Rising Global Savings Can Hide an Uneven Sustainability Picture

What remains after depletion?

Conventional savings figures do not tell the whole story of an economy’s changing productive resources. Savings may rise while energy reserves, minerals or forests are depleted. Environmental damage also complicates the picture.

Xin Xiao’s preprint, Global Trends and Heterogeneity in Modified Adjusted Net Savings, examines an accounting measure intended to bring some of those costs into view. Covering 1990–2021, it reports substantial growth in the global total alongside persistent differences between countries. As a preprint, the study should be read as preliminary research rather than treated as an established, peer-reviewed conclusion.

Its central message is about interpretation: a larger worldwide dollar total does not necessarily mean that sustainability-related savings have improved proportionately, or that progress is widely shared.

A deliberately narrower indicator

The study starts with the World Bank’s adjusted net savings series excluding particulate emission damage. This adds education expenditure to net national savings, then subtracts energy depletion, mineral depletion, net forest depletion and carbon dioxide damage.

The author constructs modified adjusted net savings, or MANS, by subtracting education expenditure from that starting series. The education term therefore cancels out. In plain language, MANS is net national savings after deductions for the specified resource depletion and carbon damage.

Removing education expenditure is intended to prevent differences in that component from dominating comparisons between countries. It does not establish that education is unimportant. Rather, the indicator deliberately leaves that investment component outside its final calculation. MANS is a derived measure, not the headline World Bank series itself.

The underlying framework is called weak sustainability. It assumes that different forms of capital can substitute for one another: gains in one category may offset losses in another. That is a consequential assumption, not a demonstration that ecological losses can always be repaired or replaced.

Looking at the same accounts several ways

The analysis uses annual World Development Indicators data. Coverage differs across countries and years, making the dataset an unbalanced panel rather than a complete record for every economy.

The paper compares global totals, values per person and MANS as a share of nominal GDP. Each answers a different question. Totals reveal the scale of recorded savings after deductions; per-person values account for population size; GDP shares place the measure alongside the size of an economy.

Maps show relative country positions in 1990, 2000, 2010 and 2021. Other displays examine country histories, medians and the spread of outcomes. Income-group comparisons and exploratory statistical analyses add detail. These are descriptive methods: they document patterns without identifying their causes.

Growth without uniform improvement

The paper reports that global MANS increased substantially in current US dollars, particularly after the early 2000s. Its share of nominal GDP followed a more volatile path. A rising total therefore cannot, by itself, demonstrate a corresponding improvement relative to economic output.

High-income economies tend to occupy stronger positions in the per-person distribution. Large economies dominate total amounts, illustrating why total and per-person rankings should not be confused. Country positions also appear persistent: relatively strong and weak performers often retain their broad standing over time.

The income-group comparisons show higher median MANS relative to GDP in high-income countries, followed by upper-middle-income countries. Lower-middle-income and low-income groups generally have lower medians and greater variability. The author also describes pronounced fluctuations in some resource-dependent economies.

Resource depletion, savings and carbon damage all enter the accounting formula, making them relevant to interpretation. But this study does not establish that resource dependence caused particular changes. Its exploratory fixed-effect results are conditional associations, not causal estimates.

Important limits on the picture

Current-dollar amounts are not adjusted for inflation or exchange-rate movements. Changes can consequently reflect prices and currency values as well as underlying economic developments. Depletion and damage estimates also depend on valuation assumptions and national accounting practices.

There are presentation limitations too. The methods and results describe the country-ranking figure inconsistently as total versus per-person MANS. Another figure’s axis units do not consistently match its description. Those discrepancies warrant caution about detailed figure-based comparisons, even though the broader narrative repeatedly distinguishes totals, per-person values and GDP shares.

Finally, neither particulate emission damage nor the education component is included in MANS. The measure cannot provide a complete ecological or human-capital assessment. The paper recommends reading it alongside physical natural-capital indicators, emissions, biodiversity and distributional measures.

The useful takeaway is therefore narrower than a global sustainability scorecard: aggregate savings after selected environmental deductions can rise while country-level gaps remain substantial.