Inequality in access to health, education, and employment opportunities is exacerbated in developing nations due to the uneven distribution of access to high-speed internet connections. In Colombia, the government enacted a policy (in 2012) to subsidize internet fees for low-income households to bridge the digital divide. The reductions were not granted to all plans and thus created incentives for consumers to switch between plans. We estimate a structural model of demand for internet connection plans, which we use to quantify the importance of switching behavior. We estimate the model using data on plans offered by all internet service providers to households in all socioeconomic (SES) groups across Colombia. Our results indicate that the subsidy caused a non-negligible fraction of low-SES households to switch internet plans - the majority of which switched to plans with lower speeds not higher speeds. Furthermore, the more wealthy households (of the lower SES groups) were twice as likely to switch plans than those in the lowest SES group. Our findings suggest that the impact, not only on internet adoption, but also on switching behavior should be taken into account when formulating subsidies designed to bridge the digital divide. JEL Classification: L15, L51, L86, D12, D31

No abstract

Often empirical researchers face many data constraints when estimating models of demand. These constraints can sometimes prevent adequate evaluation of policies. In this article, we discuss two such missing data problems that arise frequently: missing data on prices and missing information on the size of the potential market. We present some ways to overcome these limitations in the context of two recent research projects. Jacobi and Sovinsky (2018), which addresses how to incorporate unobserved price heterogeneity, and Hidalgo and Sovinsky (2018), which focuses on how to use modelling techniques to estimate missing market size. Our aim is to provide a starting point for thinking about ways to overcome common data issues.