Adapting Natural Resource Enterprises under Global Warming in South America: A Mixed Logit Analysis

The planet Earth has been warming during the past century and will continue to warm in the coming centuries.¹ A warmer temperature in this century is highly likely to hurt low-latitude developing countries, even though it could impose only moderate damage to temperate developed nations.² Low-latitude developing countries are so vulnerable because they are already located in a hot climate and rely heavily on sectors that are highly climate sensitive, such as cropping and livestock management.³ For example, these sectors account for more than a third of the economy and employ more than two-thirds of the population in agriculture-based countries.⁴ Consequently, these countries will be seriously harmed by climate change unless prudent actions are taken to adapt to climate shifts by selecting heat-tolerant species and agricultural systems.⁵ Without adaptations to climate change in these sectors, existing and future economic development programs in the region will turn out less effective.⁶

Researchers find that individuals are likely to adjust current practices in an effort to adapt to climate change. Farmers in Sri Lanka will make dramatic shifts toward commercial tree plantations such as coconut, rubber, and tea as global warming reduces paddy rice production sharply.⁷ African farmers will adopt an integrated system of farming and livestock species such as sheep and goats more frequently when climate becomes hot and dry.⁸ As climate change progresses, the ranges of most tree species such as white oak and paper birch are projected to move north, while the range of sugar maple will shift out of the United States entirely.⁹ However, these studies provide only a partial analysis focused on a single sector, while an integrated assessment of natural resource sectors is rare, if not nonexistent. Although most researchers focus on crop vulnerabilities, livestock management accounts for as much of total agricultural income as crops, while forestry income accounts for 22 percent of the total income of rural households in developing countries.¹⁰

From this background, this paper develops an integrated adaptation model of natural-resource-intensive enterprises including crops, livestock, and forests and encompassing both specialized and diversified enterprises. Specialized enterprises, focusing on crops only, livestock only, or forests only, and diversified enterprises, combining crops and forests, crops and livestock, or crops, livestock, and forests, are modeled to explain the current sensitivities and future shifts in the choices of these enterprises when climate is changed. In contrast to Africa, South America is an ideal place for such a model since it relies heavily on livestock and forests in addition to crops.¹¹ Argentina and Brazil are the world's largest exporters and consumers of beef cattle.¹² The region is heavily forested in the north, has extensive grasslands in the southern plains, and features high mountains along the Andes range.¹³ It has the world's largest rain forests in the Amazon, and its pasturelands are four to eight times larger than the cropland in major countries.¹⁴

This paper analyzes a rural household's adoption of one of these enterprises located across South America's varied climate zones. The surveys are collected from about 2,300 South American rural households in seven countries across the continent. Using a mixed logit model, this paper estimates random coefficients with normal distributions and calculates the probabilities of different enterprises to be adopted by simulation methods.¹⁵ Actual household choices are explained by climate factors, after controlling for other factors such as soils, market access, household characteristics, and country-specific effects. Based on the estimated parameters, this paper predicts dynamically expected changes in enterprise adoptions under multiple climate scenarios for 2020 and 2060, based on atmospheric-oceanic general circulation models (AOGCM).

The paper proceeds as follows. In the next section, I describe a mixed logit model that is applied to the analysis of the choice of natural resource enterprises under climate constraints. The subsequent section describes the data used in this study and their sources. The ensuing sections report estimation results, followed by simulation results for the next half century...