MSc Development Studies alum Alejandra Padin-Dujon takes a look at parametric insurance, a model designed to assist vulnerable countries with the financial burden of climate disasters, and contemplates why it isn’t delivering on its promises.

“Why is a raven like a writing desk?” asked the Mad Hatter. My riddle: “Is parametric insurance worth the hype?” If you are familiar with Alice in Wonderland, you will recall that the Mad Hatter hasn’t a clue, and neither do I. Nevertheless, time goes on, international development specialists peddle their takes, and the promise and peril of parametric insurance each vie for advantage. This is the confused world of insurance-based disaster risk management in the age of climate crisis, where the world’s poorest and “least developed” countries—the lowest historical emitters—shoulder the lion’s share of climate disaster. Is parametric insurance the hot, not-so-new thing: the “old news” yet in its infancy, portending great things for climate resilience? Or is it a red herring, with high premiums and insufficiently tailored data and tools, stealing international attention away from structural change?

What is parametric insurance?

The concept of parametric insurance is simple, on paper. It’s an alternative to traditional indemnity insurance: the kind that looks at damages before calculating payouts. Instead, parametric insurance creates an index using climate projections and asset valuations before disasters happen, deciding preemptively what intensity of disaster (decided, for example, by wind speed, storm surge, or scarce rainfall) will result in what level of compensation. Then, when disaster strikes, insured governments, communities, or individuals receive payouts aligned with the intensity of the extreme event and the value of the assets they have elected to insure. Simple.

Arguably, the simplicity of the concept has helped it gain traction with international financial institutions, international climate bodies, and many governments. Add to that its technical appeal: instead of months of costly and delayed damage assessments, insured populations (should) get quick disbursements; instead of requiring an army of damage assessors, the index (ideally) does nearly all the work and conserves government capacity. Of course, the story is more complicated than that—but we’re getting there.

Politics

Parametric insurance is not new: environmental indicator-based insurance schemes date back to the early twentieth century. Nonetheless, it truly picked up steam in the late 1990s as academic economists and World Bank employees cast new eyes upon these prior designs, looking for ways to craft insurance contracts that would minimize moral hazard (that is, the risk that beneficiaries will rely too heavily on their benefit, lacking sufficient incentive to make prudent choices). Simultaneously, the World Bank was reconceiving poverty as vulnerability to shocks (including weather shocks) rather than a static state—so, it was time to invest in ex ante resilience building. The World Bank then went on to implement experimental parametric insurance throughout the 2000s, at the individual scale all the way through to the sovereign.

Simultaneously, international climate bodies like the United Nations Framework Convention on Climate Change (UNFCCC) adopted the principle of risk sharing and insurance as a critical tool in the fight against drought, desertification, flood, and other catastrophes, especially in developing countries. This principle, calling for “risk insurance facilities, climate risk pooling and other insurance solutions” was enshrined in the Paris Agreement (2015).

The road to the inclusion of insurance in the Paris Agreement under Article 8 on loss and damage (L&D) is itself an interesting story. “Loss and damage” can be defined as present and future impacts of climate change that are not avoided through emissions reduction or adaptation. In the early 1990s, the Alliance of Small Island States began pushing (unsuccessfully) for the inclusion of L&D in international climate agreements, citing heightened and immediate climate vulnerability. While L&D was originally conceived as social protection, international insurance agencies took up the rallying cry, and what eventually ended up in the 2013 Warsaw International Mechanism for Loss and Damage—the immediate precursor to the Paris Agreement’s stance on L&D—included a risk management message that opened the door to new insurance possibilities. Thus, L&D is generally assumed to mean one of two things today: either ex post official development assistance (aid), or ex ante insurance.

Promise

Early World Bank-led iterations of parametric insurance largely followed in the vein of microfinance: small-scale contracts targeted at the individual. To recap, in rough terms: poverty was not so much about static states introduced by disadvantageous political and economic systems, as about the incidental precarity of an individual’s economic position in the face of shocks, including extreme climate events. (If you’re confused how this precarity could be disentangled from the many bigger forces that shape individuals’ lives, I’m right there with you.)

Since the early 2010s, there has been a shift away from micro-level parametric insurance toward bigger pools—especially sovereign risk pooling. The logic, in a nutshell, goes like this: at any given time, most countries should be disaster-free, which dilutes the overall risk profile of the multi-country pool. In other words, while all countries in the pool should be paying premiums, it is highly unlikely that many large payouts will need to be made at once, which would deplete the insurer’s coffers, making insurance unattractive to the insurer and expensive to the insured.

This risk dilution runs in stark contrast to micro-level parametric insurance. When northern Kenya is hit by drought, as it has been this year, practically no one in agricultural communities emerges unscathed. There is no one to balance out the risk, everyone draws on their insurance plan, and premiums skyrocket. Sovereign risk pooling—as can be seen in the Caribbean (CCRIF SPC, formerly the Caribbean Catastrophe Risk Insurance Facility, est. 2007), Africa (African Risk Capacity, est. 2012-14), and the Pacific (Pacific Catastrophe Risk Insurance Company, est. 2016)—is a popular solution.

Peril

Two key structural issues pose a threat to the efficacy of parametric insurance, whether in micro or macro form: basis risk and unaffordable premiums. These exist in addition to other notable critiques: for example, parametric insurance as a distraction from the need for more structural change in political and economic systems; the dangers of treating climate vulnerability as a latent quality of communities rather than a built-in design flaw of economic planning (“it’s a feature, not a bug”); discrepancies between initial and updated value-for-money projections; etc. Parametric insurance should arguably be just one ingredient in holistic disaster risk reduction recipes—nevertheless, I will focus on the technical design weaknesses of parametric insurance.

Tailoring indices to real contexts and projected damages is a difficult process made even more complicated in many cases by imprecise modeling and insufficient historical data. Basis risk here, roughly speaking, is the chance that, despite attempts to calibrate premiums in accordance with projected climate-related damages, there is a mismatch between real damages and payouts when the time comes to disburse. Payouts might be too little, or even nonexistent (never triggered). In practice, it is tough to say whether the mismatch is systematic and therefore indicative of a deeper problem in the index, or whether it’s simply random—but insured countries do not need this kind of evidence to feel the devastation of paying premiums that yield nothing when funds are needed most.

In 2015, Kenya experienced an especially bad year of drought. Despite its membership in the African Risk Capacity (ARC) pool, no payment was triggered, and the country suffered for it. Was this mismatch random? Looking at one event, it’s impossible to say. Regardless, after 2015, it became politically infeasible for Kenya to stay in the ARC premium-paying pool, and the country dropped out after the 2015-16 fiscal year. It has never rejoined, although it expressed some interest in 2020.

Additional challenges include affordability: the premiums linked to impactful payouts, especially in the regions that ostensibly need insurance the most due to heightened climate vulnerability, are just too high. Insurance in these contexts is an expensive business. This cost has led to a precedent of shaky aid flows subsidizing premiums. The World Bank does this with the COAST pilot initiative for fisherfolk in the Caribbean, and the UK Foreign, Commonwealth, and Development Office does this with the ARC in Africa, to name a couple.

Conclusion

There is no real conclusion to this saga, except to recognize that, just like loss and damage, parametric insurance isn’t going anywhere anytime soon. It has captured the imagination of international financial institutions, the UNFCCC, and many governments in both the Global North and Global South. Is the question, then, how to make it more effective? Should we be shifting the narrative to bigger-picture solutions? Do both? I couldn’t tell you. Why is a raven like a writing desk?

A special thanks to Prof Leigh Johnson at the University of Oregon for an in-depth conversation on the topic.

The views expressed in this post are those of the author and do not reflect those of the International Development LSE blog or the London School of Economics and Political Science.

Image credit: Drought in Kenya’s Ewaso Ngiro river basin, 2017 via Climate Centre on Flickr.