The Sustainable Development Index methodology was updated in 2026 to version 2.0, as published by Hickel, Onder and Zoomkawala in Ecological Economics. The full method is described in that paper, which seeks to resolve several remaining problems in the literature. What follows here is a brief overview.

The formula

The SDI starts with the Human Development Index (HDI) and applies a penalty (or adjustment factor, A) for ecological overshoot. The general formula is as follows:

SDI = HDI x A

To determine the adjustment factor (A), we assess each nation’s CO2 emissions and Material Footprint as a ratio of the sustainable boundaries:

We then obtain the average overshoot (AO) as a geometric mean of the two ratios. Ratios greater than 1 mean ecological pressure exceeds the boundary (overshoot). Ratios less than 1 mean pressures are within the boundary (undershoot). When undershoot occurs, the result is set at 1 by clipping the function. This prevents undershoot on one pressure indicator from compensating for overshoot on the other in the AO equation, and prevents undershoot from compensating for low development in the final SDI equation. This approach maintains the principle of strong sustainability. If both emissions and material footprint are less than the boundary, then AO=1 and the HDI is not penalised.

Next, we index the AO between 0 and 1 by subtracting 1 and then applying a Gaussian squared exponential decay function. This gives the adjustment factor (A), which declines monotonically and tends toward zero as overshoot grows.

The severity of the penalty depends on the constant k, which can be defined differently for different specifications, as described in the paper.

For emissions and material use we rely on consumption-based data as much as possible. This accounts for international trade by adding the emissions and materials embodied in imports (including the upstream materials and emissions involved in producing and shipping imported goods) and subtracting emissions and materials embodied in exports.

The specifications

For the default specification (SDI-1850), each nation’s cumulative CO2 emissions are measured from 1850 against its fair-share of the planetary boundary of 350pm concentration of CO2 in the atmosphere. Each nation’s cumulative material use is measured from 1970 (the first year of MF data) against its fair-share of the global sustainable limit of 50 Gt per year (without allowing undershoots in some years to compensate for overshoots in others). This cumulative historical approach is useful in that it fully accounts for each nation’s contributions to global ecological pressures, and implements equity principles by allowing developing countries to use their full fair-shares of planetary resources.

For the second specification (SDI-1970), the only change is to measure emissions from a later start date, 1970, providing a baseline that is more relevant to contemporary knowledge about anthropogenic climate change. Material use remains measured from 1970. For both SDI-1850 and SDI-1970 the penalty function is calibrated with a k of 0.416, which ensures that when emissions and material use exceed the boundary by 200%, the HDI is cut in half.

For the third specification (SDI-Current), historical emissions and material use are ignored. Each nation’s annual emissions is measured against its fair-share of the remaining carbon budget for a 50% chance of limiting global warming to 1.7C in any given year. Each nation’s annual material use is measured against its fair-share of the global 50 Gt limit. By ignoring historical emissions and material use this approach violates equity principles (developing countries face tighter ecological constraints imposed by the excess historical emissions and material use of rich nations), but it better reveals year-on-year changes in ecological efficiency, which are otherwise obscured by the cumulative historical specifications.

Data sources

For HDI data (life expectancy, years of schooling and GNI per capita) we use the UNDP's Human Development Report 2025, which provides data through 2023.

For CO2 emissions we use two datasets: EORA PRIMAP (v199.82), which provides consumption emissions from 1990-2019, and PRIMAP-HISTTP (v.2.8), which provides territorial emissions from 1850-2025. For the period 1850-1969 we use territorial emissions. For the period 1970-1989 we estimate consumption emissions by using the ratio of consumption to territorial emissions for 1990-1994. For the period 1990-2019 we used consumption-based emissions. For the period 2020-2023 we estimate consumption emissions by using the ratio of consumption to territorial emissions for 2015-2019.

For material footprint, data is obtained from the UN International Resource Panel Global Material Flows database.