Distributing opportunities through a highway? Infrastructure as an opportunities generator for vulnerable regions in El Salvador

Baires Montano, Wilber Isaac; Dinarte Diaz, Lelys Ileana (2016). 'Distributing opportunities through a highway? Infrastructure as an opportunities generator for vulnerable regions in El Salvador' Paper presented at the annual conference of the HDCA, Tokyo 2016.


abstract
Justification and theoretical framework
There is a wide line of literature considering the construction of public infrastructure as a motor for economic development. Empirical work, however, just recently tested this statement in a formal way, but still ignoring some important development outcomes. Faber (2014), for example, only examines the impact of the construction of highways on industrial GDP on peripheral Chinese regions. Banerjee, Duflo and Qian (2012), in other hand, limit themselves to explore the causal effect of proximity to transportation networks on per capita GDP levels across sectors. Both papers find positive effects. Also, there is evidence of short-run impact of infrastructure on another different outcomes, such as local trade (Duranton, Morrow and Turner, 2013), individual sales and wages of high qualification workers (Michaels, 2008) and firms exports (Martincus, Carballo and Cusolito, 2012).
Therefore, the objective of this paper proposal is to contribute to this literature examining the impact of infrastructure on alternative development outcomes of particular interest for a developing country context. Additionally, we will explore heterogeneous effects due to gender. In order to evaluate the effect of a infrastructure policy, we will use the construction of the Northern Transnational Highway (223 Km length) in a highly vulnerable region in El Salvador, Central America under the FOMILENIO I compact between the Millennium Challenge Corporation (MCC) and the national Government. These agreement encompasses three components: Connectivity project, Human Development Project and Productive Development Project. The first one intends to physically unify El Salvador's Northern Zone with the rest of the country, and to enable new economic opportunities for rural households, lower transportation costs, and decrease travel times to markets.
Our main outcomes are related with the components of the Human Development Index (HDI): Education, Health and Income. In terms of health, our prior here is that the highway construction improved the access to health services due to the reduction of transportation costs. The measures to prove this will be changes in child mortality, teen pregnancy, and number of women attending pregnancy controls. In terms of education, the existence of roads of high quality increased means of transports supply (before the Highway the transport options were only bikes, a few buses and cars, and the principal means of transport were the carts). The highway also connected the region with a Technical Education Center. This allowed students to reduce the transport time and costs, created the opportunity to access to technical education and therefore possibly changed the dropout rates and education performance. The measures we will present are high-school enrollment, technical education enrollment and academic performance.
In terms of income, the construction of this infrastructure could have created more economic opportunities in the region such as increase in local trade, land value, crop or main activity diversification, among others. In line with Michalopoulos and Papaioannou (2013) and Henderson et al (2009), we will use the light density at night data reported by the NASA as a proxy for income level (even more interesting, Elvidge et al. (1997) and Min (2008) show a strong correlation between the density of light at night and variables such as the provision of public goods, poverty rates and access to electricity). In the investigation, we will also attempt to analyze the mechanisms which could be generating some of these outcomes. For instance, the improvement of the economic activity could be caused by the generation of new businesses, by the increase of sales from the incumbents or both. We will try to decompose some of them.
Finally, we will present results on labor outcomes. Our hypothesis is that the highway construction could change the child and youth labor supply, but we are not sure yet on what direction. On the one hand, it is possible that due to the reduction of transportation costs, the children and youths will attend school more. But it could also occur that now they have access to more diversified job offers and therefore they prefer to work than to go to school. With our available data, we will identify those mechanisms. 
Data
Our data comes from different sources at the municipal level. The health and education data will be obtain from the Ministry of Health and Education, respectively. We will have a panel of municipalities in the Northern region of El Salvador from 2005 to 2015. The income related data will be obtained from the NASA's "Defense Meteorological Satellite Program's Operational Linescan System (DMSP-OLS)''. And finally, we will get the data of the highway construction from the MCC and the Technical Secretary of the Government of El Salvador. 
Methodology
The Northern Transnational Highway was built in 6 phases from 2007 to 2012. So, the phase at which each municipality was on each year generates a good variation to use, and they are useful to define them as part of the treatment or control group. We will take advantage of this process of construction to apply the Differences-in-Differences methodology.
We will also apply the methodology of Faber (2014) and Banerjee, Duflo and Qian (2012) which use straight lines, a quasi randomization and minimal cost routes (MCR) as a exogenous variation source on trains and highways in China.  

scroll to top