One of the most basic predictions about a competitive market is that the initial allocation of a resource should not influence the efficiency of its use. A terrific paper by Gary Libecap and Dean Lueck shows that this prediction fails dramatically in the market for land.
They look at the 116 billion square meters of land in the state of Ohio. Because of an accident of history, a large fraction of these square meters were assembled into irregularly shaped parcels via an uncoordinated process of private claims by independent individuals. The rest were assembled first into rectangular parcels along the lines of the survey called for in the Northwest Ordinance and then transfered to private ownership.
It’s worth reading the paper to get all the details, but the punch line is that this difference in the initial bundling of small bits of land had a lasting effect on how they are used. Today, more than 200 years later, a flat square meter is worth 30% less if it was initially assigned to an irregularly shaped parcel.
One of the unfortunate parts of the publication process is that the journal, which still publishes in black and white, dropped a terrific Google Earth image showing an analogous persistent difference in how land was first transfered to the private sector North and South of the US-Mexico border. The image below, inspired by one from their working paper, shows that the land on the California side of the border was transferred to the private sector in rectangular parcels and that the in Mexico was not.
In this case, it is easy to see how the differences in the initial allocation of the large tracts of agricultural land influenced the subsequent use of the land, even down to the level of influencing the layout of the street grid in cities that developed later.
Look first at the irregular pattern of streets in central Mexicali, just South of the border:
Compare with this picture of the rectangular grid in Calexico, just North of the border: