KHS 2023/24 · n = 21,347 households · All 47 counties
KHS 2023/24 · n = kaya 21,347 · Kaunti zote 47
Kenya Housing Vulnerability Atlas Ramani ya Udhaifu wa Makazi Kenya
Select a county to see its five-dimension vulnerability profile and get AI policy recommendations. Higher score = more vulnerable.
Chagua kaunti kuona profaili yake ya vipimo vitano na mapendekezo ya sera. Alama kubwa = udhaifu zaidi.
47
Counties mappedKaunti zimepigwa ramani
20.1%
Highly vulnerable (1 in 5, 80th pct)Udhaifu mkubwa (1 kati ya 5)
0.4797
National mean HFVS (corrected)Wastani wa HFVS (iliorekebishwa)
21,347
Households surveyedKaya zilizochunguzwa
0.65–0.78
Field proxy AUC (spatial CV)AUC ya mfano wa shamba
Sort worst:Panga:
D1 Financial
D2 Tenure
D3 Hazard
D4 Dwelling
D5 Utility
Click card to explore · Click 2 to compare
Bonyeza kadi kuangalia · Bonyeza 2 kulinganisha
National StatisticsTakwimu za Taifa
KHS 2023/24 · KNBS
21,347
Households surveyed across 47 counties
Kaya katika kaunti 47
10.09%
Above 90th percentile threshold — highly vulnerable (2,154 households)
Juu ya kiwango cha asilimia 90 — udhaifu mkubwa (kaya 2,154)
80th pct sensitivity band: 20.12% — about 1 in 5
Ukanda wa unyeti wa asilimia 80: 20.12%
1.81%
Triple-exposed (flood + no tenure + rent stress)
Mfiduo wa tatu
3.90%
Urban triple-exposure — 25.8× higher than rural
Mfiduo wa tatu mjini — mara 25.8 zaidi ya vijijini
96.8%
No formal loan access nationally
Hawana mkopo rasmi kitaifa
96.6%
Cook with solid fuel (driving D5)
Kupika kwa kuni au mkaa
43.7%
No grid electricity connection
Hawana umeme wa gridi
54.8%
Rent-stressed renters (>30% of expenditure)
Msongo wa kodi
54.5%
No land ownership nationally
Hawana umiliki wa ardhi
38.2%
Overcrowded households (>2 persons/room)
Msongamano wa kaya
Effective Dimension Weights — Why D5 Dominates
Uzito wa Vipimo — Kwa Nini D5 Inatawala
Weight Correction Note — Two Pipelines, Two Intentional Choices
Kumbuka ya Urekebishaji wa Uzito — Njia Mbili, Maamuzi Mawili ya Makusudi
Two pipeline choices produce two HFVS weight sets. The national headline (Notebook 1, variance-balanced z-score) raises D3 Physical Hazard to 21.5% and sets D5 Utility Deprivation at 24.3%. The county table uses the dissertation's original weighting (~6% D3) where hazard-proximity coding is sparse — a documented data limitation. The two sets are not directly comparable; see How It Works.
Uzito huu wa kitaifa unatoka kwa ujenzi upya wa z-score wa Notebook 1 unaorekebishwa. D3 inapanda hadi asilimia 21.5 na D5 inashuka hadi asilimia 24.3. Jedwali la kaunti linatumia uzito wa awali wa tasnifu (~asilimia 6 D3) — hii ni kumbukumbu ya upungufu wa data.
10 Most Vulnerable Counties
Kaunti 10 Zenye Udhaifu Zaidi
10 Least Vulnerable Counties
Kaunti 10 Zenye Udhaifu Kidogo
Who Needs to Act — and With Which Number
Nani Anapaswa Kutenda — na Takwimu Gani
For Everyone / Kwa Kila Mtu
Understanding Your County's Housing Vulnerability Kuelewa Udhaifu wa Makazi katika Kaunti Yako
No jargon. Plain language for every Kenyan — tenant, homeowner, student, farmer, or community leader.
Hakuna maneno magumu. Lugha ya kawaida kwa kila Mkenya.
What is the HFVS score?Ni nini Alama ya HFVS?
The HFVS combines five housing problems into one score from 0 to 1. Closer to 1 = more vulnerable.
HFVS inaunganisha matatizo matano katika nambari moja kati ya 0 na 1. Karibu na 1 = udhaifu zaidi.
The Five Dimensions MeasuredMatatizo Matano Yanayopimwa
D1
Financial StressMsongo wa Fedha
Paying too much income on rent, having no savings, and being unable to access any loan when you need one.Kulipa sehemu kubwa ya mapato kwa kodi, kutokuwa na akiba, na kushindwa kupata mkopo.
D2
Tenure InsecurityKukosa Haki ya Ardhi
Renting without a written lease, living on disputed land, or having no title deed. One argument and you could lose your home.Kupanga bila mkataba wa maandishi, kuishi kwenye ardhi inayoshindaniwa, au kukosa hati miliki.
D3
Physical HazardHatari ya Mazingira
Living in a flood zone or mudslide-prone hillside. One heavy rain season can destroy everything.Kuishi katika eneo la mafuriko au kilima kinachoweza kuteleza. Msimu mmoja wa mvua kubwa unaweza kuharibu kila kitu.
D4
Dwelling QualityUbora wa Makazi
Grass or leaking iron-sheet roofs, mud walls, earthen floors, and too many people in too small a space.Paa za nyasi au bati zinazovuja, kuta za udongo, sakafu ya udongo, na watu wengi katika nafasi ndogo.
D5
Utility DeprivationUkosefu wa Huduma
No electricity, cooking with firewood/charcoal, no clean water. This is the biggest national driver — affecting 96.6% of households.Hakuna umeme, kupika kwa kuni au mkaa. Hii ndiyo sababu kubwa zaidi nchini — inayoathiri asilimia 96.6.
How to Read Your County ScoreJinsi ya Kusoma Alama ya Kaunti Yako
1
Go to "Explore Counties" tabNenda kwenye "Tafuta Kaunti"
Type your county name in the search box. Tap the card to open its full profile.Andika jina la kaunti yako kwenye sanduku la utafutaji. Gonga kadi kufungua maelezo yake.
2
Find the longest bar — that's the biggest problemTafuta mstari mrefu zaidi — hiyo ndiyo tatizo kubwa zaidi
The "Worst" badge highlights your county's most critical dimension (D1–D5). The comparison bars show how you rank against the national average.Beji ya "Mbaya Zaidi" inaonyesha kipimo muhimu zaidi cha kaunti yako (D1–D5).
3
Ask the AI Policy AdvisorUliza Mshauri wa Sera wa AI
The AI Advisor at the bottom of every county profile gives specific recommendations in plain English or Swahili.Mshauri wa AI chini ya kila profaili ya kaunti anatoa mapendekezo maalum kwa Kiswahili au Kiingereza.
What You Can Do With This DataUnachoweza Kufanya na Taarifa Hii
If you are a voter or community member: ask your MCA or MP what they are doing about the dimension your county scores worst on. Data source: KNBS — official and verifiable.
Kama wewe ni mpiga kura au mwanajamii: uliza MCA wako au Mbunge wako wanachofanya kuhusu kipimo ambacho kaunti yako ina alama mbaya zaidi.
If you are a county official or MCA: the dimension breakdown maps to budget lines. D5 = electrification and clean cookstoves. D2 = land registration. D3 = drainage and flood works.
Kama wewe ni afisa wa kaunti au MCA: D5 = umeme na jiko safi. D2 = usajili wa ardhi. D3 = mifereji ya maji.
If you are a journalist or researcher: cite KNBS Kenya Housing Survey 2023/24 (n = 21,347). See How It Works for the full methodology.
Kama wewe ni mwandishi wa habari au mtafiti: HFVS inatokana na KNBS Kenya Housing Survey 2023/24 (n = 21,347).
The HFVS FormulaFomula ya HFVS
HFVS_i = (D1_i + D2_i + D3_i + D4_i + D5_i) / 5 [equal weights, n = 21,347]
Each dimension is normalised to [0,1] via z-score standardisation followed by min-max rescaling, then averaged equally. County estimates are survey-weighted using the KNBS sampling design (ratio 327.6×).
Kila kipimo kinasanifishwa hadi [0,1] kupitia usanifishaji wa z-score ikifuatiwa na upimaji wa min-max, kisha wastani sawa. Makadirio ya kaunti yana uzito wa utafiti wa KNBS (uwiano 327.6×).
Five Dimensions — Full SpecificationVipimo Vitano — Maelezo Kamili
D1 — Financial Stress (Effective weight: 23.7%)
Rent-to-expenditure burden (>30% threshold), absence of formal savings, no formal/informal loan access. Key stat: 96.8% have no formal loan instrument. Mzigo wa kodi kwa matumizi (>30%), kukosa akiba rasmi, hakuna mkopo. Takwimu: Asilimia 96.8 hawana chombo chochote cha mkopo rasmi.
Rent-to-expenditure burden (>30% threshold), absence of formal savings, no formal/informal loan access. Key stat: 96.8% have no formal loan instrument. Mzigo wa kodi kwa matumizi (>30%), kukosa akiba rasmi, hakuna mkopo. Takwimu: Asilimia 96.8 hawana chombo chochote cha mkopo rasmi.
D2 — Tenure Insecurity (Effective weight: 13.8%)
Land ownership status, written lease presence, active parcel disputes, informal settlement indicator. Key stat: 54.5% report no land ownership nationally. Hali ya umiliki wa ardhi, mkataba wa maandishi, migogoro ya viwanja. Takwimu: Asilimia 54.5 hawana umiliki wa ardhi.
Land ownership status, written lease presence, active parcel disputes, informal settlement indicator. Key stat: 54.5% report no land ownership nationally. Hali ya umiliki wa ardhi, mkataba wa maandishi, migogoro ya viwanja. Takwimu: Asilimia 54.5 hawana umiliki wa ardhi.
D3 — Physical Hazard (Effective weight: 6.2%)
Flood zone designation, mudslide zone exposure. Note: six proximity sub-columns absent from the 2023/24 extract suppressed D3. Key stats: 19.1% in flood zones; 12.6% in mudslide zones. Eneo la mafuriko, maporomoko ya udongo. Kumbuka: vipande sita vya ukaribu havikuwepo, kupunguza uzito wa D3. Takwimu: 19.1% katika maeneo ya mafuriko.
Flood zone designation, mudslide zone exposure. Note: six proximity sub-columns absent from the 2023/24 extract suppressed D3. Key stats: 19.1% in flood zones; 12.6% in mudslide zones. Eneo la mafuriko, maporomoko ya udongo. Kumbuka: vipande sita vya ukaribu havikuwepo, kupunguza uzito wa D3. Takwimu: 19.1% katika maeneo ya mafuriko.
D4 — Dwelling Quality (Effective weight: 22.0%)
Roof/wall/floor material durability, occupancy density (>2 persons/room). Key stat: 38.2% of households are overcrowded. Uimara wa nyenzo za paa/kuta/sakafu, msongamano wa kukaa. Takwimu: Asilimia 38.2 wamezidi watu.
Roof/wall/floor material durability, occupancy density (>2 persons/room). Key stat: 38.2% of households are overcrowded. Uimara wa nyenzo za paa/kuta/sakafu, msongamano wa kukaa. Takwimu: Asilimia 38.2 wamezidi watu.
D5 — Utility Deprivation (Effective weight: 34.3%)
Grid electricity, cooking fuel (solid = deprived), improved water source, improved sanitation. Dominates due to near-universal solid-fuel cooking (96.6%) and limited grid access (43.7% unconnected). Umeme wa gridi, mafuta ya kupikia, maji bora, usafi. Inatawala kutokana na upikaji wa kuni (96.6%) na upatikanaji mdogo wa umeme (43.7% hawajaunganishwa).
Grid electricity, cooking fuel (solid = deprived), improved water source, improved sanitation. Dominates due to near-universal solid-fuel cooking (96.6%) and limited grid access (43.7% unconnected). Umeme wa gridi, mafuta ya kupikia, maji bora, usafi. Inatawala kutokana na upikaji wa kuni (96.6%) na upatikanaji mdogo wa umeme (43.7% hawajaunganishwa).
Statistical ValidationUthibitisho wa Takwimu
All five dimension scores are non-normal (Shapiro-Wilk p<0.001; KS p<0.001). Parametric and nonparametric tests run in parallel — all decisions agree.
Urban households score higher than rural (Welch t=9.78, d=0.12). Female-headed households face elevated vulnerability (t=5.21, d=0.07). Education gradient confirmed (ANOVA η²=0.015). All effect sizes are small — the HFVS is not reducible to demographic proxies.
Proxy ML model (LightGBM-XGBoost blend, isotonic calibration): AUC=0.782, Brier=0.186. County-level Spearman ρ=0.960. Brier falls below the 0.22 actuarial threshold enabling direct premium-loading. Alama zote za vipimo vitano haziko katika usambazaji wa kawaida (Shapiro-Wilk p<0.001). Njia za parametric na non-parametric zilitekelezwa sambamba.
Kaya za mijini zinapata alama za juu zaidi kuliko vijijini (Welch t=9.78, d=0.12). Mfano wa ML unafikia AUC=0.782 na Brier=0.186 — chini ya kizingiti cha 0.22 cha uaktuwali.
Urban households score higher than rural (Welch t=9.78, d=0.12). Female-headed households face elevated vulnerability (t=5.21, d=0.07). Education gradient confirmed (ANOVA η²=0.015). All effect sizes are small — the HFVS is not reducible to demographic proxies.
Proxy ML model (LightGBM-XGBoost blend, isotonic calibration): AUC=0.782, Brier=0.186. County-level Spearman ρ=0.960. Brier falls below the 0.22 actuarial threshold enabling direct premium-loading. Alama zote za vipimo vitano haziko katika usambazaji wa kawaida (Shapiro-Wilk p<0.001). Njia za parametric na non-parametric zilitekelezwa sambamba.
Kaya za mijini zinapata alama za juu zaidi kuliko vijijini (Welch t=9.78, d=0.12). Mfano wa ML unafikia AUC=0.782 na Brier=0.186 — chini ya kizingiti cha 0.22 cha uaktuwali.
Data Source & CitationChanzo cha Data na Nukuu
Kenya Housing Survey 2023/24 · Kenya National Bureau of Statistics (KNBS) · Stratified multi-stage cluster sampling · 47 counties · 21,347 households · 392 raw variables · Survey weight ratio 327.6× · Fieldwork: July 2023 – March 2024.
Academic Citation: [Author]. (2026). Modelling Housing-Based Financial Vulnerability and Insurance Risk Among Kenyan Households. MSc Dissertation, Strathmore University, Nairobi. Data: KNBS, KHS 2023/24. Utafiti wa Makazi wa Kenya 2023/24 · KNBS · Kaunti 47 · Kaya 21,347 · Ukusanyaji: Julai 2023 – Machi 2024.
Academic Citation: [Author]. (2026). Modelling Housing-Based Financial Vulnerability and Insurance Risk Among Kenyan Households. MSc Dissertation, Strathmore University, Nairobi. Data: KNBS, KHS 2023/24. Utafiti wa Makazi wa Kenya 2023/24 · KNBS · Kaunti 47 · Kaya 21,347 · Ukusanyaji: Julai 2023 – Machi 2024.