Two interesting advances today:
1. For the 'railways and rents' paper: I have built a 'panel' dataset using annual total railtrack in a 40km radius of three estates: Dalemain, Holkham Hall and Petworth. I regressed the rent against the track, cattle and wheat prices. The three independent variables---track, wheat and cattle---are all significant and with the 'right' signs. This is very satisfying. I'd like to get more data specific to each estate, such as yields etc. Mi is working on this. Soon we will develop a clear picture of how agricultural rents were set in the 19th century. This is something no one has tried before. The techniques will have analytical uses in developing countries where data is --- like 19th century Britain --- a bit sparse.
2. For the 'Devon rents' paper: I built a variable which represents the amount of money a farmer would get after he had paid his farming expenses. I regressed this, together with population of nearest market town and elevation, against rent. The results are highly significant. Then I built four 'windows' moving from west to east, so that I selected only the farms inside the windows. For each window I calculated the 'elasticity', which is the percentage change in rent for a percentage change in farmer's take-home money. I think (!) that this is a measure of the 'surplus extraction' of the landlord: how much he can 'squeeze' out his tenant. What is remarkable is that the elasticity changes as we move east towards London. It more than doubles over two hundred miles. This is a fascinating result, but at the moment I am at a loss as to how to explain it! I have put the elasticities into the relevant counties in the map below....not quite the same as the moving window but I can't see how else to show you.
It has been a good day. Thank you!