Reconciling economic needs, including human health, the wellbeing of future generations, and planetary ecological health, poses exacting social and political trade-offs, which some have termed a “diabolical policy problem” (Steffen, 2011). This issue has generated much debate, but what has been less controversial is that caring for planetary health enhances human health (WHO, 2022; UNFCC, 2022). It is universally assumed that healthy humans follow healthy ecosystems: robust forests, and clean air and water (UNFCC, 2022). Contrarians argue that poverty itself pollutes in many insidious ways, so that policy agendas should prefer present poverty reduction over future ecological health, especially if the former is the best path to the latter (Lomborg, 2018). Indeed, development economists urge governments to exploit natural resources wisely so as to accelerate development and increase human wellbeing (Hausmann, 2022). This reasoning is considered tenable only if the intensity of harvesting resources does not degrade human health. We examine the following basic question: To what extent, if any, does extraction of natural resources and ecosystem services affect human health? We assess to what extent the use of environmental resources have formed human capital measured in terms of Health-Adjusted Life expectancy (HALE), a good indicator of human capital stock (Chiappini et al., 2020).
The Global Burden of Disease (GBD) data on Health-Adjusted Life Expectancy, a multi-dimensional measure of population health, (GBD 2019 Demographics Collaborators, 2020) is one of the best ways of measuring available human capital. It measures wellbeing as life years lived free of disease and injury at the national-state level over the period 1990–2019. On these data we regress a set of indicators measuring natural resources exploitation and ecosystem health. The results suggest that there is little effect of resource extraction on HALE, except that mineral rents per GDP associates with higher HALE. Ecosystem services too generally do not matter, but there is evidence to suggest that lower availability of biome associates with better human health, suggesting that human health is generated by factors quite independent of available biodiversity and protected area. Per capita income levels show the most robust relationship with healthy life expectancy, as does population density, results generally at odds with neo-Malthusian explanations about people, planet, and human wellbeing. Our results, taken together, support those who argue that development is likely both to generate better human health, and perhaps as a result, reduce local environmental harm in the longer term (Lomborg, 2018; Wendling et al., 2020).
Theory
Assessing the effects of environmental factors on human health is not straightforward. The causal nexus of resource extraction and public health outcomes is not easily disentangled, even in industrialized countries where data is more comprehensive. First, natural resources extraction: their availability enables society to harvest them for economic benefit, thus increasing population wellbeing (health, longevity, reduction of morbidity). This process enables access to finance for all purposes that society values, from good health to better environmental standards and even decarbonization (OECD, 2009; Lyatuu et al., 2021; Hausmann, 2022). But this is not how it is observed to work out even in well-studied rich-world contexts. Note some puzzling examples from the USA. Energy-intensive production areas sometimes have healthier-than-average populations and sometimes not. Plaquemines Parish in Louisiana is among the healthiest in that State, despite hosting large oil refineries. San Juan County, New Mexico, whose economy is wholly dependent on coal mining, ranks mid-pack in public health. Logan County, West Virginia, equally dependent on coal mining, ranks health-wise near the bottom among that State’s counties. Campbell County, Wyoming, while hosting a thriving natural gas sector, ranks among the healthiest in that State (Beckfield et al., 2020b). Other cases support a link between natural resource extraction and health, like the contiguous States of Texas and Louisiana: life expectancy at birth has stagnated overall, abnormally for a developed region, since 1958 up to 2018. Both States have large fossil fuel sectors and petrochemical refineries that contributed much to the public budget without improvements in the mortality rates (Beckfield et al., 2020a). In sum, natural resource extraction does not straightforwardly result in better or worse health outcomes, even in a country like the US, which is well studied.
Causal claims regarding this nexus in poorer countries abound; they lose the benefits of resource extraction because of high corruption, environmental degradation, the greed of rich and powerful business interests that exploit the poor, including their dependency on high-consumption markets in the rich world (Homer-Dixon, 2000; Andre and Platteau, 1998; Røpke, 1994). The neo-Malthusian, neo-Marxist claims are emblematic and have persisted through the globalization debates of the 1990s and 2000s (Held and McGrew, 2007; Stiglitz, 2019). Neo-Malthusians expand the scope of Thomas Malthus’s stricture, that geometric population growth must outrun arithmetically growing food production, to claim that a geometric rate of technological innovation may offset the outrun. But this entails a geometrically growing consumption of natural resources and the pollution thereof, – set against a static if not, indeed, an arithmetically declining “carrying capacity of the planet” (Mellos, 1988). As neo-Malthusians argue, poor countries are caught in a Malthusian trap because shrinking resources impose limits to growth (Meadows et al., 1993; Daily and Ehrlich, 1996). Neo-Marxists likewise expand Marx’s claims, saying that global capitalism exploits both labor and nature:
Capital accumulation and concentration are … based … also on … the “subsidies” provided by the undervaluation of genetic resources, natural goods, and ecosystem services, e.g. free access to genetic resources to be prospected, oil and gas to maintain overcapitalised agriculture and urban services, etc. (Leff, p.127)(Leff, 2000).
Poor countries’ exploitation by global capitalism and free-market structures result in payment outflows to the foreign-owned conglomerates that dominate the extractive industries (Bornschier and Chase-Dunn, 1985; Cardoso and Faletto, 1979; Leff, 2000). The net result is “immiserizing [sic] and illusory growth” that leaves behind residual pollution and depleted natural capital (Fofak, 2022; Suckling et al., 2021).
Another body of literature focuses on the so-called “natural resource curse” (van der Ploeg, 2011), usually blaming poor country institutions and governments for neglecting their populations.(Mehlum et al., 2006; Ross, 2012) In contrast to resource-rich industrialized countries, which some claim have sprung this trap (Larsen, 2006), a poor country, if it possesses enough natural resources, fails to develop the institutional and political conditions required for development through productivity growth (Mahdavy, 1970; Beblawi, 1987). If resource wealth, thus, is lost to a population because of corruption and mismanagement, the net outcome for local populations could be environmental degradation and resources depletion, with little lasting good for society (Ali and Abdellatif, 2015; Ross, 2012).
A significant minority of scholars contest the “resource curse” thesis (Brunschweiler and Bulte, 2008; Cotet and Tsui, 2013). Measuring the timing of oil discoveries and size of initial discoveries in place, Cotet and Tsui(Cotet and Tsui, 2013) found “little robust evidence of a negative relationship between oil endowment and economic performance, even after controlling for initial income”; instead, they found oil abundance associates with population growth, hinting at “a Malthusian effect which reduces the economic growth measured in per capita GDP”, possibly because positive health outcomes increase population. For many, the resource curse mainly involves extractive industries. There is evidence that mineral extraction, however, matters less than oil and gas (Ross, 2012); even so, conforming with the mainstream resource curse literature, we hypothesize that:
H1. Dependency on natural resource extraction adversely affects health-adjusted life expectancy.
Unlike the rich literature on natural resource extraction, few studies investigate the relationship between ecosystems services and human health. Several recent meta-analyses have found that significant flaws of scientific methodology often typify the field; for example, “anecdotal evidence is frequently used to support or refute particular positions or hypotheses” (McKinnon et al., 2016). And the evidentiary limitations on robustness and rigor are so severe that generalization thereon in the state of the art cannot reach reliable conclusions about the socio-economic impacts of biodiversity protection, positive or negative (Bowler et al., 2010). One difficulty is that some environmental services have positive effects on health, e.g. pollination, soil protection and fertility, pest control, fresh water, but others have negative effects, e.g. pests and diseases, allergens (Guo et al., 2022).
The best results so far obtained are from studies that relate biodiversity to human health. This matters to this study because biodiversity underpins all other ecosystem services (Cardinale et al., 2012). Sandifer et al.(Sandifer et al., 2015) identify more than 20 physiological (health) benefits from "immersion” in nature, including faster healing/recovery from surgery/illness/trauma, reduced cardiovascular and respiratory disease, and increased parasympathetic nerve activity (Sandifer et al., 2015). Biodiversity contributes to a more varied, balanced human diet and supplies it with micronutrients; to the production of marketable drug products (Veeresham, 2012); to protection from the emergence of new diseases (Wilkinson et al., 2018) and to the exposure of the human organism to beneficial bacteria, which furnishes us resistance to allergens (Ruokolainen et al., 2017).
We make a rough gauge of ecosystem services, estimating data measuring degree of biome protection and access to tree cover, wetlands, and grasslands (Wendling et al., 2020). Ecosystem health is also gauged by available biome protected. If the natural environment directly affects human health, these special biophysical factors should have a measurable impact on human health; thus, we posit a second hypothesis:
H2: Access to renewable resources measured in terms of biophysical assets increases healthy life expectancy.
We do not believe that possessing natural assets automatically improves human health, but it is more plausible that translating environmental assets into physical assets (capital) helps. We make no causal claims, as our estimation strategy is to ascertain statistical associations between ecological factors and health outcomes, relying on the above-cited theories to interpret findings. Causal analysis would require, for example, an instrumental variable approach whereby variable of interest x is instrumented with exogenous variable z that causes x but does not directly explain y. Such instruments are hard to find and often inadequately satisfy the instrument-exclusion (exogeneity) assumption that z is unrelated to y. Environmental factors could be instrumented with climatic zones, but climatic factors also directly affect human health, violating instrument-exclusion. A recent study used world-market commodity prices as an instrument to explain natural-resource rents and found that resource extraction increased life expectancy (Lyatuu et al., 2021). These authors neither address potential bias from violation of the exogeneity assumption, nor do they provide statistics testing for it. Although our dependent variables are different, we do attempt some replication of these interesting findings. Our main identification strategy, however, is to match theoretical expectation with observed association, so as to reject or confirm the hypotheses.