Opening the black box of land institutions

Since Acemoglu, Johnson and Robinson (2001), the role of historical path dependencies and long-run persistence has been reaffirmed by a wide variety of papers (Legal origin: Glaeser and Shleifer, 2001; Slave trade: Nunn, 2007; Colonial Land Tenure Systems: Banerjee and Iyer, 2004 etc.pp.). In line with Banerjee and Iyer (2004), the recent working paper by Libecap, Lueck and O'Grady (2010) released last week examines the persistence of land instutions in former British colonies.

Focusing on two different types of land demarcation - a centralized, rectangular system and a decentralized, metes and bounds  system - the authors introduce a simple framework to model the decision on which institutions to adopt: The rectangular system lowers uncertainty and transaction costs but requires centralized organizational capabilities, while the metes and bounds system is more arbitrary and insecure but easier to establish. A brief empirical analysis testing the hypothesis for US, Canada, Australia and New Zealand concludes the paper.

This paper seems to be in line with the vast body of work that goes beyond the simple "institutions matter" and actually tries to unbundle the Gordian knot of highly interdependent and jointly endogenous mechanisms. In fact, one must not even look at British colonies and run regressions to see the long-run importance of land institutions: Flying across Germany, it is very easy to tell above which part of the country you are by simply glancing at the average plot size and shapes. Using related modern image processing techniques, it could be possible to generate more testable data from satellite images such as in Alesina, Easterly and Matuszeski (2010).