Homelessness in Cape Town

Background

Cape Town has an estimated 14,357 people living on the street or in temporary shelters, with the figure rising year-on-year and the City's own official estimate of around 4,000 falling well below the count used by direct-service organisations. Rough sleeping is concentrated in the CBD but visible across Muizenberg, Mitchell's Plain, the southern suburbs, and the Voortrekker Road corridor. The population is heterogeneous: long-term rough sleepers, recently evicted households, internal migrants from less economically active provinces, people exiting prisons or hospitals without a discharge plan, and individuals struggling with untreated mental illness or substance use.

Consequences

Life expectancy on the street is markedly lower than the city median; exposure to violence, weather and disease compounds underlying conditions; the population includes children. Studies estimating the all-in cost of homelessness to the City — across health care, criminal justice, and emergency services — find it runs to a substantial multiple of what comparable supportive-housing provision would cost. Only about 12% of the unhoused population holds any formal employment, and routes back out are narrow.

Constraints

• Spatial inheritance. Apartheid-era forced removals shaped a metropolitan form in which the poor live far from work and amenities. Over three decades after democracy, this geometry constrains every intervention.
• Migration dynamics. People arrive from the Eastern Cape and elsewhere seeking work; without arrival support, many slip into rough sleeping within months.
• Mental health and addiction system limits. Hospitals and psychiatric facilities lack bed capacity; detox and rehabilitation places are insufficient. People are discharged onto the street.
• Municipal posture. The City has pursued eviction orders against rough sleepers on CBD pavements, treating visibility as the problem rather than the underlying condition, eroding trust and creating tension with NGOs.
• Funding decay. Programs run on a launch-and-decay cycle: visible at opening, underfunded within two years.
• Data fragmentation. The City and the NGO sector do not share a population register; volunteer point-in-time counts are recent and patchy, hampering targeting and evaluation.

Observed evidence

The 2024 volunteer point-in-time count — involving U-turn, Streetscapes, MES, The Haven, New Hope SA, and the Voortrekker Road Corridor Improvement District — is the most recent serious attempt at granular data. Survey work by Hopkins, Laitinen and Skinner (International Journal on Homelessness, 2024), covering 350 individuals, found that dominant responses are either punitive (displacement) or compassionate-relief (food, handouts), with neither addressing root causes. Programs combining shelter with skills, peer employment and structured exit pathways (U-turn, MES, Streetscapes) report concrete exits to housing and employment, but at small absolute scale relative to the population.