Models Fail

From an essay by Jean-Philippe Bouchaud in Nature a few weeks ago:

Compared with physics, it seems fair to say that the quantitative success of the economic sciences has been disappointing. Rockets fly to the Moon; energy is extracted from minute changes of atomic mass. What is the flagship achievement of economics? Only its recurrent inability to predict and avert crises, including the current worldwide credit crunch.

This sort of talk is going around quite a bit these days about the failure of economic / financial models. Jonah Lehrer draws a comparison between economics and the collapse of the cod industry, with good discussion — some commenters take him to task on the possibility that the models of the cod fishery were good, it was the implementation that was dishonest.

Now, Bouchaud (who appears to be both an economist and a physicist) is being a bit disingenuous. In physics, it is much easier to falsify a model, and therefore much easier to reject it. Any science that includes human input is necessarily more nuanced and difficult to understand. So revolutionizing economics, or conservation, takes longer. But it also requires a willingness of actors to acknowledge that the models aren’t working in the first place.

Attacking conservationists for being blind to what isn’t working is a pretty good straw man. We could re-forest the Amazon with the papers written fretting over what does and doesn’t work. But where is the revolution? Who is demanding real change these days in conservation?

[Trying to catch up on this Veterans Day, so some links may be a little... stale]

Posted by Tim on November 11th, 2008 • Add a comment
Tags:

Emergency Conservation

This is a post about conservation priorities, in which we discuss the possibility that the number of publications about how to alllocate conservation funds has a large and significant correlation with the amount of funds available, or alternatively where said publications grow steadily regardless of the current financial climate. In a paper in press at TREE, a slew of Australian authors (through a workshop hosted at the Applied Environmental Decision Analysis (AEDA) Centre at the University of Queensland, Australia) argue that conservation triage is not just an effective use of conservation funds, but the most efficient. I think others might disagree (I’m guessing that the Kakapo is kind of the SW Pacific’s answer to the CA Condor). Here is their conclusion:

Efficient resource allocation relies upon clear goals for what we hope our actions will achieve for biodiversity conservation. Decision making based on the principles of triage provides a defensible, rational and repeatable approach to prioritising conservation investments. By explicitly acknowledging the use of triage as a process for efficient resource allocation, we are able to clearly understand and scrutinise the tradeoffs resulting from investing in one action over another, thereby increasing confidence in investments. If doctors are willing to use triage in allocating resources to save human lives, why would conservation biologists be squeamish?

Maybe I’m missing something, but what the hell? Don’t get me wrong, the emergency room is a fantastic resource, but it’s not how I want my health care, and it’s not how any reasonable health care system should be designed. When people ask, “Which conservation organization is doing the best work?” I tend to answer: most of them. That’s the nice thing about a diversity of NGOs. Some have narrow missions that do targeted conservation triage, while others work in long-term biodiversity hot spots to ensure the continued safety of those regions, while still others target the root causes of environmental harm. I just don’t see the need to say that one is more important than the other. It’s also probably a fallacy to believe that money available to one organization would necessarily be available to another. As little influence we as scientists have over governmental policy, the idea that we could shift donor money from one “sub-optimal” organization to a “better” one is lunacy.

To be fair, this paper may be saying that all conservation is triage, but that practitioners need to be explicit about the trade-offs they’re making. I guess I just don’t see that as an issue. All big conservation organizations have been forced, mostly through the efforts of large donors, to define very explicitly how they approach conservation. When the efforts of an organization stray from their stated goals, it’s usually (as this paper acknowledges) because of other pressures from both within and outside the organization.

How about an example. TNC is an organization with about $5bn in assets. That’s a lot of money. They do work all over the globe, generally (maybe sometimes only theoretically) tailored to the demands on the ground. When the Ivory-billed woodpecker was “spotted” in Arkansas, TNC quietly started to buy up land in that region. They did so knowing that as soon as word got out that the area would be protected, land prices would probably go up. They were in a position to do so because they have TONS OF CASH. Now, I suppose in this case the authors of the paper and I would agree that this is an instance of triage that is good and necessary. But the point is that TNC couldn’t do triage if it didn’t have a stable organization doing its day-to-day thing. There was no way to anticipate the “re-discovery” of the Lord God bird, so there was no way for a local organization to be set up in time to start buying the land. To follow the metaphor, you can’t send out the ambulance without the hospital.

Bottrill, M. et al. Is conservation triage just smart decision making? TREE, in press. (doi: 10.1016/j.tree.2008.07.007)

Posted by Tim on October 23rd, 2008 • Add a comment
Tags: