Deep in this article is an explanation that this bank came to this analysis by comparing "incomes and the cost of living based on the cost of a basket of 111 goods and services."
This is exactly the concept that Peter came up with to help with my idea of a charity that would balance it against exchange rates to find out where someone's charitable donation would have the most value. Would you rather spend $50 to help a homeless man catch three bus rides in Oslo? Or to help a village repair roofs for twenty of their houses in the South Pacific?