Social networks are increasingly being recognised as having an important influence on the health market outcomes, as they facilitate the exchange of information on health care related issues. Networks reduce search costs by providing information to peers about the appropriate health care providers and details about the functioning of the health care system. In this paper, we examine the impact of information externalities generated through network membership on maternal health care utilisation in Tanzania. We further propose new approaches for quantifying the size of one's network. We adopt an econometric approach that minimises the problems of omitted variable bias. Using the Demographic and Health Survey data for Tanzania, a country characterised by low levels of maternal health care utilisation we find that social networks may enhance antenatal completion and early antenatal check-up probabilities by an additional 6-35 percent and sometimes up to 59 percent. The results suggest that failure to adequately control for omitted variables would lead to substantial under-estimation of the network eect. Finally, we show that irrespective of the measure of the size of the networks, high quality
networks have better outcomes than low quality networks.