A · Cost of artificial light
From candles to LEDs.
$785 / 1,000 lumens → $0.23
1800 to 1992. 99.97% drop in real cost.
Nordhaus 1996, J. Econ. Perspectives.
S16 The progress ledger
Is the world getting
better?
Two stories are simultaneously true. The acute deprivations that defined human history, death, disease, illiteracy, extreme poverty, are in dramatic and well-sourced retreat. Climate, conflict, forced displacement, and democratic backsliding are accumulating. A serious answer holds both.
One hundred indicators from primary sources. World Bank, WHO, UN IGME, UCDP, V-Dem, NOAA, IRENA, UNHCR, IUCN, IMF, ITU, UNESCO, FAO. Each tagged Optimistic, Mixed, or Negative and grouped by theme. The framework is Dalio's. Level, trend, and distribution shown together, so no single headline carries the verdict alone.
Filed
May 2026
Engine
Pure JS · inline SVG · no build
Scope
100 indicators · 5 themes · 1820–2025
§ I Two stories, both true
Before the pandemic, global life expectancy reached 73.1 years. Under-five mortality fell by sixty percent in a single generation. The cost of solar electricity dropped ninety percent in fourteen years. Smallpox was wiped from the planet. Guinea worm went from three and a half million cases to ten.
And. In 2024, atmospheric CO2 crossed 425 parts per million for the first time. There were sixty-one active state-based conflicts, the most since UCDP records began in 1946. One hundred and twenty-three million people were forcibly displaced, roughly one in every sixty-seven humans alive. For the first time in over twenty years, autocracies outnumbered democracies.
The honest answer to is the world getting better is neither yes nor no. The question depends on which metrics you are willing to look at and which baselines you are willing to use. The ledger below is the move past either side of that argument. It shows the level, the trend, and where in the world the change is concentrated. It refuses to vote.
§ II · FIG. 16.1 The ledger — 100 indicators
Each row is one indicator. Sparkline shows the trend. Direction badge shows whether the move is favorable, mixed, or unfavorable. Click a row to expand source and caveat. The year scrubber masks data after the chosen year. The COVID toggle highlights the 2020–21 dip on affected rows.
Theme
Direction
Year
2025
100 shown
0 Optimistic
0 Mixed
0 Negative
Indicator
Baseline
Current
Trend
Direction
Loading the ledger...
§ III · FIG. 16.2 Cost curves that changed everything
Four indicators that moved by multiple orders of magnitude in the lifetime of one investor. Log scales, real dollars, milestone markers. The case for technological abundance does not rest on a single chart. It rests on this pattern repeating.
B · Cost to sequence a genome
Faster than Moore's law.
$95M → $200
2001 to 2024. Roughly 475,000× cheaper.
NHGRI Sequencing Costs Data.
C · Utility-scale solar LCOE
From premium to default.
$417 / MWh → $43
2010 to 2024. 90% drop, real dollars.
IRENA, Renewable Power Generation Costs.
D · Lithium-ion pack price
Storage caught up.
$1,474 / kWh → $108
2010 to 2025. 93% drop.
BloombergNEF Battery Price Survey.
§ IV · FIG. 16.3 What's actually getting worse
Four indicators that have moved unmistakably the wrong way in this century. Same scale, same period, no filtering. The point is the shape, not the specific numbers.
CO2 concentration (Mauna Loa)
1990 354 ppm → 2025 425 ppm
NOAA Global Monitoring Laboratory.
Active state-based conflicts
2010 31 → 2024 61
Uppsala Conflict Data Program.
Forcibly displaced people
2010 43.7M → 2024 123.2M
UNHCR Global Trends.
Liberal democracy (pop-weighted)
2010 0.42 → 2024 0.27
V-Dem 2025 Democracy Report.
§ V Receipts · sources and caveats
Sources
26 primary publishers
World Bank, WHO, UN IGME, UNAIDS, UCDP, V-Dem, NOAA, NASA, NSIDC, IRENA, IEA, BloombergNEF, ITU, UNESCO UIS, FAO, IUCN, WWF/ZSL, UNHCR, IDMC, SIPRI, FAS, OECD, IMF, UNODC, World Inequality Lab, Carter Center.
Indicators
100 series
Twenty per theme. Each with a baseline year, a latest year, a sparkline, a direction tag, and a one-line caveat. Hand-curated, not scraped.
Historical range
1820 – 2025
Long-arc indicators (literacy, light cost, GDP per capita) reach back to the nineteenth century. Most series start 1990 or later.
License
CC BY 4.0 / source
Each indicator inherits its original publisher's license. WB, WHO, and UN data are mostly CC BY; some IEA charts are restricted.
Six named caveats that always apply.
- The Living Planet Index is not "73% of animals lost." It is a geometric mean of monitored vertebrate populations, an average of trends, not a count. Methodology contested (Charles University 2024).
- The extreme-poverty line moved in June 2025. From $2.15 to $3 in 2021 PPP. The trend survives the change. The level does not. Show both lines, never one.
- GBD revisions shift back history. IHME re-estimates global cause-of-death series each cycle. Cite the GBD release year, not just the underlying year.
- The COVID dip is real and recovers unevenly. Life expectancy, schooling, extreme poverty, vaccine coverage, child labour all dipped in 2020 to 2021. Some series have re-established trend, others have not.
- V-Dem and WGI are expert-coded. Composite governance measures aggregate underlying source disagreement. Show at least two for the same concept.
- AI benchmarks saturate. By late 2024, MMLU is functionally saturated. Frontier capability has to be tracked on harder benchmarks (GPQA, SWE-bench, ARC-AGI) that themselves shift.
§ VI Methodology & colophon
Pipeline
The hundred series are hand-curated from primary-source URLs in a single static indicators.json ↗ shipped with the page. Each entry has a baseline year and value, a latest year and value, a five-to-ten point series for the sparkline, a direction tag, and a one-line caveat. Numbers are checked against the publisher's most recent release as of May 2026.
Verdict
By the metrics that matter most to human suffering and capability, the world is dramatically better off than at any prior point. By the metrics that matter to systemic risk, planetary boundaries, and political stability, the picture is troubling and in several dimensions actively worse. The next decade will be decided less by whether humanity can invent more progress and more by whether institutions can stabilize the progress already arriving.
Reading list
Our World in Data ↗
World Bank Poverty and Inequality Platform ↗
V-Dem 2025 Democracy Report ↗
Honest caveats
A static snapshot in May 2026, not a live data layer. Indicators that revise back-history (GBD, MMEIG, PIP) will drift between annual updates. The Ray Dalio framing is intentional, an editorial choice, not a methodological claim. The page picks indicators that fit the level, trend, distribution rubric. Other curators would pick differently.