Highlights from another significant year for GIF can be found in our latest impact report.

Apply for funding

The impact of adaptation and the adaptation of impact

By Ken Chomitz, Chief Economist  |   Posted 26th May 2022

GIF wants to maximise its social impact. We also want to support climate adaptation. Are those goals compatible? Can we measure climate adaptation impact on the same yardstick we use for all our investments?

We think so. For us, the main purpose of climate adaptation is to make sure that people – in particular, those living on less than $5 per day – are better off in the face of an increasingly harsh and fickle climate. We look carefully at how our innovations help people, rain or shine, flood or drought. Resilience innovations can keep people from spiralling into poverty when climate shocks hit. Adaptation innovations help them to tailor their livelihoods and activities to the current and future weather patterns. The premise behind our climate strategy is that adaptation and resilience (A&R) innovations have been overlooked and underfunded. By supporting such innovations, we expect to find powerful ways of boosting people’s well-being.

GIF uses its Practical Impact methodology to assess the potential impact of any innovation. Practical Impact recognises that innovations are inherently risky, and that it takes a decade or more for a truly impactful innovation to reach its full scale. Therefore, our forecast impact measure is:

Impact =

Estimated probability that the innovation will successfully scale up

X

Number of poor people benefited 10 years from now if the innovation successfully scales

X

Expected % improvement in a typical beneficiary’s living standard

The long-term relationship between risk and social benefit is baked into this metric. This carries over naturally to climate adaptation, where impact depends on how weather unfolds, day to day, season to season, and even decade to decade. For example, how would the innovation protect people when the storm hits? What is the impact when the weather is fair? And what is the likelihood of stormy weather, now and as climate continues to change?

Risk assessment like this could be used to forecast impact for many types of adaptation innovations. Consider this stylised example, based on two studies (Dar et al [1] and Emerick et al [2]) of the impact of introducing a flood-resistant rice variety to semi-subsistence farmers in a flood-prone region of Odisha in India. The improved seed had been shown in test farms to be just as good as the usual variety in normal years, but much more resilient under flood conditions. Researchers conducted an experiment to gauge actual farmers’ experience, randomly selecting a treatment group to receive the new seeds, for comparison with a control group.

Fortunately for the researchers, but unfortunately for the farmers, the first year of the experiment was a flood year. As expected, the treatment group fared better than the control group, with a significantly higher rice yield. Based on that finding, we can say that the new seed improved farmers’ resilience to extreme climate events.

But that was not the end of the story. The following year had more normal weather. Based on test plots, one would not have expected any difference in yield between improved and standard seeds. But in fact, the farmers using the improved seed realised much higher yields. The researchers explain this as a story of adaptation. Farmers using the old seed were reluctant to invest much in fertiliser or in labour-intensive planting methods. Those upfront expenses might be wasted in the event of a flood. In contrast, farmers with the new seeds now faced a lower risk of losing their crop, so they were willing to invest more in fertiliser and in labour-intensive planting. In addition, they didn’t have to stockpile as much rice as a hedge against a failed crop, so they were able to sell a bit more and invest it in the farm. In short, the adaptation story goes beyond resilience to extreme events – it encompasses far-reaching changes in behaviour and outcomes through the entire range of climate possibilities.

Building on this, a simple risk-adjusted impact forecast might look like this:

Impact =

flood probability*outcome during a flood year

+

Non-flood probability *outcome during a nonflood year

This simple formulation encourages a holistic look at how people adapt their lives to climate threats and how innovations might help them. It is feasible for the wide class of adaptation challenges for which there is some ability to assign probabilities to climate outcomes. Of course, it is a feature of climate change that the probabilities are changing, particularly for the most extreme events. Nonetheless, insurers and engineers are grappling with ways to quantify climate risks, and service companies are arising that can provide model-based estimates of these risks. In the Odisha study, researchers were able to use 11 years of detailed, publicly available, remote sensing data to estimate an average annual probability of flooding at 19%. Importantly, their study found that poorer, more marginalised farmers were at greater risk – a fact that can be built into a poverty-focused assessment of impact.

As GIF builds up its climate adaptation portfolio, we look forward to adapting our impact approach to meet the challenge of climate adaptation. This approach won’t directly apply to everything we do – for instance, it is more difficult to apply to capacity-building investments. And we certainly expect to iterate and improve over time. But we hope it will hone our ability to find and support the adaptation and resilience innovations that are desperately needed. We hope it can enliven ongoing discussions about how to quantify investors’ contribution to global adaptation.

We would love to hear thoughts from readers on this blog or the accompanying paper. Please send comments, suggestions and/or questions to analytics@globalinnovation.fund with the subject: Climate Adaptation Metrics.

[1] Manzoor H. Dar et al., ‘Flood-Tolerant Rice Reduces Yield Variability and Raises Expected Yield, Differentially Benefitting Socially Disadvantaged Groups’, Scientific Reports 3, no. 1 (22 November 2013): 3315, https://doi.org/10.1038/srep03315.
[2] Kyle Emerick et al., ‘Technological Innovations, Downside Risk, and the Modernization of Agriculture’, American Economic Review 106, no. 6 (1 June 2016): 1537–61, https://doi.org/10.1257/aer.20150474.