THIS REPORTS ON a study in the low-income settlement of Kali Anyar in Jakarta – one of a group of community-based studies on crowding and health initiated in 1991. Kali Anyar, a community of about 25,000, was selected because it met the criteria of high population density, low mobility, relatively poor health conditions and low socioeconomic status. Households within the community were selected based on the level of crowding within the household and on the presence or absence of a home industry.

The core of the study was the longitudinal prospective epidemiological component, looking at the health of children under the age of three and pregnant women. Indicators for children's health were the prevalence and incidence of respiratory disease and diarrhoea, and low birth weight for infants was used as the indicator for maternal health. The study used several measurements for crowding – not only the number of persons per room but also household area per person and the number of persons per bedroom.

An analysis of the data indicated that the incidence and prevalence of both respiratory ailments and diarrhoea in children were significantly correlated to the number of people per room and to the household area per person, but not to the number of persons per bedroom. Significant confounding factors for both diarrhoea and respiratory illness were the age of the child and ethnicity. Children under one year old suffered from 30 per cent more coughs and fevers than two- to three-year olds and twice as much from diarrhoea, findings which correspond well with other studies. Children from Chinese households reported approximately 60 per cent lower morbidity rates and children from Batawi households approximately 45 per cent higher rates than the reference Sudanese group. This was explained by the family size and household conditions of each group – which pointed again to a connection between morbidity and crowding. The type of bedroom floor and the absence of an in-house toilet were also found to be significant in the occurrence of diarrhoea, whilst kitchen humidity was identified as a confounding factor for respiratory illness.

Findings on lower birth weights (and, by extension, the health of pregnant women) were unexpected. Couples with a higher socioeconomic status, living in less crowded conditions, appeared to be at greater risk of giving birth to a baby with a lower birth weight. The report offers the following explanation: when young couples live with extended family, the grandmother is responsible for teaching proper nutrition to the mother-to-be. When they live on their own, with less apparent crowding and a greater tendency to buy consumer goods, this important support is lacking.

It was recognized from the beginning of the project that, in order to translate the findings into recommendations for policy and intervention, it would be necessary to place them in a context of local concerns and perceptions. As a result, the study also included an ethnographic component, looking at community perceptions of crowding, and a qualitative investigation of crowding and health-related issues for adults. Analysis found that high densities of both people and buildings, especially in the context of poor quality housing and facilities, were most certainly perceived as problematic in the minds of inhabitants. The perception of crowding was highly nuanced, coloured by a number of factors, and more complex than the focus on in-house crowding applied in the epidemiological study. Crowding and health were related in the minds of inhabitants – both in terms of somatic symptoms and stress – but not perceived as the main health risk faced in the settlement nor as a high priority concern. High density was also seen as having a positive dimension, providing interesting social contacts and economic opportunities. In many cases, it was clear that people were willing to trade off the health drawbacks for perceived economic benefits. The research