Climate change, biodiversity loss and the unsustainable use of natural resources are rightly at the centre stage of global policy debates and interventions today. Yet these issues are not just a matter for environmental policies, as interventions across all sectors can have significant (often unintended and negative) effects on the environment.

At the closing session of the recent 4^th Conference on Evaluating Environment and Development organized by the Global Environment Facility’s Independent Evaluation Office, I had the opportunity to address the question that evaluators are now asking themselves: How and to what extent should we consider the environment in our evaluative work?

Environmental sustainability and poverty alleviation

There is a strong argument in favour of mainstreaming and integrating environmental sustainability in our evaluations. After all, the sustainable use of natural resources is key for development. This is reflected in the World Bank Group’s expanded mission not simply to end poverty, but to end poverty on a liveable planet.

Human life depends on being able to draw from natural systems. Our forests alone sustain the subsistence and livelihoods of hundreds of millions of people worldwide. We cannot achieve poverty alleviation without addressing environmental issues, and we know from our work that there are win-win solutions in terms of achieving both environmental sustainability and poverty alleviation.

It is therefore critical that we integrate environmental considerations in our evaluations at the Independent Evaluation Group, and we have made strong progress in adapting our approaches to this end. But we need to ensure we do this in a meaningful way.

Ultimately, we should think carefully about how evaluations can help decision makers and stakeholders to understand how and under what circumstances specific policy interventions can generate beneficial outcomes – both for development and the environment.

Targeting our environmental impacts with tailored approaches

In my view, we should mainstream the environment in our evaluations in a strategically selective and rigorous manner. Rather than making environment a default (but peripheral) issue in every evaluation, we should concentrate our impact by asking questions such as:

• In which areas of policy interventions can we expect the highest potential negative unintended environmental effects?
• Where is the highest potential for generating new knowledge?

This will make our evaluative work as rigorous and influential as possible, and nudge policies and programs onto a more sustainable pathway for development and environmental sustainability.

Three priorities for future evaluations

I see three priorities for strategically integrating environmental considerations into IEG’s evaluations.

The first is to adopt a systems approach in evaluation, when relevant.  Environmental change processes are complex. They are non-linear, uncertain and can involve irreversible tipping points. Looking at things from a systems point of view helps us get a fuller and more accurate picture of how change happens. It considers various factors that cause change, not just changes in policy. Moreover, it also considers the broad range of changes (economic, social and environmental) affected by policy interventions as well as the interlinkages between them. Doing this well requires the use of better theories of change and longer time frames in evaluation.  It also requires a change in our mindset and skills as evaluators.

Second, we must continue to increase our use of new data and data science techniques in our evaluations. We have already found that this expands our ability to generate more rigorous evidence on the relationships between policy interventions and environmental and socio-economic outcomes. Technology and new data can improve our evaluations by helping us to understand the evolution of geospatial phenomena such as land use conservation, pollution and flood resilience, as well as poverty and inequality.

Finally, we need to expand the scope of our data collection and analysis and better take into account different views and voices from stakeholder groups involved in or affected by the interventions we evaluate. We should broaden the range of experts that we talk to and reach out to more stakeholders and experts from civil society and academia. We must include voices of indigenous communities, business owners, farmers, local government officials and all stakeholders whose livelihoods are affected by environmental changes. The voices of women must also be included, as environmental degradation has a significant impact on women, particularly in regions where they are heavily reliant on natural resources for their livelihoods.

This will not be easy, but we should aspire to achieve it. Getting there will not only be a matter of employing the right methods or providing the right incentives, but it may also require us to rethink some of our evaluation modalities and the scope of our exercises.

But the effort will be worth it because, if independent evaluators implement these changes, they will deliver better evaluations and recommendations. These will benefit policy intervention design, implementation and ultimately impact, not only on global development, but our natural world.