Science1 publisher2 min readPublished
Four decades of EU land protection show no effect on vegetation cover or night lights
Prakash Mishra and his co-authors measured the causal effect of EU land protection over four decades and found nothing on vegetation cover or nighttime lights, and he traces the result to who chose the parcels.
The Scientist · Science desk
What happened
- A paper in The Review of Economic Studies estimated the causal effect of four decades of EU land protection, covering farmland and managed forests as well as wilderness, on vegetation cover and nighttime lights.
- The authors report no meaningful impact on either outcome, a result Mishra calls surprising.
- Mishra says the conclusion is not that land protection should stop, but that the EU's incentives for protecting land were poorly designed.
- Mishra has joined Caltech as an assistant professor of economics after a year at Dartmouth College as a Globalization Postdoctoral Fellow.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision Whoever drafts the next protection target has to decide whether to constrain where member states place parcels, since the authors' complaint is about how the incentives were designed.
- exposure Carbon and biodiversity programs that treat protected hectares as the deliverable are open to the same test, and hectares under no development pressure will not produce a measurable change in either direction.
- capability Continent-wide satellite records now let a designation regime's average effect be estimated decades after the fact, so a program can be asked for measured outcomes rather than projected ones.
- precedent A causal estimate of protection published in an economics journal raises the expected standard of evidence: the comparison becomes what would have happened to those same parcels.
If the designation had bound, the estimate should have moved on one side or the other: vegetation cover up where land came out of production, night lights down where building stopped. Neither outcome moved [2]. That pattern fits parcels that were already going to stay close to the way protection would require.
Mishra's account of how the parcels were chosen fits the same reading. "The EU Commission would set area-based targets to protect land, but it was local governments who chose how to site protected areas," he said [4]. Local governments, he said, were "beholden to their voters, chose to protect land at little risk of development and with protection frameworks that still allowed human use" [5].
The parcels were selected to be cheap to protect, so what the estimate describes is that siting rule. The general cost of taking developable land out of use is a different question. The estimate also rests on two measures a satellite can supply continent-wide for forty years. Carbon sequestration and water filtration, the services Mishra names as the focus of his research, are not among them [6].
"No meaningful impacts" is the authors' characterization of their own estimates; the interview does not report the study's identification strategy, its sample or the smallest effect it could have detected [2][11].
The question here is being put as a measurement problem. Mishra's stated approach is to use global- or continent-scale data for empirical answers to hard policy trade-offs [6], and a forty-year designation record across the EU is long enough to support that kind of estimate [1].
The design claim is narrow and portable. A target written in hectares can be satisfied with hectares under no development pressure. The area figure is the only instruction that reaches the official siting the parcels.
Mishra took a bachelor's degree in economics at Penn in 2019 and a doctorate in applied economics there six years later, in 2025 [8][10]. He said of the EU work, "This work connects to a broader research agenda of mine on designing market incentives that actually serve the dual goals of environmental conservation and economic development" [7].
What to watch
- The paper itself: the sample, the identification strategy and the confidence intervals behind "no meaningful impacts".
- Whether the next EU area-based target adds a development-risk condition to how member states site parcels.
- Replication in regions where protected boundaries fall on land under active development pressure.