How does nature enhance the prosperity of poor people? It’s a hot question among conservationists right now — but in reality, it’s irrelevant both to helping the poor and to conserving nature. That’s because poor people undoubtedly depend on nature and its services.
So here’s a better question: How will any specific conservation action aimed at changing how humans use nature affect the prosperity of poor people?
Conservationists, it turns out, have very little evidence with which to answer this question. In large part, we lack that evidence because we have failed to adopt modern methods the rest of science has developed to measure causal relationships.
This is tragic, because conservation has no shortage of good ideas for increasing the quality of ecosystems and the services they provide people — ideas ranging from protected areas to performance incentives, decentralization to zoning restrictions to information dissemination about sustainable practices. Yet each of these interventions likely translates into different effects on the poor, even if each of these interventions’ effects on nature are the same (which they likely are not).
Protected areas, for example, not only change land-use patterns near the poor, but they also can create tourism opportunities and lead to changes in infrastructure like roads and electrification. In contrast, voluntary performance payments for environmental services are more spatially diffuse and rarely affect tourism opportunities or infrastructure development. They do, however, transfer large amounts of cash to rural residents — who may or may not be the poorest households. Unfortunately, we have an acute shortage of evidence about the heterogeneous effects of different conservation interventions.
And by evidence, I mean credible inferences about causal relationships between actions and effects from many sites across the globe. What are the differences between the welfare of the poor with nature conservation interventions and what their counterfactual welfare would have been in the absence of the interventions?
Let me be clear: We need more than simulations of these causal relationships. We need conclusions drawn from observable data from real conservation programs. Only with such evidence can we begin to design policies and programs that deliver both environmental benefits and enhanced prosperity for the world’s poor.
But such evidence is nearly absent in the conservation community. In 2005, the Millennium Ecosystem Assessment reported (p. 122) as one of its main findings in its Policy Responses volume that “[f]ew well-designed empirical analyses assess even the most common… conservation measures.” Eight years later, Miteva and colleagues (2012) report that their review of the conservation evidence base “confirms previous claims that causal evidence of the effectiveness of conservation interventions…is rare.”
Gathering that evidence is not just possible, however — it is within reach, if we learn the lessons of other fields.
In the last 20 or 30 years, a revolution has taken place in the ways in which scientists draw causal inferences from observational (non-experimental) data. Fields from medicine to education, criminal justice to job training have identified creative ways to ascertain whether their interventions are having an impact. By focusing on the process by which some people or areas are exposed to an intervention and others are not, these study designs can isolate the effects of the intervention separate from other factors that also affect the measured outcomes. A hallmark of this revolution is transparency in terms of what effect is being estimated and how it is being estimated.
Conservation science and practice has largely remained unaffected by this revolution, although that situation is gradually changing. But if conservation practitioners and donors do not encourage this change, the field will be forced to continue to rely on intuition and anecdote — a reliance that will damage the credibility of conservation’s claims to be relevant to people.
For example, we should have better knowledge about whether and how protected areas — a ubiquitous tool for protecting ecosystems and their services — impact the poor.
The popular wisdom is that protected areas are subject to the following pattern of tradeoffs: the stricter the protection, the more positive (desirable) are the environmental effects, and the more negative are the social effects.
The purported mechanism for this tradeoff is simple: protected areas stop damaging human uses, but in the process impose losses on their poor human neighbors who would have benefited from these uses.
Recent impact evaluations of national protected-area systems suggest, however, that this pattern may not exist or may even be reversed in many countries (Canavire-Bacarreza & Hanauer 2013; Ferraro et al. 2011). Desirable environmental outcomes can be smaller when there are stricter rules, and poverty reduction can be larger. How can that be?
What are the differences between the welfare of the poor with nature conservation interventions and what their counterfactual welfare would have been in the absence of the interventions? Such evidence is nearly absent in the conservation community. — Paul Ferraro
For political reasons, strictly protected areas are often shunted to less productive (and thus less threatened) areas. Thus, the amount of avoided environmental loss these areas can generate is often smaller than a less strictly protected area, which can be sited on more productive lands (Ferraro et al. 2013). So while the less strictly regulated protected area may permit more damaging uses, it has a greater potential for stopping environmental losses and adding gains because it can be established on lands that could have been completely converted to oil palm or soy plantations, for instance, in the absence of protection.
But the full possibilities are even more complicated. For poverty reduction, strictly protected areas offer a mechanism that less strictly protected areas do not: tourism. While less strictly protected areas permit more uses, these uses may be low value (or co-opted by elites). These uses may also fail to compensate the poor for the more-damaging (but perhaps more lucrative) productive activities that protection stops.
For example, a recent study (under review) estimated that much of the poverty-reducing power of Costa Rica’s protected area system over the last 30 or 40 years arises from tourism opportunities. Changes in ecosystem services, if they have affected the poor at all, have only served to offset the losses from protection’s restrictions on agriculture and forestry.¹
We don’t know which patterns hold for other countries — we simply have few evaluations that can guide us. Even within a country, there can be important variation in the impacts of policies and programs. A 2013 study found that strictly protected areas in Sumatra generated more avoided deforestation than less strictly protected areas (Ferraro et al. 2013) — but a recent presentation by some of the same authors suggested that the opposite was true for the rest of Indonesia (D. Miteva, 2013 ICCB). Given that we can expect heterogeneous interactions between nature protection and poverty across the globe, conservation has a critical need for more impact evaluations of interventions on environmental outcomes and poverty.
If impact evaluation and evidence-based policy were fully developed for conservation interventions, the Holy Grail might look like a figure that shows the relative cost effectiveness, based on observable characteristics of the site, of different conservation interventions across multiple effects in which practitioners, scientists and citizens might be interested (e.g., the cost-per-unit increase in habitat connectivity and unit decrease in poverty for African savannah sites where transportation infrastructure is undeveloped and tourism potential and institutional capacity are weak).
Like the Holy Grail, such figures may be unattainable for many sites in which we work. But they represent what we should be striving for in order to advance our understanding of how protecting nature can help enhance the prosperity of people.
Fundamentally, we need to know what works and under what conditions. And to develop this knowledge, we need to ask the right questions about impacts, collect the right (not more) data², and analyze the data using modern empirical designs for causal inference. The evidence base conservation needs will not magically arise with a single global-level study. It has to be carefully built up from numerous national, regional and local studies if we are to truly understand how our actions to protect nature affect the prosperity of poor people.
¹Another study on the same system (citation) found that the poorer a community adjacent to a protected area was before protection was established, the greater the poverty reduction, on average, compared to equally poor communities that did not eventually have a protected area put near them.
²From the perspective of modern approaches to impact evaluation, the conservation community invests too many resources measuring the status and trends of environmental outcome indicators and not enough measuring factors that affect both the outcomes and exposure to the environmental programs (i.e., confounders, or rival explanations for changes in status and trends). Moreover, the community typically fails to measure social outcomes, although more recently conservation scientists have found creative ways to use existing administrative data in order to evaluate past program impacts on poverty (citations), or to collect data at the start of new conservation programs in ways that permit one to estimate the social impacts of these programs (WWF MPA Indonesia; UNEP PES Uganda).
Canavire-Bacarreza, G. & M.M. Hanauer. 2013. Estimating the impacts of Bolivia’s protected areas on poverty. World Development 41:265-285.
Ferraro, P.J., M.M. Hanauer and K.E. Sims. 2011. Conditions associated with protected area success in conservation and poverty reduction. Proceedings of the National Academy of Sciences 108(34):13913-13918.
Ferraro, P.J., M.M. Hanauer, D. Miteva, G. Canavire-Bacareza, S. Pattanayak, & K. Sims. 2013. More strictly protected areas are not necessarily more protective. Environmental Research Letters 8:02511.
Miteva, D., S.K. Pattanayak, & P.J. Ferraro. 2012. Evaluation of biodiversity policy instruments: What works and what doesn’t? Oxford Review of Economic Policy 28(6):69-92.
Miteva, D. 2013. Forests & Context: Factors shaping the effectiveness of Indonesia’s protected areas. Presentation at the 26th International Congress for Conservation Biology. Baltimore, MD. 23 July.
October 9, 2013. The views expressed above are the author’s and should not be taken as those of SNAP or its member organizations.