 Our motivation was originally that we saw one particular study being done by Will Cavendish from Oxford University at the turn of the millennium in Zimbabwe, where he was meticulously registering over a full year the different sources of household incomes, including environmental income sources. And that study found a relatively high share of environmental income. And that came to be a strong argument in the debate where those who are advocates of forests and the environment said, look, incomes can't be as high as this. But many of the sort of more agricultural development-oriented economists would say, it's just a one-off study. What can you show from that? It's a particular setting. So our idea was let's have a look, let's clone this Cavendish approach and have a look in different parts of the world to see what patterns we can find. So we associated ourselves with PhD students at different institutions. Some of them we sent out, others came to us because we could offer them a home where they could exchange experiences. We offered them methodological guidance and in turn they would gather these data as a sort of some socioeconomic baseline data for their own study, typically. They would then add data of their own that they were using for their particular hypothesis of their respective PhD studies. So we ended up with three dozens of sites of studies and more than 8,000 households in our database. That was much more than we had originally imagined. We managed to, in terms of the sample composition, we managed to get people working in different parts of the rural tropics. We had some gaps, for instance, in West Africa and managed to get funding from Danida to fill some of these gaps and do studies where we actually chose exactly the sites where we wanted our study done. So we could ensure a good geographical distribution. And we ended up, in our results, we ended up actually confirming the Cavendish hypothesis that environmental incomes continue to be very high. People, even 10,000 years after the agricultural revolution started, people actually rely to a large degree still on foraging in forests and in bushlands and other natural sources from where they can get contributions to their livelihoods. So I was actually surprised about how large that income share is. I would not have thought that we would find such a large share. But that's kind of the charm about research when you can still also yourself get some interesting surprises. We do hope that we can also make a case for the World Bank and other institutions to make a forestry module in their living standard measurement service, the LSMS service. And we are actually right now in a project with FAO, PRO4 and other partners where we are working with the World Bank to try to use some of the PEN results and other research results from other groups to elaborate such a forestry module. So that, I mean, I see our PEN project very much like a bean counting exercise. We're counting the agricultural beans and the environmental beans and the things that come from wage, et cetera, so that we have a more accurate accounting of welfare sources. And we would like to help the World Bank and their partners in many African countries, for instance, National Bureau of Statistics to do a better job in doing this environmental bean counting because it will contribute to a more realistic vision of the welfare that these households have but also where does it come from and what are the leverages that could alleviate poverty or vice versa restricted access to some of these environmental resources that could limit their welfare.