We take advantage of a quasi-natural experiment in Peru in which a privatized telecommunications company was required by the government to randomly install and operate public pay phones in small rural towns throughout the country. Using an especially designed household survey for a representative sample of rural towns, we are able to link access to telephone services with household income. We find that, regardless of income measurement, most characteristics of public telephone use are positively linked with income. Remarkably, the benefits are given at both non-farm and farm income levels. The findings hold when using propensity score matching methods. © The London School of Economics and Political Science 2008.