Index-based livestock insurance in Mongolia

Index-based livestock insurance in Mongolia is a program that pays herders on livestock mortality, counted by district and species. Losses are cut into layers split between herders, insurers and the government. Winter 2009-2010: a dzud killed on the order of eight million head in Mongolia, close to a fifth of the national herd. A dzud is the mix of summer drought, deep snow and prolonged cold that keeps animals from reaching the grass. The country had planned for this. Since the mid-2000s a pilot run with the World Bank, the Index-Based Livestock Insurance Project, had been covering herders in a few provinces. Its index is neither rainfall nor vegetation: it is observed mortality, by species, in each soum, the rural district. The figure comes from the annual livestock census run by the national statistics office, a source that goes back decades and that nobody can game. If goat mortality in a soum crosses the trigger, every insured herder in that soum is paid on his goats at the declared value per head. The layered structure is what sets the program apart. Herders keep small losses, a few percent. A middle layer, up to roughly 30% mortality, is sold by commercial insurers whose premiums are pooled so that one company does not go under in a single winter. Above that, the government takes the catastrophic loss, backed by a World Bank contingent credit line. The difference with Kenya's IBLI (Index-Based Livestock Insurance) is one of principle: Kenya predicts mortality from NDVI (Normalized Difference Vegetation Index), Mongolia counts it. Basis risk is lower, since the index measures the loss itself. But the money arrives after the count, often months after the late-winter die-off, too late to buy feed and only in time to restock. By the mid-2010s the arrangement had moved from pilot to a national program.