain score, 0–100",
    linetype = "Domain"
  ) +
  theme(legend.position = "bottom")

Indicator scores by year

year_indicators %>%
  filter(direction == "positive_core") %>%
  mutate(across(c(pct, se, ci_low, ci_high), ~round(.x, 1))) %>%
  select(year, domain, label, n_hhs, n_valid, pct, se, ci_low, ci_high) %>%
  arrange(year, domain, label) %>%
  kable(caption = "Positive indicator scores by year, complete Mozambique HHS")

Positive indicator scores by year, complete Mozambique HHS

year	domain	label	n_hhs	n_valid	pct	se	ci_low	ci_high
2019	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	1460	1389	77.9	1.1	75.7	80.1
2019	Capacity for Collective Action	Empowerment & Participation in Management	1460	757	89.0	1.1	86.8	91.3
2019	Capacity for Collective Action	Social Equity in Fisheries Benefits	1460	1203	78.2	1.2	75.9	80.6
2019	Capacity for Collective Action	Trust in Local Leadership	1460	1440	81.3	1.0	79.3	83.3
2019	Sustainable Livelihoods	Household covers needs	1460	1404	33.8	1.3	31.4	36.3
2019	Sustainable Livelihoods	Never worried about enough food	1460	1400	18.6	1.0	16.6	20.7
2021	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	2493	2464	81.2	0.8	79.7	82.8
2021	Capacity for Collective Action	Empowerment & Participation in Management	2493	1181	88.7	0.9	86.8	90.5
2021	Capacity for Collective Action	Social Equity in Fisheries Benefits	2493	2366	87.9	0.7	86.6	89.2
2021	Capacity for Collective Action	Trust in Local Leadership	2493	2478	84.5	0.7	83.1	86.0
2021	Sustainable Livelihoods	Household covers needs	2493	2425	25.6	0.9	23.9	27.4
2021	Sustainable Livelihoods	Never worried about enough food	2493	2462	13.1	0.7	11.7	14.4
2023	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	313	313	90.1	1.7	86.8	93.4
2023	Capacity for Collective Action	Empowerment & Participation in Management	313	261	98.5	0.8	97.0	100.0
2023	Capacity for Collective Action	Social Equity in Fisheries Benefits	313	312	98.7	0.6	97.5	100.0
2023	Capacity for Collective Action	Trust in Local Leadership	313	313	95.8	1.1	93.6	98.1
2023	Sustainable Livelihoods	Household covers needs	313	313	19.2	2.2	14.8	23.5
2023	Sustainable Livelihoods	Never worried about enough food	313	313	0.0	0.0	0.0	0.0
2024	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	711	711	72.2	1.7	68.9	75.4
2024	Capacity for Collective Action	Empowerment & Participation in Management	711	548	97.3	0.7	95.9	98.6
2024	Capacity for Collective Action	Social Equity in Fisheries Benefits	711	711	88.3	1.2	86.0	90.7
2024	Capacity for Collective Action	Trust in Local Leadership	711	711	88.6	1.2	86.3	90.9
2024	Sustainable Livelihoods	Household covers needs	711	711	20.1	1.5	17.2	23.1
2024	Sustainable Livelihoods	Never worried about enough food	711	711	23.3	1.6	20.2	26.5
2025	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	1865	1865	54.7	1.2	52.4	57.0
2025	Capacity for Collective Action	Empowerment & Participation in Management	1865	1250	81.1	1.1	78.9	83.3
2025	Capacity for Collective Action	Social Equity in Fisheries Benefits	1865	1790	82.7	0.9	81.0	84.5
2025	Capacity for Collective Action	Trust in Local Leadership	1865	1865	76.8	1.0	74.9	78.7
2025	Sustainable Livelihoods	Household covers needs	1865	1865	23.4	1.0	21.5	25.3
2025	Sustainable Livelihoods	Never worried about enough food	1865	1865	3.5	0.4	2.7	4.3
2026	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	455	455	51.2	2.3	46.6	55.8
2026	Capacity for Collective Action	Empowerment & Participation in Management	455	296	90.9	1.7	87.6	94.2
2026	Capacity for Collective Action	Social Equity in Fisheries Benefits	455	429	57.1	2.4	52.4	61.8
2026	Capacity for Collective Action	Trust in Local Leadership	455	455	48.8	2.3	44.2	53.4
2026	Sustainable Livelihoods	Household covers needs	455	455	36.7	2.3	32.3	41.1
2026	Sustainable Livelihoods	Never worried about enough food	455	455	7.0	1.2	4.7	9.4

ggplot(
  year_indicators %>% filter(direction == "positive_core"),
  aes(
    x = year,
    y = pct,
    group = label,
    color = label
  )
) +
  geom_errorbar(
    aes(ymin = ci_low, ymax = ci_high),
    width = 0.10,
    alpha = 0.45,
    linewidth = 0.5
  ) +
  geom_line(linewidth = 0.8) +
  geom_point(size = 2) +
  facet_wrap(~ domain, ncol = 1) +
  scale_y_continuous(
    limits = c(0, 100),
    labels = scales::label_percent(scale = 1)
  ) +
  scale_x_continuous(
    breaks = sort(unique(year_indicators$year))
  ) +
  scale_color_brewer(palette = "Dark2") +
  labs(
    title = "Positive indicator scores by year: complete Mozambique HHS",
    subtitle = "Each line is one Impact Framework indicator; error bars show approximate 95% CIs",
    x = "Survey year",
    y = "% positive",
    color = "Indicator"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    legend.position = "bottom",
    legend.title = element_text(size = 11),
    legend.text = element_text(size = 10),
    strip.text = element_text(face = "bold")
  )

Whole-dataset analysis by site

Domain scores by site

site_indicators <- hhs_if %>%
  summarise_indicators(group_vars = c("site_name", "g1_municipality", "g1_province"))

site_domains <- site_indicators %>%
  make_domain_scores(group_vars = c("site_name", "g1_municipality", "g1_province"))

site_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(domain, domain_score) %>%
  kable(caption = "Domain scores by site, complete Mozambique HHS")

Domain scores by site, complete Mozambique HHS

site_name	g1_municipality	g1_province	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
Quissanga	Ilha de Mocambique	Nampula	Capacity for Collective Action	365	66.9	1.2	64.6	69.2	4	324
Baixo Pinda	Memba	Nampula	Capacity for Collective Action	382	68.6	1.2	66.3	70.9	4	350
Memba-sede	Memba	Nampula	Capacity for Collective Action	139	70.0	1.8	66.5	73.4	4	88
Petane	Inhassoro	Inhambane	Capacity for Collective Action	99	70.0	2.3	65.5	74.5	4	33
Namalungo	Mogincual	Nampula	Capacity for Collective Action	263	71.6	1.1	69.3	73.8	4	239
Petane1	Inhassoro	Inhambane	Capacity for Collective Action	74	71.7	3.1	65.7	77.7	4	21
Vuca	Inhassoro	Inhambane	Capacity for Collective Action	195	76.4	1.6	73.3	79.5	4	88
Quissimajulo	Nacala Porto	Nampula	Capacity for Collective Action	197	76.6	1.5	73.6	79.6	4	117
Fequete	Inhassoro	Inhambane	Capacity for Collective Action	500	78.0	1.0	76.1	80.0	4	198
Mahelene	Nacala Porto	Nampula	Capacity for Collective Action	244	79.4	1.3	76.8	82.0	4	163
Simuco	Memba	Nampula	Capacity for Collective Action	180	79.5	1.5	76.6	82.4	4	11
Memba	Memba	Nampula	Capacity for Collective Action	420	79.7	1.2	77.4	82.0	4	170
Maculuvelane	Mogincual	Nampula	Capacity for Collective Action	21	81.0	2.8	75.5	86.4	4	21
Ilha Insular	Ilha de Mocambique	Nampula	Capacity for Collective Action	1124	81.3	0.6	80.1	82.4	4	572
Pomene	Massinga	Inhambane	Capacity for Collective Action	277	83.0	1.2	80.7	85.4	4	155
Nhagondzo	Inhassoro	Inhambane	Capacity for Collective Action	223	84.2	1.2	81.8	86.6	4	152
Mucocuene	Inhassoro	Inhambane	Capacity for Collective Action	305	85.1	1.1	83.1	87.2	4	118
Santa Maria	Matutuíne	Maputo	Capacity for Collective Action	192	87.3	1.2	84.9	89.6	4	167
Serissa	Memba	Nampula	Capacity for Collective Action	207	87.7	1.3	85.1	90.3	4	68
Tsondzo	Inhassoro	Inhambane	Capacity for Collective Action	324	88.1	1.0	86.3	90.0	4	174
Namige Sede	Mogincual	Nampula	Capacity for Collective Action	308	88.3	1.0	86.4	90.2	4	195
Mabuluku	Matutuíne	Maputo	Capacity for Collective Action	125	88.5	1.4	85.7	91.3	4	115
Sanculo	Ilha de Mocambique	Nampula	Capacity for Collective Action	430	89.2	0.7	87.8	90.7	4	395
Zavora	Inharrime	Inhambane	Capacity for Collective Action	451	90.1	0.7	88.7	91.4	4	205
Meculuvelane	Mogincual	Nampula	Capacity for Collective Action	22	90.9	2.9	85.2	96.6	4	22
Farol	Dondo	Sofala	Capacity for Collective Action	107	91.7	1.3	89.1	94.2	4	79
Sengo	Dondo	Sofala	Capacity for Collective Action	123	95.1	1.1	92.9	97.3	4	53
Maculuvelane	Mogincual	Nampula	Sustainable Livelihoods	21	0.0	0.0	0.0	0.0	2	21
Baixo Pinda	Memba	Nampula	Sustainable Livelihoods	382	2.5	0.6	1.4	3.6	2	380
Farol	Dondo	Sofala	Sustainable Livelihoods	107	3.3	1.2	0.9	5.7	2	105
Sanculo	Ilha de Mocambique	Nampula	Sustainable Livelihoods	430	4.8	0.7	3.4	6.3	2	424
Serissa	Memba	Nampula	Sustainable Livelihoods	207	7.0	1.3	4.6	9.5	2	199
Petane1	Inhassoro	Inhambane	Sustainable Livelihoods	74	8.8	2.3	4.3	13.3	2	72
Sengo	Dondo	Sofala	Sustainable Livelihoods	123	8.9	1.9	5.2	12.5	2	107
Mucocuene	Inhassoro	Inhambane	Sustainable Livelihoods	305	9.5	1.2	7.2	11.9	2	304
Tsondzo	Inhassoro	Inhambane	Sustainable Livelihoods	324	10.5	1.2	8.1	12.8	2	320
Quissanga	Ilha de Mocambique	Nampula	Sustainable Livelihoods	365	11.0	1.1	8.8	13.2	2	363
Petane	Inhassoro	Inhambane	Sustainable Livelihoods	99	11.1	2.2	6.8	15.4	2	99
Nhagondzo	Inhassoro	Inhambane	Sustainable Livelihoods	223	12.5	1.6	9.5	15.6	2	219
Namalungo	Mogincual	Nampula	Sustainable Livelihoods	263	13.3	1.4	10.6	16.0	2	263
Ilha Insular	Ilha de Mocambique	Nampula	Sustainable Livelihoods	1124	15.3	0.7	13.9	16.7	2	1121
Fequete	Inhassoro	Inhambane	Sustainable Livelihoods	500	16.7	1.1	14.5	19.0	2	492
Memba-sede	Memba	Nampula	Sustainable Livelihoods	139	19.8	2.2	15.5	24.1	2	139
Meculuvelane	Mogincual	Nampula	Sustainable Livelihoods	22	20.5	5.4	9.9	31.0	2	22
Vuca	Inhassoro	Inhambane	Sustainable Livelihoods	195	22.9	2.1	18.8	27.0	2	194
Simuco	Memba	Nampula	Sustainable Livelihoods	180	24.4	2.3	19.9	28.8	2	176
Mahelene	Nacala Porto	Nampula	Sustainable Livelihoods	244	25.4	1.9	21.7	29.1	2	244
Quissimajulo	Nacala Porto	Nampula	Sustainable Livelihoods	197	26.4	2.2	22.1	30.7	2	197
Pomene	Massinga	Inhambane	Sustainable Livelihoods	277	29.5	1.9	25.8	33.3	2	270
Zavora	Inharrime	Inhambane	Sustainable Livelihoods	451	32.3	1.6	29.2	35.3	2	410
Memba	Memba	Nampula	Sustainable Livelihoods	420	37.6	1.7	34.3	40.9	2	408
Mabuluku	Matutuíne	Maputo	Sustainable Livelihoods	125	42.7	3.2	36.5	49.0	2	119
Namige Sede	Mogincual	Nampula	Sustainable Livelihoods	308	42.9	2.0	39.0	46.7	2	308
Santa Maria	Matutuíne	Maputo	Sustainable Livelihoods	192	51.3	2.6	46.2	56.5	2	183

ggplot(site_domains,
       aes(x = fct_reorder(site_name, domain_score), y = domain_score)) +
  geom_col(width = 0.7) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.18) +
  coord_flip() +
  facet_wrap(~ domain, scales = "free_y") +
  scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, 25)) +
  labs(
    title = "CCRF domain scores by site: complete Mozambique HHS",
    subtitle = "Error bars show approximate 95% CIs",
    x = "Site / community",
    y = "Mean domain score, 0–100"
  )

Gap between Collective Action and Sustainable Livelihoods by site

site_domain_wide <- site_domains %>%
  select(site_name, g1_municipality, g1_province, n_hhs, domain, domain_score) %>%
  pivot_wider(names_from = domain, values_from = domain_score) %>%
  mutate(
    domain_gap_collective_action_minus_livelihoods =
      `Capacity for Collective Action` - `Sustainable Livelihoods`
  ) %>%
  arrange(desc(domain_gap_collective_action_minus_livelihoods))

site_domain_wide %>%
  mutate(across(where(is.numeric), ~round(.x, 1))) %>%
  kable(caption = "Site-level gap between collective action and sustainable livelihoods")

Site-level gap between collective action and sustainable livelihoods

site_name	g1_municipality	g1_province	n_hhs	Capacity for Collective Action	Sustainable Livelihoods	domain_gap_collective_action_minus_livelihoods
Farol	Dondo	Sofala	107	91.7	3.3	88.4
Sengo	Dondo	Sofala	123	95.1	8.9	86.2
Sanculo	Ilha de Mocambique	Nampula	430	89.2	4.8	84.4
Maculuvelane	Mogincual	Nampula	21	81.0	0.0	81.0
Serissa	Memba	Nampula	207	87.7	7.0	80.6
Tsondzo	Inhassoro	Inhambane	324	88.1	10.5	77.7
Mucocuene	Inhassoro	Inhambane	305	85.1	9.5	75.6
Nhagondzo	Inhassoro	Inhambane	223	84.2	12.5	71.7
Meculuvelane	Mogincual	Nampula	22	90.9	20.5	70.5
Baixo Pinda	Memba	Nampula	382	68.6	2.5	66.1
Ilha Insular	Ilha de Mocambique	Nampula	1124	81.3	15.3	66.0
Petane1	Inhassoro	Inhambane	74	71.7	8.8	62.9
Fequete	Inhassoro	Inhambane	500	78.0	16.7	61.3
Petane	Inhassoro	Inhambane	99	70.0	11.1	58.9
Namalungo	Mogincual	Nampula	263	71.6	13.3	58.3
Zavora	Inharrime	Inhambane	451	90.1	32.3	57.8
Quissanga	Ilha de Mocambique	Nampula	365	66.9	11.0	55.9
Simuco	Memba	Nampula	180	79.5	24.4	55.1
Mahelene	Nacala Porto	Nampula	244	79.4	25.4	54.0
Pomene	Massinga	Inhambane	277	83.0	29.5	53.5
Vuca	Inhassoro	Inhambane	195	76.4	22.9	53.5
Quissimajulo	Nacala Porto	Nampula	197	76.6	26.4	50.2
Memba-sede	Memba	Nampula	139	70.0	19.8	50.2
Mabuluku	Matutuíne	Maputo	125	88.5	42.7	45.8
Namige Sede	Mogincual	Nampula	308	88.3	42.9	45.5
Memba	Memba	Nampula	420	79.7	37.6	42.1
Santa Maria	Matutuíne	Maputo	192	87.3	51.3	35.9

ggplot(site_domain_wide,
       aes(x = domain_gap_collective_action_minus_livelihoods,
           y = fct_reorder(site_name, domain_gap_collective_action_minus_livelihoods))) +
  geom_col(width = 0.7) +
  labs(
    title = "Diagnostic gap by site",
    subtitle = "Positive values mean Capacity for Collective Action scores higher than Sustainable Livelihoods",
    x = "Capacity for Collective Action minus Sustainable Livelihoods",
    y = "Site / community"
  )

Whole-dataset analysis by site and year

This is the broadest site-year view. It includes all Mozambique HHS sites, not only BAF / Rare project sites. Use this mainly as a coverage and pattern-exploration plot.

Domain scores by site and year: heatmap

site_year_indicators <- hhs_if %>%
  summarise_indicators(group_vars = c("site_name", "g1_municipality", "g1_province", "year"))

site_year_domains <- site_year_indicators %>%
  make_domain_scores(group_vars = c("site_name", "g1_municipality", "g1_province", "year"))

site_year_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(site_name, year, domain) %>%
  kable(caption = "Domain scores by site and year, complete Mozambique HHS")

Domain scores by site and year, complete Mozambique HHS

site_name	g1_municipality	g1_province	year	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
Baixo Pinda	Memba	Nampula	2019	Capacity for Collective Action	1	50.0	NA	NA	NA	4	1
Baixo Pinda	Memba	Nampula	2019	Sustainable Livelihoods	1	0.0	NA	NA	NA	2	1
Baixo Pinda	Memba	Nampula	2021	Capacity for Collective Action	101	52.1	1.6	49.0	55.2	4	96
Baixo Pinda	Memba	Nampula	2021	Sustainable Livelihoods	101	0.0	0.0	0.0	0.0	2	99
Baixo Pinda	Memba	Nampula	2024	Capacity for Collective Action	145	96.9	0.7	95.5	98.3	4	143
Baixo Pinda	Memba	Nampula	2024	Sustainable Livelihoods	145	1.0	0.6	0.0	2.2	2	145
Baixo Pinda	Memba	Nampula	2026	Capacity for Collective Action	135	51.4	1.7	48.0	54.8	4	110
Baixo Pinda	Memba	Nampula	2026	Sustainable Livelihoods	135	5.9	1.4	3.2	8.7	2	135
Farol	Dondo	Sofala	2019	Capacity for Collective Action	107	91.7	1.3	89.1	94.2	4	79
Farol	Dondo	Sofala	2019	Sustainable Livelihoods	107	3.3	1.2	0.9	5.7	2	105
Fequete	Inhassoro	Inhambane	2019	Capacity for Collective Action	196	80.9	1.5	78.0	83.8	4	90
Fequete	Inhassoro	Inhambane	2019	Sustainable Livelihoods	196	19.7	2.0	15.9	23.6	2	189
Fequete	Inhassoro	Inhambane	2021	Capacity for Collective Action	200	78.7	1.7	75.5	82.0	4	62
Fequete	Inhassoro	Inhambane	2021	Sustainable Livelihoods	200	15.1	1.7	11.7	18.4	2	198
Fequete	Inhassoro	Inhambane	2025	Capacity for Collective Action	104	71.7	2.3	67.2	76.1	4	46
Fequete	Inhassoro	Inhambane	2025	Sustainable Livelihoods	104	14.4	2.4	9.7	19.2	2	104
Ilha Insular	Ilha de Mocambique	Nampula	2019	Capacity for Collective Action	327	75.5	1.3	72.9	78.1	4	139
Ilha Insular	Ilha de Mocambique	Nampula	2019	Sustainable Livelihoods	327	25.7	1.6	22.5	28.8	2	325
Ilha Insular	Ilha de Mocambique	Nampula	2021	Capacity for Collective Action	125	95.3	1.1	93.1	97.4	4	65
Ilha Insular	Ilha de Mocambique	Nampula	2021	Sustainable Livelihoods	125	8.9	1.8	5.4	12.4	2	124
Ilha Insular	Ilha de Mocambique	Nampula	2023	Capacity for Collective Action	106	98.8	0.5	97.8	99.9	4	55
Ilha Insular	Ilha de Mocambique	Nampula	2023	Sustainable Livelihoods	106	3.3	1.2	0.9	5.7	2	106
Ilha Insular	Ilha de Mocambique	Nampula	2025	Capacity for Collective Action	566	77.8	0.8	76.2	79.5	4	313
Ilha Insular	Ilha de Mocambique	Nampula	2025	Sustainable Livelihoods	566	13.0	0.9	11.2	14.8	2	566
Mabuluku	Matutuíne	Maputo	2019	Capacity for Collective Action	61	87.5	2.1	83.5	91.6	4	59
Mabuluku	Matutuíne	Maputo	2019	Sustainable Livelihoods	61	48.6	4.1	40.5	56.6	2	56
Mabuluku	Matutuíne	Maputo	2021	Capacity for Collective Action	64	89.4	1.9	85.7	93.1	4	56
Mabuluku	Matutuíne	Maputo	2021	Sustainable Livelihoods	64	36.5	4.1	28.5	44.5	2	63
Maculuvelane	Mogincual	Nampula	2024	Capacity for Collective Action	21	81.0	2.8	75.5	86.4	4	21
Maculuvelane	Mogincual	Nampula	2024	Sustainable Livelihoods	21	0.0	0.0	0.0	0.0	2	21
Mahelene	Nacala Porto	Nampula	2024	Capacity for Collective Action	134	91.9	1.1	89.7	94.1	4	96
Mahelene	Nacala Porto	Nampula	2024	Sustainable Livelihoods	134	19.0	2.4	14.3	23.7	2	134
Mahelene	Nacala Porto	Nampula	2026	Capacity for Collective Action	110	63.2	2.5	58.4	68.1	4	67
Mahelene	Nacala Porto	Nampula	2026	Sustainable Livelihoods	110	33.2	2.7	27.9	38.5	2	110
Meculuvelane	Mogincual	Nampula	2025	Capacity for Collective Action	22	90.9	2.9	85.2	96.6	4	22
Meculuvelane	Mogincual	Nampula	2025	Sustainable Livelihoods	22	20.5	5.4	9.9	31.0	2	22
Memba	Memba	Nampula	2019	Capacity for Collective Action	206	61.9	2.0	57.8	65.9	4	77
Memba	Memba	Nampula	2019	Sustainable Livelihoods	206	19.2	1.8	15.6	22.8	2	200
Memba	Memba	Nampula	2021	Capacity for Collective Action	214	95.1	0.8	93.5	96.7	4	93
Memba	Memba	Nampula	2021	Sustainable Livelihoods	214	55.1	2.4	50.4	59.9	2	207
Memba-sede	Memba	Nampula	2026	Capacity for Collective Action	139	70.0	1.8	66.5	73.4	4	88
Memba-sede	Memba	Nampula	2026	Sustainable Livelihoods	139	19.8	2.2	15.5	24.1	2	139
Mucocuene	Inhassoro	Inhambane	2019	Capacity for Collective Action	1	100.0	NA	NA	NA	3	0
Mucocuene	Inhassoro	Inhambane	2019	Sustainable Livelihoods	1	0.0	NA	NA	NA	2	1
Mucocuene	Inhassoro	Inhambane	2021	Capacity for Collective Action	205	89.4	1.0	87.4	91.4	4	77
Mucocuene	Inhassoro	Inhambane	2021	Sustainable Livelihoods	205	4.9	1.1	2.8	7.0	2	204
Mucocuene	Inhassoro	Inhambane	2025	Capacity for Collective Action	99	76.3	2.3	71.8	80.7	4	41
Mucocuene	Inhassoro	Inhambane	2025	Sustainable Livelihoods	99	19.2	2.7	13.9	24.5	2	99
Namalungo	Mogincual	Nampula	2024	Capacity for Collective Action	134	64.7	1.2	62.4	67.1	4	134
Namalungo	Mogincual	Nampula	2024	Sustainable Livelihoods	134	6.3	1.4	3.5	9.2	2	134
Namalungo	Mogincual	Nampula	2025	Capacity for Collective Action	129	78.3	1.8	74.9	81.8	4	105
Namalungo	Mogincual	Nampula	2025	Sustainable Livelihoods	129	20.5	2.2	16.3	24.8	2	129
Namige Sede	Mogincual	Nampula	2024	Capacity for Collective Action	151	93.5	1.2	91.1	95.8	4	68
Namige Sede	Mogincual	Nampula	2024	Sustainable Livelihoods	151	62.6	2.7	57.3	67.9	2	151
Namige Sede	Mogincual	Nampula	2025	Capacity for Collective Action	157	83.2	1.5	80.3	86.1	4	127
Namige Sede	Mogincual	Nampula	2025	Sustainable Livelihoods	157	23.9	2.0	19.9	27.9	2	157
Nhagondzo	Inhassoro	Inhambane	2021	Capacity for Collective Action	127	89.6	1.4	86.9	92.3	4	76
Nhagondzo	Inhassoro	Inhambane	2021	Sustainable Livelihoods	127	8.5	1.8	5.1	12.0	2	123
Nhagondzo	Inhassoro	Inhambane	2025	Capacity for Collective Action	96	77.5	2.1	73.4	81.6	4	76
Nhagondzo	Inhassoro	Inhambane	2025	Sustainable Livelihoods	96	17.7	2.7	12.4	23.1	2	96
Petane	Inhassoro	Inhambane	2025	Capacity for Collective Action	99	70.0	2.3	65.5	74.5	4	33
Petane	Inhassoro	Inhambane	2025	Sustainable Livelihoods	99	11.1	2.2	6.8	15.4	2	99
Petane1	Inhassoro	Inhambane	2021	Capacity for Collective Action	74	71.7	3.1	65.7	77.7	4	21
Petane1	Inhassoro	Inhambane	2021	Sustainable Livelihoods	74	8.8	2.3	4.3	13.3	2	72
Pomene	Massinga	Inhambane	2019	Capacity for Collective Action	140	81.8	1.7	78.5	85.1	4	90
Pomene	Massinga	Inhambane	2019	Sustainable Livelihoods	140	33.1	2.9	27.5	38.8	2	133
Pomene	Massinga	Inhambane	2021	Capacity for Collective Action	137	84.1	1.7	80.7	87.5	4	65
Pomene	Massinga	Inhambane	2021	Sustainable Livelihoods	137	25.9	2.4	21.2	30.6	2	137
Quissanga	Ilha de Mocambique	Nampula	2021	Capacity for Collective Action	91	89.3	1.7	85.9	92.7	4	57
Quissanga	Ilha de Mocambique	Nampula	2021	Sustainable Livelihoods	91	14.6	2.6	9.5	19.7	2	89
Quissanga	Ilha de Mocambique	Nampula	2023	Capacity for Collective Action	103	98.3	0.6	97.1	99.5	4	102
Quissanga	Ilha de Mocambique	Nampula	2023	Sustainable Livelihoods	103	24.8	2.5	19.9	29.6	2	103
Quissanga	Ilha de Mocambique	Nampula	2025	Capacity for Collective Action	171	37.6	1.4	34.8	40.4	4	165
Quissanga	Ilha de Mocambique	Nampula	2025	Sustainable Livelihoods	171	0.9	0.5	0.0	1.9	2	171
Quissimajulo	Nacala Porto	Nampula	2024	Capacity for Collective Action	126	83.9	1.6	80.9	87.0	4	86
Quissimajulo	Nacala Porto	Nampula	2024	Sustainable Livelihoods	126	19.4	2.5	14.5	24.3	2	126
Quissimajulo	Nacala Porto	Nampula	2026	Capacity for Collective Action	71	61.0	3.3	54.5	67.5	4	31
Quissimajulo	Nacala Porto	Nampula	2026	Sustainable Livelihoods	71	38.7	3.9	31.1	46.3	2	71
Sanculo	Ilha de Mocambique	Nampula	2021	Capacity for Collective Action	106	94.0	1.2	91.6	96.3	4	90
Sanculo	Ilha de Mocambique	Nampula	2021	Sustainable Livelihoods	106	10.4	2.1	6.2	14.6	2	100
Sanculo	Ilha de Mocambique	Nampula	2023	Capacity for Collective Action	104	90.4	1.3	87.8	93.0	4	104
Sanculo	Ilha de Mocambique	Nampula	2023	Sustainable Livelihoods	104	1.0	0.7	0.0	2.3	2	104
Sanculo	Ilha de Mocambique	Nampula	2025	Capacity for Collective Action	220	86.3	1.1	84.2	88.5	4	201
Sanculo	Ilha de Mocambique	Nampula	2025	Sustainable Livelihoods	220	4.1	0.9	2.3	5.9	2	220
Santa Maria	Matutuíne	Maputo	2019	Capacity for Collective Action	104	93.6	1.2	91.2	96.0	4	95
Santa Maria	Matutuíne	Maputo	2019	Sustainable Livelihoods	104	65.2	3.4	58.7	71.8	2	96
Santa Maria	Matutuíne	Maputo	2021	Capacity for Collective Action	88	80.1	2.1	76.0	84.2	4	72
Santa Maria	Matutuíne	Maputo	2021	Sustainable Livelihoods	88	36.0	3.6	28.9	43.1	2	87
Sengo	Dondo	Sofala	2019	Capacity for Collective Action	94	93.6	1.5	90.6	96.5	4	38
Sengo	Dondo	Sofala	2019	Sustainable Livelihoods	94	6.5	1.8	3.1	10.0	2	92
Sengo	Dondo	Sofala	2021	Capacity for Collective Action	29	100.0	0.0	100.0	100.0	4	15
Sengo	Dondo	Sofala	2021	Sustainable Livelihoods	29	23.3	6.7	10.3	36.4	2	15
Serissa	Memba	Nampula	2021	Capacity for Collective Action	207	87.7	1.3	85.1	90.3	4	68
Serissa	Memba	Nampula	2021	Sustainable Livelihoods	207	7.0	1.3	4.6	9.5	2	199
Simuco	Memba	Nampula	2021	Capacity for Collective Action	180	79.5	1.5	76.6	82.4	4	11
Simuco	Memba	Nampula	2021	Sustainable Livelihoods	180	24.4	2.3	19.9	28.8	2	176
Tsondzo	Inhassoro	Inhambane	2021	Capacity for Collective Action	221	92.0	0.9	90.3	93.8	4	109
Tsondzo	Inhassoro	Inhambane	2021	Sustainable Livelihoods	221	7.8	1.3	5.4	10.3	2	217
Tsondzo	Inhassoro	Inhambane	2025	Capacity for Collective Action	103	80.8	2.1	76.7	84.8	4	65
Tsondzo	Inhassoro	Inhambane	2025	Sustainable Livelihoods	103	16.0	2.5	11.1	21.0	2	103
Vuca	Inhassoro	Inhambane	2021	Capacity for Collective Action	96	86.1	1.6	82.9	89.3	4	32
Vuca	Inhassoro	Inhambane	2021	Sustainable Livelihoods	96	26.2	3.1	20.1	32.4	2	95
Vuca	Inhassoro	Inhambane	2025	Capacity for Collective Action	99	68.0	2.4	63.4	72.7	4	56
Vuca	Inhassoro	Inhambane	2025	Sustainable Livelihoods	99	19.7	2.8	14.2	25.2	2	99
Zavora	Inharrime	Inhambane	2019	Capacity for Collective Action	223	89.2	1.1	87.0	91.3	4	89
Zavora	Inharrime	Inhambane	2019	Sustainable Livelihoods	223	31.6	2.3	27.1	36.2	2	185
Zavora	Inharrime	Inhambane	2021	Capacity for Collective Action	228	90.8	0.9	89.0	92.6	4	116
Zavora	Inharrime	Inhambane	2021	Sustainable Livelihoods	228	32.8	2.2	28.5	37.1	2	219

ggplot(site_year_domains,
       aes(x = factor(year), y = fct_reorder(site_name, domain_score, .fun = mean_na), fill = domain_score)) +
  geom_tile(color = "white") +
  facet_wrap(~ domain, ncol = 2) +
  scale_fill_gradient(low = "grey95", high = "grey20", limits = c(0, 100), na.value = "white") +
  labs(
    title = "CCRF domain scores by site and year: complete Mozambique HHS",
    subtitle = "Blank/missing combinations reflect no data or insufficient valid indicator responses. See table for confidence intervals.",
    x = "Survey year",
    y = "Site / community",
    fill = "Score
0–100"
  )

Domain scores by site and year: small multiples

ggplot(site_year_domains,
       aes(x = year, y = domain_score, group = domain, linetype = domain)) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.10, alpha = 0.45) +
  geom_line(linewidth = 0.6) +
  geom_point(size = 1.8) +
  facet_wrap(~ site_name, ncol = 4) +
  scale_y_continuous(limits = c(0, 100), breaks = c(0, 50, 100)) +
  scale_x_continuous(breaks = sort(unique(site_year_domains$year))) +
  labs(
    title = "CCRF domain scores by site and year: complete Mozambique HHS",
    subtitle = "Use cautiously: some site-year estimates are based on small samples; error bars show approximate 95% CIs",
    x = "Survey year",
    y = "Mean domain score, 0–100",
    linetype = "Domain"
  ) +
  theme(
    legend.position = "bottom",
    axis.text.x = element_text(angle = 45, hjust = 1),
    strip.text = element_text(face = "bold")
  )

7. Project-site CCRF analysis

This section focuses only on the BAF / Rare project sites. This is the main case-study subset.

hhs_project <- hhs_if %>%
  filter(is_project_site)

Project-site overall domain scores

project_site_indicators <- hhs_project %>%
  summarise_indicators(group_vars = c("project_site", "district", "program_maturity"))

project_site_domains <- project_site_indicators %>%
  make_domain_scores(group_vars = c("project_site", "district", "program_maturity"))

project_site_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(district, project_site, domain) %>%
  kable(caption = "Domain scores by Wilipihera / Rare project site")

Domain scores by Wilipihera / Rare project site

project_site	district	program_maturity	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
Ilha Insular	Ilha de Mocambique	Former/older Rare site	Capacity for Collective Action	1124	81.3	0.6	80.1	82.4	4	572
Ilha Insular	Ilha de Mocambique	Former/older Rare site	Sustainable Livelihoods	1124	15.3	0.7	13.9	16.7	2	1121
Quissanga	Ilha de Mocambique	Former/older Rare site	Capacity for Collective Action	365	66.9	1.2	64.6	69.2	4	324
Quissanga	Ilha de Mocambique	Former/older Rare site	Sustainable Livelihoods	365	11.0	1.1	8.8	13.2	2	363
Sanculo	Ilha de Mocambique	Former/older Rare site	Capacity for Collective Action	430	89.2	0.7	87.8	90.7	4	395
Sanculo	Ilha de Mocambique	Former/older Rare site	Sustainable Livelihoods	430	4.8	0.7	3.4	6.3	2	424
Baixo Pinda	Memba	Former/older Rare site	Capacity for Collective Action	382	68.6	1.2	66.3	70.9	4	350
Baixo Pinda	Memba	Former/older Rare site	Sustainable Livelihoods	382	2.5	0.6	1.4	3.6	2	380
Memba Sede	Memba	Former/older Rare site	Capacity for Collective Action	559	77.8	1.0	75.8	79.7	4	258
Memba Sede	Memba	Former/older Rare site	Sustainable Livelihoods	559	33.1	1.4	30.4	35.8	2	547
Meculuvelane	Mogincual	Newer expansion site	Capacity for Collective Action	22	90.9	2.9	85.2	96.6	4	22
Meculuvelane	Mogincual	Newer expansion site	Sustainable Livelihoods	22	20.5	5.4	9.9	31.0	2	22
Namalungo	Mogincual	Newer expansion site	Capacity for Collective Action	263	71.6	1.1	69.3	73.8	4	239
Namalungo	Mogincual	Newer expansion site	Sustainable Livelihoods	263	13.3	1.4	10.6	16.0	2	263
Namige Sede	Mogincual	Newer expansion site	Capacity for Collective Action	308	88.3	1.0	86.4	90.2	4	195
Namige Sede	Mogincual	Newer expansion site	Sustainable Livelihoods	308	42.9	2.0	39.0	46.7	2	308
Mahelene	Nacala Porto	Newer expansion site	Capacity for Collective Action	244	79.4	1.3	76.8	82.0	4	163
Mahelene	Nacala Porto	Newer expansion site	Sustainable Livelihoods	244	25.4	1.9	21.7	29.1	2	244
Quissimajulo	Nacala Porto	Newer expansion site	Capacity for Collective Action	197	76.6	1.5	73.6	79.6	4	117
Quissimajulo	Nacala Porto	Newer expansion site	Sustainable Livelihoods	197	26.4	2.2	22.1	30.7	2	197

ggplot(project_site_domains,
       aes(x = fct_reorder(project_site, domain_score), y = domain_score, fill = domain)) +
  geom_col(position = position_dodge(width = 0.75), width = 0.7) +
  geom_errorbar(
    aes(ymin = ci_low, ymax = ci_high),
    position = position_dodge(width = 0.75),
    width = 0.18
  ) +
  coord_flip() +
  facet_wrap(~ district, scales = "free_y") +
  scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, 25)) +
  labs(
    title = "Project-site CCRF domain scores",
    subtitle = "Error bars show approximate 95% CIs",
    x = "Project site",
    y = "Mean domain score, 0–100",
    fill = "Domain"
  ) +
  theme(legend.position = "bottom")

Project-site indicator scores

project_site_indicators %>%
  filter(direction == "positive_core") %>%
  mutate(across(c(pct, se, ci_low, ci_high), ~round(.x, 1))) %>%
  select(project_site, district, domain, label, n_hhs, n_valid, pct, se, ci_low, ci_high) %>%
  arrange(district, project_site, domain, label) %>%
  kable(caption = "Positive indicator scores by project site")

Positive indicator scores by project site

project_site	district	domain	label	n_hhs	n_valid	pct	se	ci_low	ci_high
Ilha Insular	Ilha de Mocambique	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	1124	1124	67.2	1.4	64.4	69.9
Ilha Insular	Ilha de Mocambique	Capacity for Collective Action	Empowerment & Participation in Management	1124	572	93.2	1.1	91.1	95.2
Ilha Insular	Ilha de Mocambique	Capacity for Collective Action	Social Equity in Fisheries Benefits	1124	1071	83.3	1.1	81.1	85.5
Ilha Insular	Ilha de Mocambique	Capacity for Collective Action	Trust in Local Leadership	1124	1124	81.4	1.2	79.1	83.7
Ilha Insular	Ilha de Mocambique	Sustainable Livelihoods	Household covers needs	1124	1121	26.9	1.3	24.3	29.5
Ilha Insular	Ilha de Mocambique	Sustainable Livelihoods	Never worried about enough food	1124	1123	3.7	0.6	2.6	4.7
Quissanga	Ilha de Mocambique	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	365	364	57.1	2.6	52.1	62.2
Quissanga	Ilha de Mocambique	Capacity for Collective Action	Empowerment & Participation in Management	365	324	54.9	2.8	49.5	60.4
Quissanga	Ilha de Mocambique	Capacity for Collective Action	Social Equity in Fisheries Benefits	365	362	95.0	1.1	92.8	97.3
Quissanga	Ilha de Mocambique	Capacity for Collective Action	Trust in Local Leadership	365	364	60.4	2.6	55.4	65.5
Quissanga	Ilha de Mocambique	Sustainable Livelihoods	Household covers needs	365	363	19.3	2.1	15.2	23.3
Quissanga	Ilha de Mocambique	Sustainable Livelihoods	Never worried about enough food	365	364	2.7	0.9	1.1	4.4
Sanculo	Ilha de Mocambique	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	430	429	85.5	1.7	82.2	88.9
Sanculo	Ilha de Mocambique	Capacity for Collective Action	Empowerment & Participation in Management	430	395	94.4	1.2	92.2	96.7
Sanculo	Ilha de Mocambique	Capacity for Collective Action	Social Equity in Fisheries Benefits	430	428	81.1	1.9	77.4	84.8
Sanculo	Ilha de Mocambique	Capacity for Collective Action	Trust in Local Leadership	430	429	95.8	1.0	93.9	97.7
Sanculo	Ilha de Mocambique	Sustainable Livelihoods	Household covers needs	430	424	7.8	1.3	5.2	10.3
Sanculo	Ilha de Mocambique	Sustainable Livelihoods	Never worried about enough food	430	428	1.9	0.7	0.6	3.2
Baixo Pinda	Memba	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	382	381	73.0	2.3	68.5	77.4
Baixo Pinda	Memba	Capacity for Collective Action	Empowerment & Participation in Management	382	350	74.9	2.3	70.3	79.4
Baixo Pinda	Memba	Capacity for Collective Action	Social Equity in Fisheries Benefits	382	379	80.7	2.0	76.8	84.7
Baixo Pinda	Memba	Capacity for Collective Action	Trust in Local Leadership	382	382	45.8	2.6	40.8	50.8
Baixo Pinda	Memba	Sustainable Livelihoods	Household covers needs	382	380	4.5	1.1	2.4	6.6
Baixo Pinda	Memba	Sustainable Livelihoods	Never worried about enough food	382	381	0.5	0.4	0.0	1.3
Memba Sede	Memba	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	559	551	78.0	1.8	74.6	81.5
Memba Sede	Memba	Capacity for Collective Action	Empowerment & Participation in Management	559	258	84.1	2.3	79.6	88.6
Memba Sede	Memba	Capacity for Collective Action	Social Equity in Fisheries Benefits	559	529	74.9	1.9	71.2	78.6
Memba Sede	Memba	Capacity for Collective Action	Trust in Local Leadership	559	554	74.0	1.9	70.4	77.7
Memba Sede	Memba	Sustainable Livelihoods	Household covers needs	559	547	42.0	2.1	37.9	46.2
Memba Sede	Memba	Sustainable Livelihoods	Never worried about enough food	559	551	24.1	1.8	20.6	27.7
Meculuvelane	Mogincual	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	22	22	72.7	9.7	53.7	91.8
Meculuvelane	Mogincual	Capacity for Collective Action	Empowerment & Participation in Management	22	22	95.5	4.5	86.5	100.0
Meculuvelane	Mogincual	Capacity for Collective Action	Social Equity in Fisheries Benefits	22	22	100.0	0.0	100.0	100.0
Meculuvelane	Mogincual	Capacity for Collective Action	Trust in Local Leadership	22	22	95.5	4.5	86.5	100.0
Meculuvelane	Mogincual	Sustainable Livelihoods	Household covers needs	22	22	40.9	10.7	19.9	61.9
Meculuvelane	Mogincual	Sustainable Livelihoods	Never worried about enough food	22	22	0.0	0.0	0.0	0.0
Namalungo	Mogincual	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	263	263	31.6	2.9	25.9	37.2
Namalungo	Mogincual	Capacity for Collective Action	Empowerment & Participation in Management	263	239	95.0	1.4	92.2	97.8
Namalungo	Mogincual	Capacity for Collective Action	Social Equity in Fisheries Benefits	263	263	95.4	1.3	92.9	98.0
Namalungo	Mogincual	Capacity for Collective Action	Trust in Local Leadership	263	263	64.3	3.0	58.5	70.1
Namalungo	Mogincual	Sustainable Livelihoods	Household covers needs	263	263	26.6	2.7	21.3	32.0
Namalungo	Mogincual	Sustainable Livelihoods	Never worried about enough food	263	263	0.0	0.0	0.0	0.0
Namige Sede	Mogincual	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	308	308	80.8	2.2	76.4	85.2
Namige Sede	Mogincual	Capacity for Collective Action	Empowerment & Participation in Management	308	195	91.3	2.0	87.3	95.3
Namige Sede	Mogincual	Capacity for Collective Action	Social Equity in Fisheries Benefits	308	307	90.9	1.6	87.7	94.1
Namige Sede	Mogincual	Capacity for Collective Action	Trust in Local Leadership	308	308	90.3	1.7	86.9	93.6
Namige Sede	Mogincual	Sustainable Livelihoods	Household covers needs	308	308	48.4	2.9	42.8	54.0
Namige Sede	Mogincual	Sustainable Livelihoods	Never worried about enough food	308	308	37.3	2.8	31.9	42.7
Mahelene	Nacala Porto	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	244	244	74.6	2.8	69.1	80.1
Mahelene	Nacala Porto	Capacity for Collective Action	Empowerment & Participation in Management	244	163	89.0	2.5	84.1	93.8
Mahelene	Nacala Porto	Capacity for Collective Action	Social Equity in Fisheries Benefits	244	230	68.7	3.1	62.7	74.7
Mahelene	Nacala Porto	Capacity for Collective Action	Trust in Local Leadership	244	244	85.2	2.3	80.8	89.7
Mahelene	Nacala Porto	Sustainable Livelihoods	Household covers needs	244	244	36.9	3.1	30.8	43.0
Mahelene	Nacala Porto	Sustainable Livelihoods	Never worried about enough food	244	244	13.9	2.2	9.6	18.3
Quissimajulo	Nacala Porto	Capacity for Collective Action	Collective Efficacy for Fisheries Management Score	197	197	64.5	3.4	57.8	71.2
Quissimajulo	Nacala Porto	Capacity for Collective Action	Empowerment & Participation in Management	197	117	89.7	2.8	84.2	95.3
Quissimajulo	Nacala Porto	Capacity for Collective Action	Social Equity in Fisheries Benefits	197	186	63.4	3.5	56.5	70.4
Quissimajulo	Nacala Porto	Capacity for Collective Action	Trust in Local Leadership	197	197	88.8	2.2	84.4	93.2
Quissimajulo	Nacala Porto	Sustainable Livelihoods	Household covers needs	197	197	31.0	3.3	24.5	37.4
Quissimajulo	Nacala Porto	Sustainable Livelihoods	Never worried about enough food	197	197	21.8	3.0	16.0	27.6

project_site_indicators_clean <- project_site_indicators %>%
  filter(direction == "positive_core") %>%
  mutate(
    label = stringr::str_wrap(label, width = 28),
    project_site = stringr::str_wrap(project_site, width = 16)
  )

project_site_indicators_clean <- project_site_indicators %>%
  filter(direction == "positive_core") %>%
  mutate(
    label = stringr::str_wrap(label, width = 28),
    project_site = stringr::str_wrap(project_site, width = 16)
  )

# Define site order
site_levels <- project_site_indicators_clean %>%
  distinct(project_site) %>%
  arrange(project_site) %>%
  mutate(site_id = row_number())

project_site_indicators_clean <- project_site_indicators_clean %>%
  left_join(site_levels, by = "project_site")

# Alternating background bands
site_bands <- site_levels %>%
  mutate(
    xmin = site_id - 0.5,
    xmax = site_id + 0.5,
    shade = site_id %% 2 == 0
  ) %>%
  filter(shade)

ggplot(
  project_site_indicators_clean,
  aes(
    x = site_id,
    y = pct,
    color = label,
    group = label
  )
) +
  geom_rect(
    data = site_bands,
    aes(
      xmin = xmin,
      xmax = xmax,
      ymin = -Inf,
      ymax = Inf
    ),
    inherit.aes = FALSE,
    fill = "grey95",
    color = NA
  ) +
  geom_errorbar(
    aes(ymin = ci_low, ymax = ci_high),
    position = position_dodge(width = 0.55),
    width = 0.25,
    linewidth = 0.5,
    alpha = 0.6
  ) +
  geom_point(
    position = position_dodge(width = 0.55),
    size = 2.4
  ) +
  coord_flip() +
  facet_wrap(
    ~ domain,
    ncol = 1
  ) +
  scale_x_continuous(
    breaks = site_levels$site_id,
    labels = site_levels$project_site,
    expand = expansion(add = 0.3)
  ) +
  scale_y_continuous(
    limits = c(0, 100),
    breaks = seq(0, 100, 25),
    labels = scales::label_percent(scale = 1)
  ) +
  scale_color_brewer(palette = "Dark2") +
  labs(
    title = "Positive indicator scores by project site",
    subtitle = "Points show mean indicator scores; error bars show approximate 95% CIs",
    x = NULL,
    y = "% positive",
    color = "Indicator"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    legend.position = "bottom",
    legend.title = element_text(size = 10),
    legend.text = element_text(size = 9),
    strip.text = element_text(face = "bold"),
    panel.grid.minor = element_blank(),
    panel.grid.major.y = element_blank()
  )

Project-site domain scores by year

project_site_year_indicators <- hhs_project %>%
  summarise_indicators(group_vars = c("project_site", "district", "program_maturity", "year"))

project_site_year_domains <- project_site_year_indicators %>%
  make_domain_scores(group_vars = c("project_site", "district", "program_maturity", "year"))

project_site_year_domains %>%
  mutate(across(c(domain_score, domain_se, ci_low, ci_high), ~round(.x, 1))) %>%
  arrange(district, project_site, year, domain) %>%
  kable(caption = "Project-site domain scores by year")

Project-site domain scores by year

project_site	district	program_maturity	year	domain	n_hhs	domain_score	domain_se	ci_low	ci_high	n_indicators_available	min_n_valid
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2019	Capacity for Collective Action	327	75.5	1.3	72.9	78.1	4	139
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2019	Sustainable Livelihoods	327	25.7	1.6	22.5	28.8	2	325
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2021	Capacity for Collective Action	125	95.3	1.1	93.1	97.4	4	65
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2021	Sustainable Livelihoods	125	8.9	1.8	5.4	12.4	2	124
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2023	Capacity for Collective Action	106	98.8	0.5	97.8	99.9	4	55
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2023	Sustainable Livelihoods	106	3.3	1.2	0.9	5.7	2	106
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2025	Capacity for Collective Action	566	77.8	0.8	76.2	79.5	4	313
Ilha Insular	Ilha de Mocambique	Former/older Rare site	2025	Sustainable Livelihoods	566	13.0	0.9	11.2	14.8	2	566
Quissanga	Ilha de Mocambique	Former/older Rare site	2021	Capacity for Collective Action	91	89.3	1.7	85.9	92.7	4	57
Quissanga	Ilha de Mocambique	Former/older Rare site	2021	Sustainable Livelihoods	91	14.6	2.6	9.5	19.7	2	89
Quissanga	Ilha de Mocambique	Former/older Rare site	2023	Capacity for Collective Action	103	98.3	0.6	97.1	99.5	4	102
Quissanga	Ilha de Mocambique	Former/older Rare site	2023	Sustainable Livelihoods	103	24.8	2.5	19.9	29.6	2	103
Quissanga	Ilha de Mocambique	Former/older Rare site	2025	Capacity for Collective Action	171	37.6	1.4	34.8	40.4	4	165
Quissanga	Ilha de Mocambique	Former/older Rare site	2025	Sustainable Livelihoods	171	0.9	0.5	0.0	1.9	2	171
Sanculo	Ilha de Mocambique	Former/older Rare site	2021	Capacity for Collective Action	106	94.0	1.2	91.6	96.3	4	90
Sanculo	Ilha de Mocambique	Former/older Rare site	2021	Sustainable Livelihoods	106	10.4	2.1	6.2	14.6	2	100
Sanculo	Ilha de Mocambique	Former/older Rare site	2023	Capacity for Collective Action	104	90.4	1.3	87.8	93.0	4	104
Sanculo	Ilha de Mocambique	Former/older Rare site	2023	Sustainable Livelihoods	104	1.0	0.7	0.0	2.3	2	104
Sanculo	Ilha de Mocambique	Former/older Rare site	2025	Capacity for Collective Action	220	86.3	1.1	84.2	88.5	4	201
Sanculo	Ilha de Mocambique	Former/older Rare site	2025	Sustainable Livelihoods	220	4.1	0.9	2.3	5.9	2	220
Baixo Pinda	Memba	Former/older Rare site	2019	Capacity for Collective Action	1	50.0	NA	NA	NA	4	1
Baixo Pinda	Memba	Former/older Rare site	2019	Sustainable Livelihoods	1	0.0	NA	NA	NA	2	1
Baixo Pinda	Memba	Former/older Rare site	2021	Capacity for Collective Action	101	52.1	1.6	49.0	55.2	4	96
Baixo Pinda	Memba	Former/older Rare site	2021	Sustainable Livelihoods	101	0.0	0.0	0.0	0.0	2	99
Baixo Pinda	Memba	Former/older Rare site	2024	Capacity for Collective Action	145	96.9	0.7	95.5	98.3	4	143
Baixo Pinda	Memba	Former/older Rare site	2024	Sustainable Livelihoods	145	1.0	0.6	0.0	2.2	2	145
Baixo Pinda	Memba	Former/older Rare site	2026	Capacity for Collective Action	135	51.4	1.7	48.0	54.8	4	110
Baixo Pinda	Memba	Former/older Rare site	2026	Sustainable Livelihoods	135	5.9	1.4	3.2	8.7	2	135
Memba Sede	Memba	Former/older Rare site	2019	Capacity for Collective Action	206	61.9	2.0	57.8	65.9	4	77
Memba Sede	Memba	Former/older Rare site	2019	Sustainable Livelihoods	206	19.2	1.8	15.6	22.8	2	200
Memba Sede	Memba	Former/older Rare site	2021	Capacity for Collective Action	214	95.1	0.8	93.5	96.7	4	93
Memba Sede	Memba	Former/older Rare site	2021	Sustainable Livelihoods	214	55.1	2.4	50.4	59.9	2	207
Memba Sede	Memba	Former/older Rare site	2026	Capacity for Collective Action	139	70.0	1.8	66.5	73.4	4	88
Memba Sede	Memba	Former/older Rare site	2026	Sustainable Livelihoods	139	19.8	2.2	15.5	24.1	2	139
Meculuvelane	Mogincual	Newer expansion site	2025	Capacity for Collective Action	22	90.9	2.9	85.2	96.6	4	22
Meculuvelane	Mogincual	Newer expansion site	2025	Sustainable Livelihoods	22	20.5	5.4	9.9	31.0	2	22
Namalungo	Mogincual	Newer expansion site	2024	Capacity for Collective Action	134	64.7	1.2	62.4	67.1	4	134
Namalungo	Mogincual	Newer expansion site	2024	Sustainable Livelihoods	134	6.3	1.4	3.5	9.2	2	134
Namalungo	Mogincual	Newer expansion site	2025	Capacity for Collective Action	129	78.3	1.8	74.9	81.8	4	105
Namalungo	Mogincual	Newer expansion site	2025	Sustainable Livelihoods	129	20.5	2.2	16.3	24.8	2	129
Namige Sede	Mogincual	Newer expansion site	2024	Capacity for Collective Action	151	93.5	1.2	91.1	95.8	4	68
Namige Sede	Mogincual	Newer expansion site	2024	Sustainable Livelihoods	151	62.6	2.7	57.3	67.9	2	151
Namige Sede	Mogincual	Newer expansion site	2025	Capacity for Collective Action	157	83.2	1.5	80.3	86.1	4	127
Namige Sede	Mogincual	Newer expansion site	2025	Sustainable Livelihoods	157	23.9	2.0	19.9	27.9	2	157
Mahelene	Nacala Porto	Newer expansion site	2024	Capacity for Collective Action	134	91.9	1.1	89.7	94.1	4	96
Mahelene	Nacala Porto	Newer expansion site	2024	Sustainable Livelihoods	134	19.0	2.4	14.3	23.7	2	134
Mahelene	Nacala Porto	Newer expansion site	2026	Capacity for Collective Action	110	63.2	2.5	58.4	68.1	4	67
Mahelene	Nacala Porto	Newer expansion site	2026	Sustainable Livelihoods	110	33.2	2.7	27.9	38.5	2	110
Quissimajulo	Nacala Porto	Newer expansion site	2024	Capacity for Collective Action	126	83.9	1.6	80.9	87.0	4	86
Quissimajulo	Nacala Porto	Newer expansion site	2024	Sustainable Livelihoods	126	19.4	2.5	14.5	24.3	2	126
Quissimajulo	Nacala Porto	Newer expansion site	2026	Capacity for Collective Action	71	61.0	3.3	54.5	67.5	4	31
Quissimajulo	Nacala Porto	Newer expansion site	2026	Sustainable Livelihoods	71	38.7	3.9	31.1	46.3	2	71

ggplot(project_site_year_domains,
       aes(x = year, y = domain_score, group = domain, linetype = domain)) +
  geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0.10, alpha = 0.55) +
  geom_line(linewidth = 0.8) +
  geom_point(size = 2.1) +
  facet_wrap(~ project_site, ncol = 2) +
  scale_y_continuous(limits = c(0, 100), breaks = seq(0, 100, 25)) +
  scale_x_continuous(breaks = sort(unique(project_site_year_domains$year))) +
  labs(
    title = "Project-site domain scores by year",
    subtitle = "Use cautiously: sample sizes and implementation context vary across site-years; error bars show approximate 95% CIs",
    x = "Survey year",
    y = "Mean domain score, 0–100",
    linetype = "Domain"
  ) +
  theme(
    legend.position = "bottom",
    strip.text = element_text(face = "bold"),
    axis.text.x = element_text(angle = 45, hjust = 1)
  )

Project-site domain scores by year: heatmap version

ggplot(project_site_year_domains,
       aes(x = factor(year), y = fct_reorder(project_site, domain_score, .fun = mean_na), fill = domain_score)) +
  geom_tile(color = "white") +
  geom_text(aes(label = if_else(is.na(domain_score), "", round(domain_score, 0) %>% as.character())), size = 3) +
  facet_wrap(~ domain, ncol = 2) +
  scale_fill_gradient(low = "grey95", high = "grey20", limits = c(0, 100), na.value = "white") +
  labs(
    title = "Project-site domain scores by year",
    subtitle = "Numbers are mean scores on a 0–100 scale",
    x = "Survey year",
    y = "Project site",
    fill = "Score\n0–100"
  )

8. Sustainable Livelihoods (negative)

These plots show the negative livelihood conditions directly.

project_livelihood_diagnostics <- project_site_indicators %>%
  filter(indicator %in% c("financial_strain", "food_worry_any", "food_worry_often"))

project_livelihood_diagnostics %>%
  mutate(across(c(pct, se, ci_low, ci_high), ~round(.x, 1))) %>%
  select(project_site, district, label, n_hhs, n_valid, pct, se, ci_low, ci_high) %>%
  arrange(district, project_site, label) %>%
  kable(caption = "Livelihood diagnostic indicators by project site")

Livelihood diagnostic indicators by project site

project_site	district	label	n_hhs	n_valid	pct	se	ci_low	ci_high
Ilha Insular	Ilha de Mocambique	Household covers needs with difficulty	1124	1121	73.1	1.3	70.5	75.7
Ilha Insular	Ilha de Mocambique	Often worried about food	1124	1123	36.1	1.4	33.3	38.9
Ilha Insular	Ilha de Mocambique	Sometimes/often worried about food	1124	1123	96.3	0.6	95.3	97.4
Quissanga	Ilha de Mocambique	Household covers needs with difficulty	365	363	80.7	2.1	76.7	84.8
Quissanga	Ilha de Mocambique	Often worried about food	365	364	46.4	2.6	41.3	51.6
Quissanga	Ilha de Mocambique	Sometimes/often worried about food	365	364	97.3	0.9	95.6	98.9
Sanculo	Ilha de Mocambique	Household covers needs with difficulty	430	424	92.2	1.3	89.7	94.8
Sanculo	Ilha de Mocambique	Often worried about food	430	428	40.7	2.4	36.0	45.3
Sanculo	Ilha de Mocambique	Sometimes/often worried about food	430	428	98.1	0.7	96.8	99.4
Baixo Pinda	Memba	Household covers needs with difficulty	382	380	95.5	1.1	93.4	97.6
Baixo Pinda	Memba	Often worried about food	382	381	24.9	2.2	20.6	29.3
Baixo Pinda	Memba	Sometimes/often worried about food	382	381	99.5	0.4	98.7	100.0
Memba Sede	Memba	Household covers needs with difficulty	559	547	58.0	2.1	53.8	62.1
Memba Sede	Memba	Often worried about food	559	551	4.9	0.9	3.1	6.7
Memba Sede	Memba	Sometimes/often worried about food	559	551	75.9	1.8	72.3	79.4
Meculuvelane	Mogincual	Household covers needs with difficulty	22	22	59.1	10.7	38.1	80.1
Meculuvelane	Mogincual	Often worried about food	22	22	4.5	4.5	0.0	13.5
Meculuvelane	Mogincual	Sometimes/often worried about food	22	22	100.0	0.0	100.0	100.0
Namalungo	Mogincual	Household covers needs with difficulty	263	263	73.4	2.7	68.0	78.7
Namalungo	Mogincual	Often worried about food	263	263	14.8	2.2	10.5	19.1
Namalungo	Mogincual	Sometimes/often worried about food	263	263	100.0	0.0	100.0	100.0
Namige Sede	Mogincual	Household covers needs with difficulty	308	308	51.6	2.9	46.0	57.2
Namige Sede	Mogincual	Often worried about food	308	308	6.5	1.4	3.7	9.2
Namige Sede	Mogincual	Sometimes/often worried about food	308	308	62.7	2.8	57.3	68.1
Mahelene	Nacala Porto	Household covers needs with difficulty	244	244	63.1	3.1	57.0	69.2
Mahelene	Nacala Porto	Often worried about food	244	244	34.4	3.0	28.5	40.4
Mahelene	Nacala Porto	Sometimes/often worried about food	244	244	86.1	2.2	81.7	90.4
Quissimajulo	Nacala Porto	Household covers needs with difficulty	197	197	69.0	3.3	62.6	75.5
Quissimajulo	Nacala Porto	Often worried about food	197	197	36.0	3.4	29.3	42.8
Quissimajulo	Nacala Porto	Sometimes/often worried about food	197	197	78.2	3.0	72.4	84.0

ggplot(project_livelihood_diagnostics,
       aes(x = fct_reorder(project_site, pct), y = pct, fill = label)) +
  geom_col(position = position_dodge(width = 0.75), width = 0.7) +
  geom_errorbar(
    aes(ymin = ci_low, ymax = ci_high),
    position = position_dodge(width = 0.75),
    width = 0.18
  ) +
  coord_flip() +
  facet_wrap(~ district, scales = "free_y") +
  scale_y_continuous(limits = c(0, 100), labels = label_percent(scale = 1)) +
  labs(
    title = "Sustainable Livelihoods diagnostic indicators by project site",
    subtitle = "Higher values here indicate more livelihood stress; error bars show approximate 95% CIs",
    x = "Project site",
    y = "% of valid responses",
    fill = "Diagnostic indicator"
  ) +
  theme(legend.position = "bottom")

9. Income context

Income is useful context but still requires more cleaning and checks for outliers, currency consistency, recall period, and inflation. Prefer medians and IQRs.

income_summary <- hhs_if %>%
  summarise(
    n_valid_income = sum(!is.na(g13_hh_average_income)),
    min = min(g13_hh_average_income, na.rm = TRUE),
    p25 = p25_na(g13_hh_average_income),
    median = median_na(g13_hh_average_income),
    mean = mean_na(g13_hh_average_income),
    p75 = p75_na(g13_hh_average_income),
    p95 = as.numeric(quantile(g13_hh_average_income, 0.95, na.rm = TRUE)),
    max = max(g13_hh_average_income, na.rm = TRUE)
  )

income_summary %>%
  mutate(across(where(is.numeric), ~round(.x, 1))) %>%
  kable(caption = "Raw average monthly household income summary")

Raw average monthly household income summary

n_valid_income	min	p25	median	mean	p75	p95	max
6877	0	2000	5000	9420.7	9800	25000	9e+05

# Clean income variable for plotting only --------------------------------

income_plot_data <- hhs_if %>%
  mutate(
    income_raw = g13_hh_average_income,
    income_mzn = readr::parse_number(as.character(g13_hh_average_income))
  ) %>%
  filter(
    !is.na(income_mzn),
    income_mzn >= 0
  )

# Use the 99th percentile as a conservative plotting cutoff.
# This removes extreme values that dominate the plot but keeps most observations.
income_cutoff_99 <- quantile(
  income_plot_data$income_mzn,
  probs = 0.99,
  na.rm = TRUE
)

income_plot_data_clean <- income_plot_data %>%
  filter(income_mzn <= income_cutoff_99)

income_outlier_summary <- tibble::tibble(
  records_with_income = nrow(income_plot_data),
  records_kept_for_plots = nrow(income_plot_data_clean),
  records_removed_from_income_plots = nrow(income_plot_data) - nrow(income_plot_data_clean),
  income_cutoff_99 = income_cutoff_99
)

income_outlier_summary %>%
  knitr::kable(
    caption = "Income values excluded from income plots using the 99th percentile cutoff"
  )

Income values excluded from income plots using the 99th percentile cutoff

records_with_income	records_kept_for_plots	records_removed_from_income_plots	income_cutoff_99
6877	6811	66	75000

ggplot(
  income_plot_data_clean,
  aes(x = income_mzn)
) +
  geom_histogram(
    bins = 40,
    fill = "grey40",
    color = "white"
  ) +
  scale_x_continuous(
    labels = scales::comma
  ) +
  labs(
    title = "Reported average monthly household income",
    subtitle = paste0(
      "Values above the 99th percentile excluded for readability; cutoff = ",
      scales::comma(round(income_cutoff_99, 0))
    ),
    x = "Average monthly household income",
    y = "HHS records"
  ) +
  theme_minimal(base_size = 12)

project_income <- hhs_project %>%
  group_by(project_site, district) %>%
  summarise(
    n_valid_income = sum(!is.na(g13_hh_average_income)),
    median_income = median_na(g13_hh_average_income),
    income_p25 = p25_na(g13_hh_average_income),
    income_p75 = p75_na(g13_hh_average_income),
    .groups = "drop"
  )

project_income %>%
  mutate(across(where(is.numeric), ~round(.x, 1))) %>%
  kable(caption = "Income context by project site")

Income context by project site

project_site	district	n_valid_income	median_income	income_p25	income_p75
Baixo Pinda	Memba	380	1500	800	4850
Ilha Insular	Ilha de Mocambique	1117	8000	5000	12000
Mahelene	Nacala Porto	244	7000	3000	13000
Meculuvelane	Mogincual	22	2820	2040	3550
Memba Sede	Memba	482	500	200	2000
Namalungo	Mogincual	263	6000	3800	7000
Namige Sede	Mogincual	308	10000	4475	36250
Quissanga	Ilha de Mocambique	359	2500	1900	8825
Quissimajulo	Nacala Porto	197	5000	2500	8007
Sanculo	Ilha de Mocambique	418	6000	3800	14475

income_plot_data_clean %>%
  filter(!is.na(project_site)) %>%
  ggplot(
    aes(
      x = reorder(project_site, income_mzn, median, na.rm = TRUE),
      y = income_mzn
    )
  ) +
  geom_boxplot(outlier.alpha = 0.25) +
  coord_flip() +
  scale_y_continuous(
    labels = scales::comma
  ) +
  labs(
    title = "Reported average monthly household income by project site",
    subtitle = paste0(
      "Values above the 99th percentile excluded for readability; cutoff = ",
      scales::comma(round(income_cutoff_99, 0))
    ),
    x = NULL,
    y = "Average monthly household income"
  ) +
  theme_minimal(base_size = 12)

income_plot_data_clean %>%
  filter(!is.na(year)) %>%
  ggplot(
    aes(
      x = factor(year),
      y = income_mzn
    )
  ) +
  geom_boxplot(outlier.alpha = 0.25) +
  scale_y_continuous(
    labels = scales::comma
  ) +
  labs(
    title = "Reported average monthly household income by year",
    subtitle = paste0(
      "Values above the 99th percentile excluded for readability; cutoff = ",
      scales::comma(round(income_cutoff_99, 0))
    ),
    x = "Survey year",
    y = "Average monthly household income"
  ) +
  theme_minimal(base_size = 12)

project_site_income_year <- income_plot_data_clean %>%
  filter(
    !is.na(project_site),
    !is.na(year),
    !is.na(income_mzn)
  ) %>%
  group_by(project_site, year) %>%
  summarise(
    n = n(),
    median_income = median(income_mzn, na.rm = TRUE),
    mean_income = mean(income_mzn, na.rm = TRUE),
    q25 = quantile(income_mzn, 0.25, na.rm = TRUE),
    q75 = quantile(income_mzn, 0.75, na.rm = TRUE),
    .groups = "drop"
  )

project_site_income_year %>%
  mutate(
    median_income = round(median_income, 0),
    mean_income = round(mean_income, 0),
    q25 = round(q25, 0),
    q75 = round(q75, 0)
  ) %>%
  arrange(project_site, year) %>%
  knitr::kable(
    caption = "Reported average monthly household income by project site and year"
  )

Reported average monthly household income by project site and year

project_site	year	n	median_income	mean_income	q25	q75
Baixo Pinda	2019	1	5000	5000	5000	5000
Baixo Pinda	2021	99	1000	994	700	1300
Baixo Pinda	2024	145	5000	4163	3000	6000
Baixo Pinda	2026	135	1500	1972	650	2450
Ilha Insular	2019	325	5500	7105	3500	9000
Ilha Insular	2021	120	5000	4966	3000	6000
Ilha Insular	2023	104	9900	17176	8225	24300
Ilha Insular	2025	557	9500	11506	6500	13000
Mahelene	2024	134	4500	6451	1500	8450
Mahelene	2026	110	10125	13330	6000	18000
Meculuvelane	2025	22	2820	2967	2040	3550
Memba Sede	2019	153	1200	2210	200	3500
Memba Sede	2021	190	400	566	200	600
Memba Sede	2026	139	1200	2322	400	2700
Namalungo	2024	133	6000	8130	6000	8000
Namalungo	2025	129	3900	5375	2000	6000
Namige Sede	2024	109	25000	29954	15000	40000
Namige Sede	2025	157	4600	5471	3000	6300
Quissanga	2021	85	3000	3331	1000	5000
Quissanga	2023	102	14000	13578	8562	18000
Quissanga	2025	171	2000	3206	1500	2500
Quissimajulo	2024	126	4000	4739	2000	6950
Quissimajulo	2026	71	8000	11642	5250	12500
Sanculo	2021	94	3500	3471	1850	5000
Sanculo	2023	102	16305	18022	10000	21450
Sanculo	2025	213	5200	9589	4000	10000

income_plot_data_clean %>%
  filter(
    !is.na(project_site),
    !is.na(year),
    !is.na(income_mzn)
  ) %>%
  mutate(
    project_site = stringr::str_wrap(project_site, width = 18)
  ) %>%
  ggplot(
    aes(
      x = factor(year),
      y = income_mzn
    )
  ) +
  geom_boxplot(
    outlier.alpha = 0.15,
    width = 0.65
  ) +
  facet_wrap(~ project_site, ncol = 3, scales = "free_y") +
  scale_y_continuous(labels = scales::comma) +
  labs(
    title = "Reported average monthly household income by project site and year",
    subtitle = paste0(
      "Values above the 99th percentile excluded for readability; cutoff = ",
      scales::comma(round(income_cutoff_99, 0))
    ),
    x = "Survey year",
    y = "Average monthly household income"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    strip.text = element_text(face = "bold"),
    panel.grid.minor = element_blank()
  )

bootstrap_median_ci <- function(x, n_boot = 1000, conf = 0.95) {
  
  x <- x[!is.na(x)]
  
  if (length(x) < 5) {
    return(
      tibble::tibble(
        median_income = median(x, na.rm = TRUE),
        ci_low = NA_real_,
        ci_high = NA_real_
      )
    )
  }
  
  boot_medians <- replicate(
    n_boot,
    median(sample(x, size = length(x), replace = TRUE), na.rm = TRUE)
  )
  
  alpha <- 1 - conf
  
  tibble::tibble(
    median_income = median(x, na.rm = TRUE),
    ci_low = quantile(boot_medians, probs = alpha / 2, na.rm = TRUE),
    ci_high = quantile(boot_medians, probs = 1 - alpha / 2, na.rm = TRUE)
  )
}

# Project-site income summary by year ------------------------------------

project_site_income_year <- income_plot_data_clean %>%
  filter(
    !is.na(project_site),
    !is.na(year),
    !is.na(income_mzn)
  ) %>%
  group_by(project_site, year) %>%
  summarise(
    n = n(),
    mean_income = mean(income_mzn, na.rm = TRUE),
    q25 = quantile(income_mzn, 0.25, na.rm = TRUE),
    q75 = quantile(income_mzn, 0.75, na.rm = TRUE),
    median_ci = list(bootstrap_median_ci(income_mzn)),
    .groups = "drop"
  ) %>%
  tidyr::unnest(median_ci)


ggplot(
  project_site_income_year,
  aes(
    x = year,
    y = median_income,
    group = project_site,
    color = project_site
  )
) +
  geom_errorbar(
    aes(
      ymin = ci_low,
      ymax = ci_high
    ),
    width = 0.12,
    alpha = 0.45,
    linewidth = 0.5
  ) +
  geom_line(linewidth = 0.8) +
  geom_point(aes(size = n), alpha = 0.85) +
  scale_y_continuous(labels = scales::comma) +
  scale_x_continuous(
    breaks = sort(unique(project_site_income_year$year))
  ) +
  scale_color_brewer(palette = "Paired") +
  labs(
    title = "Median reported monthly household income by project site and year",
    subtitle = "Error bars show bootstrap 95% CIs for the median; point size reflects number of HHS records with non-missing income",
    x = "Survey year",
    y = "Median monthly household income",
    color = "Project site",
    size = "N"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    legend.position = "bottom",
    panel.grid.minor = element_blank()
  )

income_heatmap_data <- project_site_income_year %>%
  mutate(
    project_site = forcats::fct_reorder(
      project_site,
      median_income,
      .fun = median,
      na.rm = TRUE
    ),
    label = paste0(
      scales::comma(round(median_income, 0)),
      "\n",
      "n=",
      n
    ),
    income_scaled = scales::rescale(
      median_income,
      to = c(0, 1),
      from = range(median_income, na.rm = TRUE)
    ),
    label_color = if_else(income_scaled < 0.45, "white", "black")
  )

ggplot(
  income_heatmap_data,
  aes(
    x = factor(year),
    y = project_site,
    fill = median_income
  )
) +
  geom_tile(color = "white") +
  geom_text(
    aes(
      label = label,
      color = label_color
    ),
    size = 3
  ) +
  scale_color_identity() +
  scale_fill_viridis_c(
    labels = scales::comma,
    option = "C"
  ) +
  labs(
    title = "Median reported monthly household income by project site and year",
    subtitle = "Cell labels show median income and number of HHS records",
    x = "Survey year",
    y = NULL,
    fill = "Median income"
  ) +
  theme_minimal(base_size = 12) +
  theme(
    panel.grid = element_blank()
  )

10. Suggested interpretation for a CCRF case study

The Mozambique HHS provides a useful demonstration of the CCRF as a diagnostic framework. The likely core finding is an imbalance between the two measurable components:

• Capacity for Collective Action tends to score higher. This suggests that many respondents express positive views on social equity, trust in local decision-making, community management ability, and the value of participation in management.
• Sustainable Livelihoods tends to score lower. Many households report difficulty covering basic needs and concern about having enough food.
• The CCRF therefore helps identify Sustainable Livelihoods as a priority resilience weakness, even in places where collective-action foundations appear stronger.

Suggested manuscript language:

In the Mozambique case study, HHS results showed a marked imbalance across CCRF domains. Indicators linked to collective action and co-management were comparatively stronger, suggesting a foundation for local fisheries stewardship. In contrast, Sustainable Livelihoods indicators were consistently weaker: many households reported difficulty covering basic needs and concern about food availability. The CCRF therefore identified livelihoods resilience as a priority management gap, helping justify targeted responses such as fisheries-based microenterprises, Farmer Field Schools, TVET, poultry, aquaculture, apiculture, market linkages, savings groups, and nutrition/WASH interventions.