To this model I would add the transaction costs for vetting a transaction, the cost of identifying and engaging transaction partners, and the relative sensitivity to a negative outcome (the stake as a percentage of total stake).

I believe that would enable you to identify more or less corrupt industries.

Unfortunately, both stakes and information costs make governance prone to abuse. To see why it’s not nearly as corrupt as one might expect from this model, you’d need reputation cost and benefit, where trusted governments and leaders attract higher functioning citizens and industries.