Does the Endangered Species Act work?
at the Andrew Young School of Policy Studies

The earliest piece of federal legislation written to protect and conserve wildlife in its natural habitat, the Endangered Species Act of 1973, is no less controversial today than it was 30 years ago. Why? According to Paul Ferraro, an environmental economist and assistant professor at AYSPS, no one has conducted a well-designed empirical study of the ESA’s effectiveness.

“Many people believe the Act places an unfair burden on private landowners,” says Ferraro. “On the other side, many biologists and economists are concerned that it may be ineffective at achieving substantial recovery of endangered species. But even after thirty years, no one knows exactly how well the act works.”

“This is a huge deal. Biodiversity is declining worldwide, and ESA is our centerpiece policy tool for dealing with the issue.” In fact, as many as one-third of the native species in the United States are at risk, according to NatureServe, a nonprofit conservation organization that collects and manages biodiversity data on more than 20,000 plant and animal species. In a 2002 Nature Conservancy report that analyzed this data, Georgia ranked among the top 10 states with the greatest level of risk for disappearing species.

“If the act is ineffective,” says Ferraro, “then everybody should be aware of that.”

So Ferraro and Monica Ospina, an AYSPS student pursuing a Ph.D. in Economics, have found a way to use empirical research in a unique, ongoing project to detect whether the ESA has had a positive or a negative effect on endangered species in the United States. “We are examining the question in a way no one has tried,” he says.

Ferraro believes the key problem is that no one knows whether endangered species would be any worse off if they were not listed. “But we can’t observe that, because they’re listed. We don’t know what the counterfactual looks like,” he says. Earlier researchers either did not use a control group or they simply used unlisted species as a control. “You must have a control,” he says, “and you must have the right control, a group similar to the endangered species, except that one is listed and one is not. You want them to be roughly similar in all the characteristics you think would affect recovery.”

For their research, Ferraro and Ospina are using “matching,” a technique that he says is relatively new in policy work. “Matching is mainly used in labor economics,” he says. “It has been used probably less than 10 years in the economics and policy fields. Nobody has published its use in conservation economics.

“Matching allows us to take a larger data set and use a more sophisticated method to get at the question: can we detect a net effect?” Using measures of species recovery between 1993 and 1997, Ferraro and Ospina have estimated the size and significance of ESA listings on species recovery for 490 vertebrates.

Ferraro volunteers what their research has found to date, but only with a caveat: “Our preliminary findings suggest that the ESA listing, by itself, does not appear to affect the probability that a species will recover or decline. However, we are also finding that the combination of substantial government funds for recovery initiatives with an ESA listing does lead to a substantial increase in the probability of recovery and a decrease in the probability of decline,” he says.

“Our caveat, and this is critical, is that we don’t believe our analysis right now is a good enough case without expanding and improving the data set. Statistical studies can often change: you add a new variable, you expand the data set to cover more years, and all of a sudden your results disappear or change.” He and Ospina will improve and expand their data set before they release a final report.