December 30, 2019

Persistence and reversibility

When assessing the merits of different interventions to improve wild animal welfare, two prominent factors are (1) how persistent a certain intervention is and (2) how easy it is to reverse the effects given a bad outcome.

Reversibility is important because we can’t always predict upfront exactly how an intervention will turn out. If things go wrong, we want to be able to reverse the intervention to contain the damage. We anticipate having a better empirical understanding in the future, so to preserve option value, we probably want to delay persistent and hard-to-reverse interventions until we’re highly confident they’ll have positive effects.

Persistence, or the duration of the effect in expectation, is important because it will affect how much effort we would have to put into improving wild animal welfare by a certain amount. If the effects are very persistent, we only have to initiate the process once, then we can reap the benefits over a long period of time. 

However, the benefits of higher cost-effectiveness of persistent interventions seem to be strongly outweighed by other factors in most cases. Firstly, most highly persistent interventions are very hard to reverse, which means we have to give up a lot of option value. Secondly, trying to implement highly persistent and hard-to-reverse interventions could stir up a lot of animosity, which would be damaging for the long-term stability of the wild animal welfare project.

Thus, it seems like if we have to choose between an intervention with low persistence but high reversibility and an intervention with high persistence but low reversibility, we should choose the former rather than the latter.

Striking a balance

There might be a way, however, to increase persistence in some scenarios without paying with reduced reversibility. More precisely: if we could find interventions that are persistent in the face of natural processes, while at the same time being highly reversible by humans, we could get some of the benefits of persistence, while avoiding the negative consequences of implementing irreversible interventions. 

With continued work on their efficacy and safety, gene drives might possess these characteristics in the future. A gene drive is a genetic modification that makes itself more likely to be inherited. Gene drives can be persistent because they are able to modify the genome of every individual in a population or species. Gene drives can also, under some circumstances, be reversible. A secondary gene drive can target the primary gene drive, halting its spread and removing the changes it made. However, it is important to note that the safety and reversibility of gene drives is critically dependent on how much effort is put into making them safe and reversible, which is one reason why the development of policies regulating gene drives will be of high importance.

Despite gene drives being an important exception, most actions will probably be either non-persistent and reversible, or persistent and non-reversible. In my report, I explore three examples in the second category: extinctions, climate change, and energy efficiency improvements in plants. Even though all of these examples have somewhat unclear effects on wild animal welfare, they should probably be avoided due to the fact that they seem very hard to reverse, and will persist for a long time.

Cover photo: Adapted from “Loggerhead sea turtle hatchlings crawl to sea,” © Blair Witherington, Creative Commons