Payments for Forest Conservation

I spent a good chunk of last week working with US and Mexican colleagues, looking at data from a forest conservation program in Mexico. This “Payment for Environmental Services” (PES) program is in the highly threatened area where Monarch butterflies stay over winter. The program pays producers and communities not to log their forestland, as a means to halt the rapid deforestation that has threatened the area for years. While in many cases, PES programs have been placed in regions with little initial threat of logging, this region is at the opposite end of the spectrum. The program operates in an area with a very high threat of logging, and the real worry is that the most tempting lands left are those with the relatively pristine, dense forests – the ones in the core of the Monarch butterfly area.

At first glance, one might think the program looks like a great success. The land inside the program area is much more heavily forested than the land outside, so a quick comparison of average forest cover argues that the program is working. At a second glance, the program looks like an abject failure. Comparing deforestation rates, we see more logging inside the core region than outside, so clearly the payments aren’t working. The real problem of course is twofold: first, the core land was selected to be in the program because it was more forested. Second, as other lands are deforested, the core zone becomes increasingly attractive to loggers, and it becomes harder and harder to keep the illegal loggers out.

These PES programs, where villagers are paid for providing an environmental service, are notoriously hard to evaluate. First is the problem that many such programs are put in places where people would likely have never cut down the forest in the first place. So the program may be only paying people for doing something they would have already done anyway (“additionality”). To determine the true opportunity cost of preserving this land – i.e. to determine whether these communities would have kept their forests intact without the payment – it’s helpful to have a control group to compare them to. But finding true control groups is tricky. For example, in places where there is a lot of logging pressure, by taking some land out of production, other neighboring land becomes more valuable, and therefore subject to higher logging pressure (“spillover effect”).
If this second effect is in place, just comparing the land use of those receiving the payment to those who are not will give a biased result. Since the program is driving increased logging into the control group, it’s going to look like the program is working amazingly well, even though overall deforestation hasn’t changed.

After trying to explain the pressure for logging using physical characteristics of the land and transportation cost, I found that when one looks at the map of deforestation, deforestation follows community boundaries amazingly well. So, for all that one would expect the high-value forest to get cut first, or land close to roads to be particularly at risk, deforestation, at least in part, seems to come down to community governance. Some communities have not only ceased logging, they are patrolling their forest to keep their trees safe. Other communities are not being so diligent. In some cases the illegal logging is spilling over from neighboring logged regions. In other cases, the community itself is split, and the faction that didn’t agree with the decision to participate in the conservation program just went ahead and logged anyway. We also observe some communities who refused to take the money even though they were formally losing the right to log. Rumors are that leaders in these communities did not want to give up the opaque payments they receive from the forestry companies in exchange for the very transparent payments they would receive from the conservation fund. Besides, illegal logging is rarely prosecuted, so losing the right to log was not apparently as great a threat as one might hope.

Why do we care? One reason we might is that a number of countries are looking at imposing these PES programs to pay for carbon sequestration (often referred to as “reducing emissions from deforestation in developing countries, or REDD).   The point is that when we consider any form of payment for environmental service, we need to consider the community institutions needed for such a program to be successful. Now, the situation in this conservation program in Mexico is complicated by the fact that land is often managed communally, but even if one is relying on a group or institution to organize private property owners, one could see similar problems. Second, much land in developing countries, particularly land that might be the target of a carbon-based reforestation program, is owned collectively. If all we are interested in is preserving carbon, we may just want to target well governed communities, to ensure that PES actually produce what was purchased. If we want to deliver development goals, we may need for some thought to go into designing PES programs that not only pay communities, but also help them establish the institutional structure needed to deliver the services purchased.