January 28, 2014
News /
Posted by NCID

Anna Tompsett from the University of Columbia presented a paper titled: "Bridges Transport Infrastructure and Economic Geography on the Mississippi and Ohio 1860-2000". “This paper is really about how infrastructure influences economic structure,” she said. “Governments and international markets will spend up to $45 billion between now and 2050 on infrastructure projects”, and after a pause for effect she continued, “85% of this will be in non-OECD countries...thESE ARE the developing countries.”

She originally extracted her data from the National Bridge Inventory (NBI) dataset in the United States. The resulting dataset contained information on every bridge ever constructed over the Mississippi and Ohio Rivers. She then cross-checked the data extracted from the original database with both satellite imagery and alternative sources of information. Population data was drawn from historical censuses.

The location of a bridge is never chosen randomly. Multiple factors influence choices about bridge location including physical characteristics about the river and the difficulty of bridge construction. Normally bridges will be constructed before a river confluence since upstream there will be less water than downstream.

Using the value of agricultural land as a proxy for production, she showed that the accumulation in population is precipitated by an immediate local rise in per capita production. After 30 years, production and population density remain elevated in areas with relatively good transport access. She instituted various robustness checks including country level unobservables with county fixed effects and implemented a country quadratic trend.

“I am confident that the population of a county that experiences a 50% reduction in distance to a bridge grows by an additional 3% over the following 30 years, relative to a median growth rate of 15% over the same time period” she said in conclusion.