Universal education is key to enhanced climate adaptation

See allHide authors and affiliations

Science  28 Nov 2014:
Vol. 346, Issue 6213, pp. 1061-1062
DOI: 10.1126/science.1257975

Over the coming years, enormous amounts of money will likely be spent on adaptation to climate change. The international community recently made pledges of up to $100 billion per year by 2020 for the Green Climate Fund. Judging from such climate finance to date, funding for large projects overwhelmingly goes to engineers to build seawalls, dams, or irrigation systems (1). But with specific projections of future changes in climate in specific locations still highly uncertain, such heavy concrete (in both meanings) and immobile investments that can lock countries into certain paths may not be the best way to go (2). Our new study suggests that it may be efficient and effective to give part of this fund to educators rather than engineers. Public investment in universal education in poor countries in the near future should be seen as a top priority for enhancing societies' adaptive capacity vis-à-vis future climate change.

Recent research suggests that general empowerment of populations through universal primary and secondary education is not only essential to poverty alleviation and economic growth but also to reducing vulnerability to natural disasters (3, 4). It is not unreasonable to assume that factors that helped reduce vulnerability to floods, tropical storms, and droughts over the past decades will help reduce future vulnerability to climate change. We present findings from the most comprehensive global-level assessment of the effects of education on disaster fatalities (measured as the logged number of deaths per million of population) from hydro-meteorological hazards that are likely to be intensified by climate change, e.g., floods, droughts, storms, and extreme temperatures. The data cover 167 countries for the period 1970 to 2010. Data on disasters come from the Emergency Events Database (EM-DAT), which provides the best available information on the number of disasters and reported fatalities from around the world (5).

EDUCATE FEMALES, REDUCE FATALITIES. Because the literature on disaster vulnerability has conventionally emphasized economic growth while disregarding education, our statistical analysis focuses on the relative assessment of these two factors as measured by Gross Domestic Product (GDP) per capita and the proportion of women aged 20 to 39 with at least junior secondary education. The latter was shown to be a good indicator for recent improvements in human capital in other contexts (3).

To account for differences in the frequency of natural hazards experienced and size of the countries affected, we include as controls the number of registered disasters per population, total arable land area, a dummy variable for being landlocked, the recent rate of population growth to capture stress on infrastructure, and 43 regional fixed effects for countries with comparable settings and climate zones. As documented in the supplementary materials (SM) (table S1 and sensitivity analysis in table S2 and fig. S1), several alternative model specifications combined with different estimation techniques resulted in very consistent findings: When estimating the relative effects of income and education in the same models, GDP per capita turns out to be insignificant, whereas female education is highly significant across all models with the expected negative sign. Hence, this empirical analysis of national-level time series clearly indicates that female education is indeed strongly associated with a reduction in disaster fatalities.

Population pyramids by age, sex, and level of education.

Alternative scenarios for 2035: SSP3 (Stalled Development) on the left and SSP1 (rapid development) on the right. Data from (9).

Assuming that this robust association between education and lower mortality risk from natural disasters will continue in the future, we present alternative scenarios for future disaster-related fatalities as a function of alternative future education and population trends. When studying the effects of improvements in school enrollment on the human capital stock of the adult population, it is essential to account for significant inertia in the process of human capital formation. Because primary and junior secondary education tend to happen almost exclusively during childhood, it will take several decades until an expansion of education among children translates into higher human capital for men and women around age 50. This process of human capital formation along cohort lines can be appropriately modeled using the tools of multidimensional demography (6).

This approach has recently been applied to produce a new set of SSP (Shared Socioeconomic Pathways) scenarios for the international integrated assessment and vulnerability, risk, and adaptation research communities replacing the older Special Report on Emissions Scenarios which contained only total population size and GDP as socioeconomic variables (7). The SSPs were defined to address simultaneously the socioeconomic challenges to climate change mitigation and adaptation (8). Besides many other economic and technological variables, alternative population scenarios by age, sex, and seven levels of educational attainment for all countries form the “human core” of the full SSPs (9). SSP1 illustrates the case of rapid social development in all parts of the world associated with rapidly expanding education (see the first chart). SSP2 is the middle-of-the-road scenario where current development trends continue while SSP3 anticipates a fragmented world with stalled socioeconomic development. The figure also illustrates the great inertia of progress in improving educational attainment where, by 2035, the differences between the scenarios are only evident for the younger cohorts.

Predicted decadal number of disaster deaths (in millions).

Difference in deaths resulting from estimated education and population effects according to the contrasting scenarios SSP1 and SSP3 to 2100. See SM for details.

The results of combining the estimated coefficients (table S1) with two contrasting SSP scenarios (SSP1 and SSP3) for the rest of the century are shown in the second chart. We did this by taking the time-varying population and education variables from the respective SSPs. Different assumptions were made for the frequency of disasters representing possible greater future hazards. The solid lines in the second chart show the hypothetical case of constant hazard (i.e., no climate change). Under SSP1, this results in a significant decline of disaster deaths because of underlying progress in educational expansion. If we assume stalled development, which also implies higher fertility and thus higher population size, we observe almost no change under SSP3. The dashed lines assume an increase in the number of hydro-meteorological extreme events of on average 10% per decade (Climate Change +10%). Although there is still a slight reduction in future disaster deaths for SSP1, we observe a strong increase according to SSP3. The more extreme assumption of the hazard increasing on average by 20% per decade (Climate Change +20%; dotted line) leads to an increase in future disaster deaths in the longer run for all SSPs, although to different degrees.

COGNITIVE CAPACITY, SOCIAL SPILLOVER. Our macrolevel finding that education reduces disaster-related mortality is consistent with evidence from recent empirical studies for different parts of the world and at different levels of analysis (from individual-, household-, and community-level to global-level data). These studies demonstrate that education contributes to vulnerability reduction and adaptive capacity enhancement in the predisaster phase and during disaster events and the disaster aftermath [for review, see (2)].

Before a disaster, disaster mitigation efforts like living in low-risk areas or undertaking disaster preparedness measures, such as stockpiling emergency supplies, are found to be greater among more highly educated individuals and households (10). Similarly, loss of life, injury, morbidity, and physical damage from natural disasters were reported to be lower in communities and countries with a higher proportion of populations with at least a junior secondary education (11). The better educated were also found to cope better with both income loss and the psychological impacts of natural disasters (12). Most of these studies explicitly compare the effects of education to those of household income with education consistently emerging as more important. Given such systematically strong associations and a sound causal narrative described below, there is firm ground to assume functional causality of the effects of education on reducing vulnerability. This implies that a continuation of this association in the future can be reasonably assumed.

One important mechanism through which education influences human well-being is neurocognitive development. Learning basic literacy, numeracy, and abstraction skills enhances cognitive capacities through raising the efficiency of individuals' cognitive processes and logical reasoning (13). Accordingly, because preventive action is initiated by stressors, such as perception of risk, followed by assessments of one's ability to respond to the threat, the more educated tend to have greater risk awareness because of better understanding of the consequences of their actions, e.g., as found in the case of smoking and cancer prevention (14). In addition to these individual-level effects, there are also spillover effects of education at the community level as is evident for the effect of female education on lowering infant mortality (15). Opportunities of social interaction with more-educated members may speed up the diffusion of information and knowledge, or access to institutions that favor disaster risk reduction.

Of course, in our study the association between educational level and disaster vulnerability has only been estimated on the basis of the past 40 years and can change in the longer-term future because of all kinds of uncertainties. Instead of assuming different percentage changes in the hazard as we did, more differentiated global climate models could be applied. But our calculations show a clear picture of the strong effects of empowerment through education on reducing disaster vulnerability and enhancing adaptive capacity to climate change, which is unlikely to change when using more sophisticated models. Accordingly, given uncertainty about the precise manifestations of climate change in specific areas, it seems beneficial to increase general flexibility and enhance human and social capital in order to empower populations to better and more flexibly cope with climate change in a way best for their long-term benefit.

References and Notes

  1. Acknowledgments: The work leading to this paper was in part funded by the European Research Council Advanced Investigator Grant on “Forecasting societies' adaptive capacities to climate change” (ERC-2008-AdG 230195-FutureSoc).
View Abstract

Stay Connected to Science

Navigate This Article