The following is the established format for referencing this article:Rigal, S. 2022. Social-environmental index: combining social and biophysical indicators reveals limits to growth. Ecology and Society 27(2):33.
ABSTRACTThe negative impact of the dominant socioeconomic paradigm on the biosphere, on the climate, and on societies themselves is acute. Yet, the success of countries is measured by indicators known to be limited because they target a socially attractive but environmentally unsustainable model of society. A myriad of indicators have been proposed to address this lack of relevant measurements that assess the real social achievements of countries. At the same time, the impacts of human societies on life and climate have become increasingly well monitored and have shown the ecological deadlock of the dominant development model. Although social and environmental thresholds have been highlighted, combined indices that allow countries to track their trajectories toward greater social justice without exceeding biophysical thresholds are still lacking. A combined socio-environmental index (SEI) is constructed here to fill this gap. The relationship between SEI and gross domestic product (GDP), population density and sustainable development index (SDI) is then analyzed. This allows for a re-examination of the Easterlin paradox from a social and environmental perspective. In addition, considering population density allows to test the influence of population on the country’s sustainability success. It is shown that combining social and environmental thresholds into a combined index is not only feasible but provides a useful complementary tool to detailed and specific social or environmental indicators. SEI highlights and quantifies the limits, already exceeded in many countries, beyond which economic development is clearly related to a decline in social achievements and a crossing of biophysical thresholds. Unlike GDP, population density and population growth were not found as being related to the current unsustainable development model. Therefore, the results call for a degrowth in the environmental impact of high-income countries, which may result in social improvements and yield room of maneuver for the development of social foundations in other countries. All of these transformations require new narratives, goals, and measurement tools that can be partly provided by SEI.
The spread of the imperial mode of living (Brand and Wissen 2017) across the globe is responsible for the greatest damage to our planet (Ripple et al. 2017, Kolasi 2018, Marques 2020). Yet, this way of life, characteristic of Western countries, remains a target for most nations without considering the biophysical impossibility of extending its high environmental consequences to over 7 billion people (Ganivet 2020). However, the social achievements than can accompany the mode of living of Western societies, help to understand its growing success. Because the limits of our planet are finite and meeting the social need of most humans is a matter of justice and humanity, estimating whether and how it is possible to combine a good life for all within planetary boundaries is crucial (O’Neill et al. 2018).
The limitations of mainstream indicators for assessing human well-being are widely recognized, and for GDP in particular, they have been acknowledged since its introduction in national accounts (Kuznets 1934). Not only are non-market activities left aside, but it is based on an economic conception of well-being, i.e., the possibility of possessing more and more marketable goods. These limitations have led to the emergence of many other indicators aimed at encompassing the multiple facets of well-being. Among the most universally used, the HDI (Human Development Index) accounts for life expectancy and education in addition to GNI (Gross National Income, derived from GDP; Stanton 2007), and the GPI (Genuine Progress Indicator) corresponds to economic well-being adjusted by negatively accounting for social costs of economic activities (Talberth et al. 2007). Other happiness indicators or sociological surveys have also been developed to more accurately assess the “Easterlin Paradox” that highlights the decorrelation between GDP per capita and people happiness above given income thresholds already exceeded in Western countries (Easterlin 1974, Max-Neef 1995, Di Tella and MacCulloch 2008, Fanning and O’Neill 2019). However, few indices have been proposed to directly encompass human well-being or happiness, with what constitutes the foundation of all human life and economic activity, namely the preservation of functioning ecosystems on a livable planet (Abdallah et al. 2009, Hickel 2020).
To say nothing of maintaining any economic activity, the very possibility of living in some currently populated areas in the near future is in question because of the consequences of global warming (Raymond et al. 2020). In addition, the erosion of biodiversity deeply affects the daily lives and lifestyle of millions of people (Begotti and Peres 2020). More particularly, planetary boundaries (Rockström et al. 2009) are already partially exceeded by the imperial mode of living as highlighted by the tremendous differences between wealthy, mainly Western countries and others, even when scaling down to the individual level, i.e., comparing environmental footprint or carbon emissions per capita (Duro and Teixidó-Figueras 2013, Piketty and Chancel 2015).
Finding a path toward sustainability, therefore, requires evaluating social achievements in light of the environmental degradation produced by socioeconomic changes. In this sense, a recent analysis by O’Neill et al. (2018), proposed placing countries in a two-dimensional plane defined by social minimum bars and overcoming biophysical boundaries. This first breakthrough remained qualitative and based on results from the doughnut economy defining a space between social foundation and environmental ceiling (Raworth 2013). The environmental ceiling corresponds to the notion of planetary boundaries (Rockström et al. 2009) that remain debated in their measurements (can they be measured?) and definitions (do they correspond to tipping points?; Montoya et al. 2018a, 2018b). O’Neill et al. (2018) overcame the main criticism against boundaries, in particular in biodiversity, by using a measurable biodiversity boundary (Running 2012). Moreover, a further improvement added quantitative overshoot of planet boundaries (Hickel 2019) by accounting for the fact that some countries only slightly exceed biophysical thresholds while others are well beyond. This has resulted, in particular, in an amelioration of the HDI in a Sustainable Development Index (SDI; Hickel 2020), which takes into account a combination of social indicators (life expectancy, education, and income) and biophysical indicators (CO2 and material footprint) while remaining partially based on GDP. Such approaches have clarified the trade-off between countries with high social achievement but large environmental impacts and countries with lighter ecological impacts but low social achievement (Fig. 1a; Scherer et al. 2018). However, some countries are encouragingly escaping this valley of trade-offs. For instance, Vietnam has managed to achieve six social minimum bars while exceeding only one biophysical boundary (O’Neill et al. 2018).
Yet two-dimensional analyses (social vs. biophysical axis) remain difficult to use for cross-country comparisons or even for comparisons between years within a country. If GDP-based policies are to be abandoned, an index that is easy to manipulate must be provided. The use of two dimensional indicators could therefore be limiting for public policy, as they can hardly help to assess the improvement of a country over time. By using the information available in the two-dimensional indicators (one social and one environmental axis), it is possible to construct one combined index. This index can then be easily used in public policies to estimate the gap between the current state of countries and the goal to be reached, namely the achievement of all social standards within the limits of the planet, if this is ever possible. Yet merging two dimensions into one results in a loss of information. It also makes the assumption of substitutability between social achievements and biophysical thresholds considered as a weak sustainability hypothesis (O’Neill 2012). Therefore, the construction of a combined index must meet three criteria to be meaningful and useful (Meadows 1998): (1) retain as much information as possible from the initial indicators, (2) be used as a complement, not a replacement, to the set of social and biophysical indicators to ensure strong sustainability, and (3) define a target to reach.
The challenge is then to determine how to combine social and environmental information while losing as little information as possible and defining a target. In this study, a socio-environmental index (SEI) is proposed that is based on quantitative and qualitative information on social achievements and biophysical impacts. In addition to the qualitative information provided by threshold approaches (either below or above thresholds), it is also possible to consider quantitative data on social achievements and overshoot of biophysical thresholds. The robustness of this index and its relevance compared to SDI, the most advanced socio-environmental index, is then tested. Using the combined SEI as a complement to detailed social or environmental indicators, one finally questions three broad issues: (1) Are socio-environmental success and population density interconnected (the population question being controversial since the long-standing debate on the role of population to achieve sustainability [Ehrlich and Ehrlich 2009, Crist et al. 2017])?; (2) Does the “Easterlin paradox” exist when combining social and environmental indicators?; (3) Which social or environmental indicators determine the GDP threshold and does that threshold allow to reach a safe and just space for all?
Social and biophysical indicators of countries
The 11 social indicators (life satisfaction, healthy life expectancy, nutrition, sanitation, income, access to energy, education, social support, democratic quality, equality, and employment) were selected from the safe and just space framework proposed by Raworth (2012). The seven biophysical indicators (CO2 emissions, phosphorus, nitrogen, blue water, embodied human appropriation of net primary production (eHANPP), ecological footprint, and material footprint) were selected from the safe operating space proposed by Rockström et al. (2009).
All those social and biophysical indicators were downscaled by O’Neill et al. (2018), which provided their threshold (see details in Appendix 1) and values in 2011 (see details in O’Neill et al. 2018). Countries for which fewer than six social indicators reported (out of 11) or fewer than four biophysical indicators were reported (out of seven) were dropped from the analysis, leaving 151 countries with sufficient data to compute the spatial analysis.
In addition to the available dataset described above, social and biophysical values were reconstructed for each social and biophysical indicator at different times steps for each country based on the same sources (Krausmann et al. 2013, Lenzen et al. 2013, Helliwell et al. 2018, World Bank 2015). Because not all required data were available together for the same time period, with some years missing, even in the last two decades, the accessible information was averaged to obtain social and biophysical scores at three time steps (2005, 2010, 2015) for each country. For each indicator, the 2005 values correspond to the average of the values between 2003 and 2007, the 2010 values to the average of the values between 2008 and 2012, and the 2015 values to the average of the values between 2013 and 2017. Because of data limitations, this could not be done for all 151 countries but for 135 countries. This step allows tracking temporal changes in development pathways for the different countries.
Combined socio-environmental index
To account for both social minimum bars and biophysical limits, the social and biophysical scores were combined into one socio-environmental index (SEI). For each country, a biophysical score was calculated by combining each biophysical indicator (step 1) and a social score was calculated by combining each social indicator (step 2). The two scores were then combined to calculate the SEI (step 3). Hereafter, a score corresponds to an average value, either social or biophysical, whereas an indicator always refers to a given social or biophysical indicator (as presented in Appendix 1).
Step 1. The biophysical score was constructed by integrating the intensity of biophysical transgression while keeping all qualitative information from the number of biophysical thresholds overshot. To do so, the coordinate of each country c on the qualitative x-axis (Fig. 1b) was calculated as the number of overshot thresholds cTo divided by the number of recorded biophysical indicators cTr (Eq. 1):
with cx the coordinate of c on the x-axis, cTo the number of overshot biophysical thresholds by country c and cTr the number of recorded biophysical indicators for c.
Then, the coordinate of each country c on the quantitative k-axis (Fig. 1b) was calculated as the mean of the overshoot for all biophysical indicator values (the overshoot being the difference between the threshold and the observed indicator; Eq. 2). Each indicator was scaled by its threshold value to attribute the same weight to each indicator. Note that if a country reaches but does not go beyond several limits, ck = 0.
with ck the coordinate of c on the k-axis, ci the biophysical indicator values above thresholds and cTo the number of overshot biophysical thresholds by country c.
To attribute the same weight to social success and biophysical overshoot, x and k were combined into a single z axis. To do so the position of each country c(cx, cy, ck) was re-projected from the 3-dimensional space (x, y, k) to the plane (z, y). That is, coordinates cx and ck were combined in a new coordinate cz, which corresponds to the orthogonal projection of c' (c projected on (x, k), Fig. 1b) on the segment [BW]. This was done by transforming the coordinate system as follows (Eq. 3):
with cz the coordinate of on the z-axis, cx the coordinate of c on the x-axis, ck the coordinate of c on the k-axis, Bx the coordinate of B on the x-axis and Bk the coordinate of B on the k-axis.
Step 2. The social score (y-axis) was obtained by adapting the modification proposed by Hickel (2019). For the nutrition indicator, because exceeding it can lead to social and health problems, all values above the minimum bar (2700 kilo-calories per person per day) were set to one. For the other social indicators, the raw social values (i.e., not scaled by the minimum bar value) were used to obtain the distance of each indicator to its maximal theoretic value. Those maximal theoretic values were directly available because each indicator fell within a bounded interval. The only exception was healthy life expectancy for which the maximum observed value was used (Singapore = 75 yr). The social score was computed as the average of all distances between social indicators and their maximal values (Eq. 4). By doing so, quantitative information was obtained for the social score while avoiding the shortcoming of equalizing all indicators above minimum bar to one. This allows for more accurate assessments of actual social achievements. For instance, notable differences exist between countries with a democratic quality of 0.8 (corresponding to the minimum bar and close to the value for the United States) and democracy largely above 0.8 (e.g., in Scandinavian countries). The same goes for the other social indicators. For clarity purposes, the optimal position B was set at the origin (0, 0, 0) of the 3-dimensional space (x, (1-y), k; Fig. 1b) by defining c'y as one minus the social score:
with c'y the coordinate of c on the (1-y)-axis, cy the coordinate of c on the y-axis, cj the value of the social indicator j, cjmax the maximum of the social indicator j and cSr the number of recorded social indicator for country c.
Step 3. For a country, represented by c in Fig. 1b, the SEI value is based on the distance [Bc''] in the Euclidean space ((1-y), z). More precisely, the SEI value corresponds to the distance between c'', the projection of c on the plane ((1-y), z), and B, scaled by the distance [BW], W being the worst position (low social indicators and high number of largely overshoot biophysical thresholds). To obtain an index increasing when approaching B, SEI was defined as follows (Eq. 5):
with SEIc the SEI value of c, cz the coordinate of c on the z-axis, c'y the coordinate of c on the (1-y)-axis, Bz the coordinate of B on the z-axis and By the coordinate of B on the y-axis.
External socioeconomic indices
The current mainstream socioeconomic index for assessing a country’s success and estimating (economic) well-being, the gross domestic product (GDP), was compared with SEI. It should be noted that GDP is not directly present in the social performance indicators nor in the biophysical indicators (Appendix 1). The sustainable development index (SDI) corresponds to the most advanced index to account both for social and environmental scores (Hickel 2020) and represents a significant update of the human development index (HDI). However, it takes into account a limited number of social and environmental variables and remains, by construction, directly linked to GDP. The relationship between SEI and the HDI, which is the primary index proposed as a complement to GDP, is analyzed in Appendix 2. The genuine progress indicator (GPI) was not used because it was only available for 17 countries and not for all time periods considered (Kubiszewski et al. 2013). Finally, the relationship between SEI and population density was analyzed. UN temporal data GDP per capita are based on purchasing power parity (PPP) in constant 2017 international dollars (Int$; UN 2020a). World GDP was taken from the World Bank (2015), population density (from population data and country surface [World Bank 2015, UN 2020b]) and SDI (available at https://www.sustainabledevelopmentindex.org/) were used to calculate the averaged value for each time step (2005, 2010, and 2015) and for each country as described for social and biophysical scores.
Two separate analyses were conducted. First, the spatial pattern, i.e., the relationships between SEI and GDP, population or SDI with 2011 values was studied. To maximize sample size, the 2010 averages of GDP, population density, and SDI were used to analyze their spatial relationships with SEI. Next, a temporal analysis was conducted as the space-for-time hypothesis may mislead the interpretation of spatial relationships between SEI and economic activity, population or SDI (Damgaard 2019). The temporal pattern was thus analyzed, i.e., it was tested whether temporal changes in SEI were congruent in time with the link observed in space between SEI and GDP, population density or SDI. The 2005–2015 changes for GDP, population density and SDI were used to maximize the number of countries and the time period considered. Because GDP and population density had skewed distributions, they were log-transformed for subsequent analyses. Countries with density greater than 1000 inhab.km-2 were removed as they correspond to islands and city-states that are difficult to compare with larger states in terms of per capita impact.
Spatial relationships between SEI and external indices were analyzed by applying a segmented regression (Muggeo 2008) to assess potential turning points in SEI along GDP, density, or SDI gradients. Temporal changes in SEI and external indices were expressed as relative changes compared to 2005 values. Temporal relationships between SEI and the socioeconomic indices were analyzed using a multiple linear regression with the change in SEI between 2005 and 2015 as the response variable and change between 2005 and 2015 in the index considered (either GDP, density, or SDI) in interaction with the base index value (2005 value) as explanatory variables.
Relationship with GDP
Overall, SEI was robust to the removal of indicators (see robustness test in Appendix 2) and ranked high and low GDP countries in the bottom positions, whereas middle-income countries were often ranked higher in terms of SEI than in terms of GDP per capita (Fig. 2 and see detail by score in Appendix 3). Specifically, in space, SEI showed an increasing relationship with GDP below a GDP value equal to Int$6830 (CI = [5615, 8310]) and then a negative relationship for higher GDP values (Fig. 3a) (slope1 = 0.08, sd1 = 0.02, tvalue1 = 5, pvalue1 = 8.10−7, slope2 = −0.21, sd2 = 0.02, tvalue2 = -12, pvalue2 < 1.10−10, r2 = 0.69). This was due to the positive relationship between GDP and biophysical score, which is no longer compensated by the positive relationship between GDP and social score (see Appendix 3).
The inverse relationship between SEI and GDP at low and high GDP values was confirmed by the temporal analysis (Fig. 3 and Table 1). The change in SEI was not significantly negatively related to GDP, positively related to the change in GDP, but negatively related to the interaction term between GDP and the change in GDP. In other words, changes in SEI were positively related to changes in GDP when the initial GDP was low, but negatively related to changes in GDP when the initial GDP value was high (Fig. 3d and Appendix 4). The relationship reversed at a GDP value equal to Int$5900 (see details of the calculations in Appendix 4). This threshold was within the confidence interval found by the spatial analysis and was already exceeded by 110 countries (Appendix 5) inhabited by 4.3 billion people in 2011.
The opposite relationship between SEI and GDP was due to two factors. The initial increasing relationship was due to the positive relationship between the change in social achievements and the change in GDP between 2005 and 2015 (Fig. 4a and Table 1). The negative relationship beyond the threshold was due to the negative relationship between the change in social achievements and spatial GDP (Table 1) and to the positive relationships between 2005 and 2015 change in biophysical overshoot and the change in GDP (Fig. 4d and Table 1) as well as between the change in biophysical overshoot and spatial GDP (Table 1). For instance, in Chad, GDP per capita increased by 20% between 2005 (Int$1460) and 2015 (Int$1760) and at the same time all eight recorded social indicators increased and all biophysical indicators decreased except for ecological footprint and eHANPP. Conversely, in Russia, GDP per capita increased by 27% between 2005 (Int$20,390) and 2015 (Int$25,940) and at the same time 7 of the 10 recorded social indicators increased while democratic quality and access to energy decreased, and all biophysical indicators increased except CO2 emissions and eHANPP. In addition, all of the highest SEI values were found for low to medium GDP per capita (e.g., Moldova or Nepal with a GDP per capita in 2010 equal to Int$8610 and Int$2290, respectively). Note that all negative GDP changes were related to a decrease in the biophysical impact and almost all negative GDP changes were also related to a decline in the social score (except from Madagascar).
Interestingly, global GDP per capita (Int$14,275 in 2011 (Fig. 3) and Int$16,911 in 2019) was already above the turning points found in the spatial and temporal relationships between SEI and GDP. Moreover, for most of the social indicators that make up the social score, there were examples of countries close or above the minimum bar, albeit their GDP was lower or equal to SEI optimal GDP (Int$5900; Appendix 6). The only exceptions were sanitation and democratic quality for which no country with a GDP around Int$5900 was close to the minimum bar. Furthermore, the SEI-optimal GDP corresponds to the point where most countries become unable to stay below the biophysical limits (Appendix 6). These results remain valid when the 2010 world GDP per capita is used as a limit instead of the SEI-optimal GDP.
Relationship with population density
SEI showed a slightly positive relationship with population density in space (slope = 0.02, sd = 0.01, tvalue = 2, pvalue = 0.004, r² = 0.04) but not in time (Fig. 3b and e). In space, the slightly positive relationship was due to the low biophysical score, which only appears in medium to highly densely populated countries (Appendix 3).
The lack of a temporal relationship was due to the lack of a significant link between changes in social score and in population density (Fig. 4b and Table 2) and between the change in social score and population density in space (Table 2). However, there was a weak negative relationship between the change in biophysical impact and the change in population density (Fig. 4e and Table 2). Confirming the lack of a strong relationship between proximity of the safe and just space for all and population density, social achievements were met and biophysical indicators were below thresholds both in countries with high or low population density (Appendix 6). For instance, in Sri Lanka, population density increased by 7% between 2005 (298 inhab.km−2) and 2015 (319 inhab.km−2) with no transgression of biophysical thresholds and an increase in six of the seven recorded social indicators. Conversely, in China, population density increased by 9% between 2005 (133 inhab.km−2) and 2015 (145 inhab.km−2) with an increase in six out of seven recorded biophysical indicators.
Relationship with SDI
SDI was strongly related to SEI in space (slope = 0.37, sd = 0.04, tvalue = 10, pvalue < 1−10, r² = 0.42; Fig. 3c). Decoupling the analysis by social or biophysical scores (Appendix 3) showed two groups of countries. In the first, the social score was well correlated to SDI but the biophysical score was not. In the second group, high social scores were achieved for low SDI value, but the biophysical score was negatively related to SDI. Therefore, in each group, at least one of the two components of SEI was related to SDI leading to an overall correlation between SEI and SDI in space.
However, change in SEI was significantly negatively related to SDI (Fig. 3f and Table 3), not related to SDI change or to the interaction term between SDI and change in SDI (Table 3). Negative changes in SDI were all related to an increase in biophysical score whereas positive changes in SDI can be related to either a decrease or an increase in biophysical score (Fig. 4f). Conversely, changes in SDI and social score were not related (Fig. 4c and Table 3). In other words, many countries, although they changed significantly in SEI between 2005 and 2015, were not changing in SDI. Conversely, many countries although they changed significantly in SDI between 2005 and 2015 did not change significantly in SEI. In addition, as expected by construction, SDI was limited to account for other biophysical limits than CO2 emission and material footprint (Appendix 6).
Overall, this study demonstrates that it is possible to merge into a single socio-environmental index (SEI) both social achievements and environmental overshoots. The pathway to sustainability between social foundations and environmental ceiling may be narrow (Wiedmann et al. 2020) but this study shows the possibility for each country to assess its current position in the safe and just place, to track its historical pathway and to evaluate its progress toward a good life within the limits of the planet.
SEI and limits of SDI
SDI corresponds to the most advanced index measuring human development while accounting for CO2 emission and material footprint (Hickel 2020). It was constructed to assess the ecological efficiency of human development whereas the HDI did not take into account any environmental impact due to human development. Assessing social achievements in the light of most environmental degradation is crucial to establish an index that measures strong sustainability. In this study, the combined SEI was explicitly constructed to qualitatively and quantitatively encompass all social minimum bars and biophysical thresholds defining the safe and just space (O’Neill et al. 2018). SDI did not manage to encompass all social and environmental impacts of human development, while SEI proved successful by incorporating seven biophysical boundaries, 11 social indicators, and using distributive-based income indicator (Appendix 1). Furthermore, SEI matches the three requirements for constructing an effective socio-environmental index because it retains most of the information from the original indicators, it can be used as a complement of these indicators by summarizing them into social and biophysical scores, and it explicitly defines a goal to be achieved (Fig. 1).
Revisiting the Easterlin paradox
The analysis of SEI in relation to GDP reinforced the claim that GDP fails to measure progress toward a truly sustainable state. The results showed that above a given threshold, the socio-environmental index is negatively related to GDP not only in space but also in time, meaning one cannot be improved while the other increases. Furthermore, the negative relationship at high GDP values is not only caused by a degraded environment but also to a decline in social success (Table 1).
Temporal analyses are a key to validate the existence of a GDP threshold and propose to quantify it. If the threshold can be approached with spatial data for a given year, the shortcoming of space for time reasoning (Damgaard 2019) could have led to opposite interpretations. Yet the temporal approach qualitatively and quantitatively confirmed the spatial pattern. The results are consistent with previous findings from works on degrowth and steady state economy (SSE; Daly and Daly 1973, Kerschner 2010) but not with expectations of sustainable development (SD). SD and weak sustainability theories are based on the assumption of a Kuznet curve (Dasgupta et al. 2002, Dinda 2004) that an increase in GDP would have a negative impact on the environment when the value of GDP is low, but would then have a positive impact the environment above a threshold. It also assumes that growth will benefit social indicators. Thus, one should have seen a relationship between the combined SEI and negative or stable GDP (if social gains compensated environmental losses) for low GDP values, followed by a positive relationship above the Kuznet minimum. In other words, the Kuznet curve hypothesis would have resulted in a pattern opposite to that observed over space and time. Degrowth and SSE promote the idea that GDP growth may (or may not) benefit low-income countries, as GDP growth may be a reasonable proxy for increases in real welfare and decent living standards in some developing countries (Rao and Min 2018, and see Appendix 3). But unlike SD, degrowth and SSE assume the social and environmental drawbacks will quickly reverse the relationship beyond a given threshold, which is supported by the results.
Specifically, the existence of this GDP threshold for social achievements has been emphasized by the Easterlin paradox (Easterlin 1974, Welsch 2002, Majeed and Mumtaz 2017). Beyond a threshold already exceeded by many countries, non-economic well-being has become decorrelated or even anti-correlated with GDP. The most common explanation would be that environmental impacts (Otero et al. 2020) and social drawbacks of increasing GDP become stronger than social gains, breaking the correlation between welfare and GDP or even reversing it as hypothesized by Max-Neef (1995). Consistent with this expectation, there was an increasing or stable relationship between GDP and SEI for low values of GDP and a negative relationship above a threshold, and more specifically, the change in social score was negatively linked with the GDP value.
Defusing the population bomb
The number of people intensifies the per capita impacts on the planet as stated by the equation I = P × A × T in Ehrlich and Holdren (1971) and Daily and Ehrlich (1994), with I the environmental impact, P the population, A the affluence, and T the technology. It also increases the cost required to achieve the social foundations. Despite this mathematical intuition, the link between SEI and population density shows that different levels of population density appear to be compatible with high and increasing values in SEI, at least to a certain extent and in some countries. The absence of a negative relationship over time implies it is, in theory, possible to achieve social standards within planetary boundaries even in countries with increasing populations. At the same time, the results also show that a large population is not necessary to achieve high social standards, as proponents of human capital argue (Eberstadt 2007), because low-density countries achieve social minima as well as high-density countries. Overall, population density does not seem to be the main factor explaining the low social or high biophysical scores of many countries. Nevertheless, the relationship between the combined index and population density is valid only for countries with a population density of less than 1000 inhab.km−2, as there were no examples of non-city-state countries above this density. It cannot be excluded that population density becomes a predominant factor in the transgression of biophysical thresholds for higher densities as suggested by city-state examples. Furthermore, biophysical thresholds were downscaled to per capita values. Therefore, assuming that per capita values are fixed, population growth in countries that are currently below the thresholds at the national level will lead to threshold transgression. This phenomenon could even intensify if per capita values increase. On the contrary, population reduction in countries that are currently above the biophysical threshold at the national level could help them return below the threshold (Kopnina and Washington 2016). However, the time-scale required for such demographic changes (Bradshaw and Brook 2014, Ganivet 2020) makes lowering per capita values an inevitable and more effective tool.
A call for degrowth in the North
Most of humanity already lives in countries with per capita GDP above the GDP threshold values found in this study, although not all citizens from these countries receive the same share. Most social minimum bars can be achieved by countries with per capita income below the order of magnitude of the GDP threshold values. Moreover, it is even necessary to stay at these income values in order not to exceed any biophysical threshold. The current dynamics in developed and developing countries that had transgressed biophysical thresholds and experienced positive economic growth do not allow for the possibility of falling below the thresholds again at least for the next few decades (Appendix 3).
More than providing an exact threshold value for GDP, this study emphasized that achieving a safe and just space for all requires a redistribution of current wealth rather than growth in global GDP. There is currently enough wealth being produced, if not too much, as GDP values of the top-ranked countries as well as the optimal GDP values to maximize SEI are lower than the current global GDP per capita. A decrease in the environmental impact, which might result in a GDP degrowth, is needed to bring the impact of high footprint countries into a safe and just space. So far, the observed decreases in GDP, while related to a diminution in the biophysical impact (in particular CO2 emission and ecological footprint, see Appendix 3), were consequences of external pressures such as political troubles (e.g., Central African Republic) or economic crises (e.g., Spain, Italy, Greece). This implies a decrease in social achievements (especially life satisfaction) in parallel with the decrease in GDP.
A degrowth in the environmental impact is also necessary to provide room to maneuver for countries that have not met social needs, some of which potentially require increased use in energy, material, and space (Krausmann et al. 2017, Wiedenhofer et al. 2020), e.g., education, sanitation, access to energy, income, and nutrition (see Hickel 2019). Such a degrowth, if it results from political willingness and not economic recession, could also lead in developed countries to an improvement in the social part of SEI (D’Alessandro et al. 2020), although the possibility to reverse unsustainable development pathways remains to be empirically demonstrated. More than a marginal economic change, political degrowth corresponds to a paradigm shift (Hickel 2021) in which transformative societal changes (Parrique 2019) should avoid the deleterious social consequences of recession such as declining life satisfaction, employment, income, and democratic quality in developed countries, and social support and nutrition in developing countries (Appendix 3). But beyond GDP per capita, which emphasized inequalities between countries, the relationship between the socio-environmental index and inequalities within countries needs to be analyzed because these inequalities have both social and environmental consequences (Thorbecke and Charumilind 2002, Rao 2014, Piketty and Chancel 2015).
RESPONSES TO THIS ARTICLEResponses to this article are invited. If accepted for publication, your response will be hyperlinked to the article. To submit a response, follow this link. To read responses already accepted, follow this link.
We thank three anonymous reviewers for their helpful comments which greatly improved the manuscript.
All analyses were performed using the R software (version 3.4.4). The data and code that support the findings of this study are openly available in Dryad at https://doi.org/10.5061/dryad.866t1g1rq https://doi.org/10.5061/dryad.866t1g1rq.
Abdallah, S., S. Thompson, J. Michaelson, N. Marks, and N. Steuer. 2009. The Happy Planet Index 2.0: Why good lives don’t have to cost the Earth. New Economics Foundation, London, UK.
Begotti, R. A., and C. A. Peres. 2020. Rapidly escalating threats to the biodiversity and ethnocultural capital of Brazilian Indigenous lands. Land Use Policy 96:104694. https://doi.org/10.1016/j.landusepol.2020.104694
Bradshaw, C. J., and B. W. Brook. 2014. Human population reduction is not a quick fix for environmental problems. Proceedings of the National Academy of Sciences 111(46):16610-16615. https://doi.org/10.1073/pnas.1410465111
Brand, U., and M. Wissen. 2017. The imperial mode of living. Pages 152-161 in C. Spash, editor. Routledge handbook of ecological economics: nature and society. Routledge, London, UK. https://doi.org/10.4324/9781315679747-19
Crist, E., C. Mora, and R. Engelman. 2017. The interaction of human population, food production, and biodiversity protection. Science 356(6335):260-264. https://doi.org/10.1126/science.aal2011
D’Alessandro, S., A. Cieplinski, T. Distefano, and K. Dittmer. 2020. Feasible alternatives to green growth. Nature Sustainability 3(4):329-335. https://doi.org/10.1038/s41893-020-0484-y
Daily, G. C., and P. R. Ehrlich. 1994. Population, sustainability, and Earth’s carrying capacity. Pages 435-450 in F. B. Samson and F. L. Knopf, editors. Ecosystem management. Springer, New York, New York, USA. https://doi.org/10.1007/978-1-4612-4018-1_32
Daly, H. E., and H. E. Daly. 1973. Toward a steady-state economy. WH Freeman, San Francisco, California, USA.
Damgaard, C. 2019. A critique of the space-for-time substitution practice in community ecology. Trends in Ecology & Evolution 34(5):416-421. https://doi.org/10.1016/j.tree.2019.01.013
Dasgupta, S., B. Laplante, H. Wang, and D. Wheeler. 2002. Confronting the environmental Kuznets curve. Journal of Economic Perspectives 16(1):147-168. https://doi.org/10.1257/0895330027157
Di Tella, R., and R. MacCulloch. 2008. Gross national happiness as an answer to the Easterlin Paradox? Journal of Development Economics 86(1):22-42. https://doi.org/10.1016/j.jdeveco.2007.06.008
Dinda, S. 2004. Environmental Kuznets curve hypothesis: a survey. Ecological Economics 49(4):431-455. https://doi.org/10.1016/j.ecolecon.2004.02.011
Duro, J. A., and J. Teixidó-Figueras. 2013. Ecological footprint inequality across countries: the role of environment intensity, income and interaction effects. Ecological Economics 93:34-41. https://doi.org/10.1016/j.ecolecon.2013.04.011
Easterlin, R. A. 1974. Does economic growth improve the human lot? Some empirical evidence. Pages 89-125 in P. A. David and M. W. Reder, editors. Nations and households in economic growth. Academic, New York, New York, USA. https://doi.org/10.1016/B978-0-12-205050-3.50008-7
Eberstadt, N. 2007. China’s one-child mistake. Wall Street Journal, 17 September. https://www.wsj.com/articles/SB118997971938828979
Ehrlich, P. R., and A. H. Ehrlich. 2009. The population bomb revisited. Electronic Journal of Sustainable Development 1(3):63-71.
Ehrlich, P. R., and J. P. Holdren. 1971. Impact of population growth. Science 171(3977):1212-1217. https://doi.org/10.1126/science.171.3977.1212
Fanning, A. L., and D. W. O’Neill. 2019. The wellbeing-consumption paradox: happiness, health, income, and carbon emissions in growing versus non-growing economies. Journal of Cleaner Production 212:810-821. https://doi.org/10.1016/j.jclepro.2018.11.223
Ganivet, E. 2020. Growth in human population and consumption both need to be addressed to reach an ecologically sustainable future. Environment, Development and Sustainability 22(6):4979-4998. https://doi.org/10.1007/s10668-019-00446-w
Helliwell, J., R. Layard, and J. Sachs. 2018. World happiness report 2018. Sustainable Development Solutions Network, New York, New York, USA.
Hickel, J. 2019. Is it possible to achieve a good life for all within planetary boundaries? Third World Quarterly 40(1):18-35. https://doi.org/10.1080/01436597.2018.1535895
Hickel, J. 2020. The sustainable development index: measuring the ecological efficiency of human development in the Anthropocene. Ecological Economics 167:106331. https://doi.org/10.1016/j.ecolecon.2019.05.011
Hickel, J. 2021. What does degrowth mean? A few points of clarification. Globalizations 18(7):1105-1111. https://doi.org/10.1080/14747731.2020.1812222
Kerschner, C. 2010. Economic de-growth vs. steady-state economy. Journal of Cleaner Production 18(6):544-551. https://doi.org/10.1016/j.jclepro.2009.10.019
Kolasi, E. 2018. The physics of capitalism. Monthly Review 70(1):29-43. https://doi.org/10.14452/MR-070-01-2018-05_4
Kopnina, H., and H. Washington. 2016. Discussing why population growth is still ignored or denied. Chinese Journal of Population Resources and Environment 14(2):133-143. https://doi.org/10.1080/10042857.2016.1149296
Krausmann, F., K.-H. Erb, S. Gingrich, H. Haberl, A. Bondeau, V. Gaube, C. Lauk, C. Plutzar, and T. D. Searchinger. 2013. Global human appropriation of net primary production doubled in the 20th century. Proceedings of the National Academy of Sciences 110(25):10324-10329. https://doi.org/10.1073/pnas.1211349110
Krausmann, F., H. Schandl, N. Eisenmenger, S. Giljum, and T. Jackson. 2017. Material flow accounting: measuring global material use for sustainable development. Annual Review of Environment and Resources 42:647-675. https://doi.org/10.1146/annurev-environ-102016-060726
Kubiszewski, I., R. Costanza, C. Franco, P. Lawn, J. Talberth, T. Jackson, and C. Aylmer. 2013. Beyond GDP: measuring and achieving global genuine progress. Ecological Economics 93:57-68. https://doi.org/10.1016/j.ecolecon.2013.04.019
Kuznets, S. 1934. National income, 1929-1932. Pages 1-12 in National income, 1929-1932. National Bureau of Economic Research, New York, New York, USA.
Lenzen, M., D. Moran, K. Kanemoto, and A. Geschke. 2013. Building Eora: a global multi-region input-output database at high country and sector resolution. Economic Systems Research 25(1):20-49. https://doi.org/10.1080/09535314.2013.769938
Majeed, M. T., and S. Mumtaz. 2017. Happiness and environmental degradation: a global analysis. Pakistan Journal of Commerce and Social Sciences 11(3):753-772.
Marques, L. C. 2020. Capitalism and environmental collapse. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-030-47527-7
Max-Neef, M. 1995. Economic growth and quality of life: a threshold hypothesis. Ecological Economics 15(2):115-118. https://doi.org/10.1016/0921-8009(95)00064-X
Meadows, D. H. 1998. Indicators and information systems for sustainable development. The Sustainability Institute, Hartland, Vermont, USA.
Montoya, J. M., I. Donohue, and S. L. Pimm. 2018a. Planetary boundaries for biodiversity: implausible science, pernicious policies. Trends in Ecology & Evolution 33(2):71-73. https://doi.org/10.1016/j.tree.2017.10.004
Montoya, J. M., I. Donohue, and S. L. Pimm. 2018b. Why a planetary boundary, if it is not planetary, and the boundary is undefined? A reply to Rockström et al. Trends in Ecology & Evolution 33(4):234. https://doi.org/10.1016/j.tree.2018.01.008
Muggeo, V. M. 2008. Segmented: an R package to fit regression models with broken-line relationships. R News 8(1):20-25.
O’Neill, D. W. 2012. Measuring progress in the degrowth transition to a steady state economy. Ecological Economics 84:221-231. https://doi.org/10.1016/j.ecolecon.2011.05.020
O’Neill, D. W., A. L. Fanning, W. F. Lamb, and J. K. Steinberger. 2018. A good life for all within planetary boundaries. Nature Sustainability 1(2):88-95. https://doi.org/10.1038/s41893-018-0021-4
Otero, I., K. N. Farrell, S. Pueyo, G. Kallis, L. Kehoe, H. Haberl, C. Plutzar, P. Hobson, J. García-Márquez, B. Rodríguez-Labajos, et al. 2020. Biodiversity policy beyond economic growth. Conservation Letters 13(4):e12713. https://doi.org/10.1111/conl.12713
Parrique, T. 2019. The political economy of degrowth. Dissertation. Clermont Auvergne, Clermont-Ferrand, France.
Piketty, T., and L. Chancel. 2015. Carbon and inequality: from Kyoto to Paris. Trends in the Global Inequality of Carbon Emissions (1998-2013) and Prospects for An Equitable Adaptation Fund. Paris School of Economics, Paris, France.
Rao, N. D. 2014. International and intranational equity in sharing climate change mitigation burdens. International Environmental Agreements: Politics, Law and Economics 14(2):129-146. https://doi.org/10.1007/s10784-013-9212-7
Rao, N. D., and J. Min. 2018. Decent living standards: material prerequisites for human wellbeing. Social Indicators Research 138(1):225-244. https://doi.org/10.1007/s11205-017-1650-0
Raworth, K. 2012. A safe and just space for humanity: can we live within the doughnut? Oxfam, Oxford, UK.
Raworth, K. 2013. Defining a safe and just space for humanity. Pages 28-38 in The Worldwatch Institute. State of the World 2013. Island, Washington, D.C., USA. https://doi.org/10.5822/978-1-61091-458-1_3
Raymond, C., T. Matthews, and R. M. Horton. 2020. The emergence of heat and humidity too severe for human tolerance. Science Advances 6(19):eaaw1838. https://doi.org/10.1126/sciadv.aaw1838
Ripple, W. J., C. Wolf, T. M. Newsome, M. Galetti, M. Alamgir, E. Crist, M. I. Mahmoud, W. F. Laurance, and 15,364 scientist signatories from 184 countries. 2017. World scientists’ warning to humanity: a second notice. BioScience 67(12):1026-1028. https://doi.org/10.1093/biosci/bix125
Rockström, J., W. Steffen, K. Noone, Å. Persson, F. S. Chapin, III, E. Lambin, T. M. Lenton, M. Scheffer, C. Folke, H. Schellnhuber, B. Nykvist, C. A. De Wit, T. Hughes, S. van der Leeuw, H. Rodhe, S. Sörlin, P. K. Snyder, R. Costanza, U. Svedin, M. Falkenmark, L. Karlberg, R. W. Corell, V. J. Fabry, J. Hansen, B. Walker, D. Liverman, K. Richardson, P. Crutzen, and J. Foley. 2009. Planetary boundaries: exploring the safe operating space for humanity. Ecology and Society 14(2):32. https://doi.org/10.5751/ES-03180-140232
Running, S. W. 2012. A measurable planetary boundary for the biosphere. Science 337(6101):1458-1459. https://doi.org/10.1126/science.1227620
Scherer, L., P. Behrens, A. de Koning, R. Heijungs, B. Sprecher, and A. Tukker. 2018. Trade-offs between social and environmental Sustainable Development Goals. Environmental Science & Policy 90:65-72. https://doi.org/10.1016/j.envsci.2018.10.002
Stanton, E. A. 2007. The Human Development Index: a history. PERI Working Paper 127. Political Economy Research Institute, University of Massachusetts, Amherst, USA.
Talberth, J., C. Cobb, and N. Slattery. 2007. The genuine progress indicator 2006. Redefining Progress, Oakland, California, USA.
Thorbecke, E., and C. Charumilind. 2002. Economic inequality and its socioeconomic impact. World Development 30(9):1477-1495. https://doi.org/10.1016/S0305-750X(02)00052-9
UN. 2020a. GDP per capita, PPP (constant 2017 international $). United Nations Statistics Division, New York, New York, USA. http://data.un.org/Data.aspx?q=GDP+per+capita%2c+PPP&d=WDI&f=Indicator_Code%3aNY.GDP.PCAP.PP.KD
UN. 2020b. Total population, both sexes combined (thousands). United Nations Statistics Division, New York, New York, USA. http://data.un.org/Data.aspx?q=+Total+population%2c+both+sexes+combined&d=PopDiv&f=variableID%3a12
Welsch, H. 2002. Preferences over prosperity and pollution: environmental valuation based on happiness surveys. Kyklos 55(4):473-494. https://doi.org/10.1111/1467-6435.00198
Wiedenhofer, D., D. Virág, G. Kalt, B. Plank, J. Streeck, M. Pichler, A. Mayer, F. Krausmann, P. Brockway, and A. Schaffartzik. 2020. A systematic review of the evidence on decoupling of GDP, resource use and GHG emissions, part I: bibliometric and conceptual mapping. Environmental Research Letters 15(6):063002. https://doi.org/10.1088/1748-9326/ab8429
Wiedmann, T., M. Lenzen, L. T. Keyßer, and J. K. Steinberger. 2020. Scientists’ warning on affluence. Nature Communications 11:3107. https://doi.org/10.1038/s41467-020-16941-y
World Bank. 2015. World development indicators. World Bank, Washington, D.C., USA. https://datatopics.worldbank.org/world-development-indicators/
Table 1: Table 1: Effects of GDP and change in GDP (δGDP) on change in SEI (δSEI) and its components (δSocial score and δBiophysical score). Coefficient estimates (Estimate), standard error (Std. error), associated tvalue, significance level (p-value) and goodness of fit (adjusted r²).
|Response||Explanatory||Estimate||Std. error||tvalue||p-value||adjusted r²|
Table 2: Effects of population density and change in population density (δDensity) on change in SEI (δSEI) and its components (δSocial score and δBiophysical score). Coefficient estimates (Estimate), standard error (Std. error), associated tvalue, significance level (p-value) and goodness of fit (adjusted r²).
|Response||Explanatory||Estimate||Std. error||tvalue||p-value||adjusted r²|
Table 3: Table 3: Effects of SDI and change in SDI (δSDI) on change in SEI (δSEI) and its components (δSocial score and δBiophysical score). Coefficient estimates (Estimate), standard error (Std. error), associated tvalue, significance level (p-value) and goodness of fit (adjusted r²).
|Response||Explanatory||Estimate||Std. error||tvalue||p-value||adjusted r²|