Estimating the Willingness to Pay for Tenure Security in Brazilian Favelas

This paper examines willingness to pay for housing tenure security in favelas in six Brazilian states, Ceara, Paraiba, Pernambuco and in the north-east, Minas Gerais and Sao Paulo in the south-east, and Rio Grande do Sul in the south, using data from the national household survey (Pesquisa Nacional por Amostra de Domicilios, PNAD) for 2002. We use a quasi-experimental technique by combining the inverse probability weighting estimator proposed by Hirano and Imbens (2001) with an interval regression model to shed light on what might be seen as the effect of title ownership on housing values in Brazilian favelas once selection on observed characteristics are controlled for. We also use state fixed effects to control for idiosyncratic characteristics of states. We interpret the resulting estimate in the same vein as Friedman et al (1988) as an estimate of the willingness to pay for tenure security. As far as we know, this is the first attempt at estimating willingness to pay for tenure security in Brazilian favelas. Our results suggest that people living in favelas are willing to pay on average an additional 18% of the value of their house for tenure security. However we find that among the poor willingness to pay for tenure security is substantially less, and for some disappears entirely. We suggest that this may be either because households in smaller and longer established favelas may have developed informal mechanisms to ensure their security or because the poor are skeptical about the real security the title represents in that environment. We argue that these estimates are useful for public policy in Brazil, where millions of households live in informal dwellings and interventions involving land regularization do not appear to account for households’ willingness to pay. We also simulate the likely cost of a titling programme and the consequences for household debt burdens.