Twenty years ago, the landmark paper “Money for Nothing?” argued that biodiversity conservation relied too little on empirical evidence. It called for more evaluations of conservation effectiveness based on explicit counterfactuals, comparing observed outcomes with those that would likely have occurred in the absence of intervention. To assess progress towards this goal, we systematically reviewed the study designs used to evaluate one of the most widely implemented conservation interventions: protected areas. Across 614 studies published over the past two decades, half still relied on simple Before-After or Control-Impact designs that do not reliably support causal inferences, although their use has declined in recent years. The other half used more formal causal identification strategies, most commonly conditioning strategies that control for observed confounders. However, most of these studies lacked pre-protection outcome data, limiting their ability to address unobserved confounders. Because causal claims depend on causal assumptions, it is notable that few studies stated these assumptions explicitly, let alone interrogated their plausibility. Although the design of conservation impact evaluations has advanced substantially, much remains to be improved. Combining causal inference methods with expanding data streams from remote sensing and biodiversity monitoring offers a major opportunity to strengthen the evidence base for conservation.

To develop evidence-based conservation programs, practitioners need high-quality empirical evaluations that quantify program impacts and test the veracity of the theories of change that motivate program designs. One important option for achieving this goal is to embed randomized experiments into program implementation. To create a culture of experimentation in organizations that implement agri-environmental programs in the United States, the United States Department of Agriculture funded the Center for Behavioral and Experimental Agri-environmental Research. Here, we summarize the lessons learned from this center's first decade of activities, focusing on the barriers to embedding experimentation in program implementation and ways to overcome these barriers. A particularly important barrier is the conflict between the incentives for staff to generate and use evidence and the incentives for staff to demonstrate success within their organizations.

Achieving sustainability goals requires that humans change their behavior not just once but persistently. Yet despite decades of research on the adoption of conservation and climate-smart agricultural practices, little is known about the extent to which these practices persist over time. One key reason is the lack of longitudinal, field-level data. Using ground-verified, longitudinal data on cover cropping across thousands of farm parcels in Indiana (USA), we find that persistence is low and contrasts sharply with the predictions made by Indiana conservation experts. We also find low persistence in a new national dataset of self-reported cover cropping by farm operators. The potential for low behavioral persistence in sustainable agricultural practices raises essential questions about the design of conservation programs and the modeling and valuation of ecosystem services.

Ecologists seek to understand the intermediary ecological processes through which changes in one attribute in a system affect other attributes. A causal understanding of mediating processes is important for testing theory and developing resource management and conservation strategies. Yet, quantifying the causal effects of these mediating processes in ecological systems is challenging, because it requires defining what we mean by a "mediated effect", determining what assumptions are required to estimate mediation effects without bias, and assessing whether these assumptions are credible in a study. To address these challenges, scholars have made significant advances in research designs for mediation analysis. Here, we review these advances for ecologists. To illustrate both the advances and the challenges in quantifying mediation effects, we use a hypothetical ecological study of drought impacts on grassland productivity. With this study, we show how common research designs used in ecology to detect and quantify mediation effects may have biases and how these biases can be addressed through alternative designs. Throughout the review, we highlight how causal claims rely on causal assumptions, and we illustrate how different designs or definitions of mediation effects can relax some of these assumptions. In contrast to statistical assumptions, causal assumptions are not verifiable from data, and so we also describe procedures that we can use to assess the sensitivity of a study's results to potential violations of its causal assumptions. The advances in causal mediation analyses reviewed herein equip ecologists to communicate clearly the causal assumptions necessary for valid inferences, and to examine and address potential violations to these assumptions using suitable experimental and observational designs, which will enable rigorous and reproducible explanations of intermediary processes in ecology.

We reconsider one of the most widely studied behavioral biases: anchoring effects. We estimate that study designs in this literature, including replication studies, routinely fail to achieve statistical power of more than 30%. This study replicates an anchoring study that reported an effect size of a 31% increase in participants' bids. In the replication, we increased the design's statistical power from 46% to 96%, reducing the average exaggeration of a statistically significant result by a factor of seven. Our replication results reject the size of the original estimated effects. We find an estimated effect of 3.4% (95% CI [−3.4%, 10%]).

Incentive payments could cost-effectively and equitably achieve biodiversity conservation goals but could also trigger unintended countervailing actions. Here, we report on a preregistered, randomized controlled trial of a pay-to-release program among small-scale, Indonesian fishing vessels for the release of two critically endangered marine taxa from fishing gear: hammerhead sharks and wedgefish. A conventional monitoring approach, which quantifies impacts based on conservation-relevant actions (i.e., numbers of live releases), implies that the program was successful: a 71 and 4% reduction in wedgefish and hammerhead shark mortality, respectively. The experimental data, however, imply that the pay-to-release program also induced some vessels to increase their catch, thereby decreasing wedgefish mortality by only 25% [confidence interval (CI): -49 to 10%] and increasing hammerhead mortality by 44% (CI: 8 to 92%). Our results do not imply that pay-to-release programs cannot work but rather demonstrate the complexity of designing incentive-based conservation programs and the importance of piloting them using experimental designs before scaling up.

Experimental research in behavioral economics focuses on consumer behaviors. Similar experimental research on profit-maximizing producers is rare. In three field experiments involving commercial agricultural producers in the United States, we detect evidence of anchoring in competitive auctions for conservation contracts related to nutrient and pest management that were worth, on average, nearly $9,000. In these auctions, the value of the starting cost-share bid was randomized to be either 0% or 100%. When the starting value was 100%, final bids were 46% higher, on average. We find weak evidence that experience with conservation contracts may modestly attenuate the anchoring effect.

To address climate change and global biodiversity loss, the world must hit three important international conservation targets by 2030: protect 30% of terrestrial and marine areas, halt and reverse forest loss, and restore 350 Mha of degraded and deforested landscapes. Here, we (1) provide estimates of the gaps between these globally agreed targets and business-as-usual trends; (2) identify examples of rapid past trend-shifts towards achieving the targets; and (3) link these past trend-shifts to different levers. Our results suggest that under a business-as-usual scenario, the world will fail to achieve all three targets. However, trend-shifts that rapidly "bend the curve" have happened in the past and these should therefore be fostered. These trend-shifts are linked to transformative change levers that include environmental governance, economic factors, values, and knowledge. Further research on trend-shifts, as well as bold action on underlying levers, is urgently needed to meet 2030 global conservation targets.

In the United States, agriculture is responsible for the majority of consumptive water use. To reduce consumptive use in water scarce regions, policymakers have implemented a number of costly interventions. These interventions range from land retirement to subsidies that encourage the adoption of efficient irrigation technologies. In nonagricultural contexts, costly policy interventions have been complemented by low-cost interventions inspired by behavioral economics. Whether these behavioral interventions are effective in the context of commercial farming is not well understood. In a preregistered, randomized field intervention, we estimate the impact of social (peer) comparisons on agricultural groundwater users in Colorado and Kansas. More than three thousand irrigators were randomized to receive either an annual peer comparison or no comparison. The peer comparison contrasted each irrigator's groundwater use to the distribution of use by neighboring irrigators. The comparison intervention reduced average annual groundwater use by 4.05% [95% CI (-5.87%, - 2.21%)], resulting in an aggregate reduction of more than 21,000 acre-feet per year at a cost less than $1.31 per acre-foot conserved. The estimated treatment effect was larger among irrigators with lower pre-intervention water use. In the 3-year experiment, we observed no evidence that the treatment effect substantially attenuated over time. We did, however, detect within-irrigator spillovers in the treatment group: groundwater use also declined among wells that were not included in the peer comparisons (peer comparisons included a maximum of three wells). The results imply that social comparisons can be a cost-effective tool, alongside other policy interventions, aimed at reducing agricultural water use.

Economic experiments have emerged as a powerful tool for agricultural policy evaluations. In this perspective, we argue that involving stakeholders in the design of economic experiments is critical to satisfy mandates for evidence-based policies and encourage policymakers' usage of experimental results. To identify advantages and disadvantages of involving stakeholders when designing experiments, we synthesize observations from six experiments in Europe and North America. In these experiments, the primary advantage was the ability to learn within realistic decision environments and thus make relevant policy recommendations. Disadvantages include complicated implementation and constraints on treatment design. We compile 12 recommendations for researchers.

Ecologists seek to understand the intermediary ecological processes through which changes in one attribute in a system affect other attributes. Yet, quantifying the causal effects of these mediating processes in ecological systems is challenging. Researchers must define what they mean by a “mediated effect”, determine what assumptions are required to estimate mediation effects without bias, and assess whether these assumptions are credible for a study. To address these challenges, scholars in fields outside of ecology have made significant advances in mediation analysis over the past three decades. Here, we bring these advances to the attention of ecologists, for whom understanding mediating processes and deriving causal inferences are important for testing theory and developing resource management and conservation strategies. To illustrate both the challenges and the advances in quantifying mediation effects, we use a hypothetical ecological study. With this study, we show how common research designs used in ecology to detect and quantify mediation effects may have biases and how these biases can be addressed through alternative designs. Throughout the review, we highlight how causal claims rely on causal assumptions, and we illustrate how different designs or definitions of mediation effects can relax some of these assumptions. In contrast to statistical assumptions, causal assumptions are not verifiable from data, so we also describe procedures that researchers can use to assess the sensitivity of a study’s results to potential violations of its causal assumptions. The advances in causal mediation analyses reviewed herein will provide ecological researchers with approaches to clearly communicate the causal assumptions necessary for valid inferences and examine potential violations to these assumptions, which will enable rigorous and reproducible explanations of intermediary processes in ecology.

In many scientific disciplines, common research practices have led to unreliable and exaggerated evidence about scientific phenomena. Here we describe some of these practices and quantify their pervasiveness in recent ecology publications in five popular journals. In an analysis of over 350 studies published between 2018 and 2020, we detect empirical evidence of exaggeration bias and selective reporting of statistically significant results. This evidence implies that the published effect sizes in ecology journals exaggerate the importance of the ecological relationships that they aim to quantify. An exaggerated evidence base hinders the ability of empirical ecology to reliably contribute to science, policy, and management. To increase the credibility of ecology research, we describe a set of actions that ecologists should take, including changes to scientific norms about what high-quality ecology looks like and expectations about what high-quality studies can deliver.

Adaptation requires investing now to avoid future damages, and thus adaptation is shaped by discount rates. Although the role of social discount rates in climate policy design has been well documented, the role of private discount rates has been ignored. We illustrate the importance of private discount rates in shaping adaptation investments by empirically demonstrating how household discount rates are negatively correlated with investments in water storage tanks in Central America. High private discount rates are common throughout the world and are a barrier to private adaptation investments. To overcome this barrier, adaptation policies targeted at private actors should ensure that benefits accrue sooner or that costs are lowered or accrue later. Governments or private companies could also offer long-term loans that exploit the differential between the discount rate of the lender and the private borrower.

Resource-conserving technologies are widely reported to benefit both the people who adopt them and the environment. Evidence for these "win-win" claims comes largely from modeling or nonexperimental designs and mostly from the energy sector. In a randomized trial of water-efficient technologies, the ex ante engineering estimate of water use reductions was three times higher than the experimental estimate, a divergence arising from engineering and behavioral reasons other than the rebound effect. Using detailed cost information and experimentally elicited time and risk preferences, we infer that the private welfare gains from adoption are, on average, negative, implying no "efficiency paradox."

Resource-conserving technologies are widely reported to benefit both the people who adopt them and the environment. Evidence for these “win-win” claims comes largely from modeling or nonexperimental designs and mostly from the energy sector. In a randomized trial of water-efficient technologies, the ex ante engineering estimate of water use reductions was three times higher than the experimental estimate, a divergence arising from engineering and behavioral reasons other than the rebound effect. Using detailed cost information and experimentally elicited time and risk preferences, we infer that the private welfare gains from adoption are, on average, negative, implying no “efficiency paradox.”

This special issue aims to extend the active discourse on applying behavioral science-based tools to policymaking in the fields of food, agriculture, and agri-environmental issues. For this introductory article, we collected and analyzed data from the 91 submissions we received for this special issue to identify knowledge gaps and priorities for future policy research. In the submitted papers, the impacts of the behavioral interventions were small when implemented in isolation, but they were larger when coupled with other policy tools. Yet, we also found that most of the interventions that were evaluated in the submitted papers focused either on consumers or producers and thus offered little insight into other actors in the supply chain. Moreover, many of the submitted papers had shortcomings that are common in the behavioral science literature, including the use of hypothetical or low-stake incentives, a focus on short-term behavior change, and a lack of discussion about cost-effectiveness and mech-anisms. We argue that better research designs and practices are needed to improve the credibility of behavioral science-based research in food policy. We conclude by presenting insights and recommendations for researchers and practitioners that arise from this special issue.