How efficient is capital allocation in rural developing economies? II

This MIT working paper that discovers that interest rates do not equal the marginal product of capital in rural Indian bank lending that I found in Brad DeLong's blog, here are my comments (no doubt I'll discover that I'm full of it once I find detailed analytical writings with or without math on this topic, the search begins) :

The author of the paper points out the legal system may cause some moral hazard distortions, but doesn't provide many background references to the general problem in the paper.

I can imagine institutional historical details that might cause MPK to deviate from interest rates. Legal delays when the economy goes into a tailspin, defaults increase, eventually leading to a culture of default, then the spiggot of funds gets turned off to the bank, the land is repossessed by the government, resold at ridiculous prices by the bank, but what the buyer is really buying here is the loan, because loans are so scarce, what I'm describing here is post-1997 reality in Chiang Rai, Thailand, where I actually live next to farmers. Sometimes, land holders even work out a deal with the bank, you keep this plot and I'll do a pig farm on the plot next to it.

What the moral hazard effect are of this labyrinth of institution on the rural banking system over the long-term, there might be a neoclassical description if augmented with institutions, the actors are very rational, even if the institutions aren't always.

More interesting, a small agriculture community that was socially homogeneous and insulated from the outside world, with their own moneylenders and their own, more efficient modes of debt collection, a little world of Ricardo's time, let's say,...it would be interesting to see whether neoclassical MPK = interest rates held then.

In a broader context with EXIM loan guarantees or MIGA political risk insurance, then there's a different moral hazard problem that probably causes MPK not equal interest rate.