In 88 poor countries for which we have data, in each and every one of the 88, the PPP for food shows that poor people can buy less food than you would expect from the PPP that the World Bank is using. The reason for this is obvious on reflection. It has to do with the fact that most foodstuffs are tradable commodities: basic foodstuffs, such as rice, flour and beans, can easily be conveyed across national borders and their prices will therefore roughly mirror the exchange rates among currencies.

Our global institutional arrangements - the basic ground rules that govern our world economy - are human-made. They don't exist naturally, nor are they God-given. We make these rules, those of the WTO [World Trade Organization] Treaty for instance, which fill tens of thousands of pages. These words have been strung together by human beings and are also interpreted and enforced by human beings.

The massive corruption common in so many developing countries would be quite impossible if Western countries did not provide convenient opportunities to ship ill-gotten funds out of the country. It wouldn't make much sense for a ruler to store in his basement large quantities of stolen cash in his own country's currency. A corrupt ruler wants to be able to keep this money safe and to be able to spend it. And for this, he needs to convert it into a Western currency and store it in a bank abroad, where it can also earn investment returns and be bequeathed to his heirs.

What is really nice about the Health Impact Fund is that it is a win-win, something that without much cost to anyone makes a lot of people better off.

You can get an ice-cold Coke for around fifty cents in most developing countries, not just in the major population centers, but at the most remote and surprising places. The logistical challenges to resupplying all these outlets are enormous - and yet, the entire system works with incredible efficiency as is confirmed by the price of the product.

Companies are actually much better than governments and other bureaucracies at organizing in a holistically efficient way the extremely complex path from the examination of molecules all the way to the delivery of medicines to patients. Already in the conception and selection of research projects, companies would anticipate all the challenges down the line that they will need to overcome in order to achieve actual health impact. Bureaucratic organizations, by contrast, are notoriously bad at this sort of optimizing.

Domestic power structures are shaped in good part by global arrangements.