Subscribe
Global Edition
AI IN SOCIETY

China

East Asia & Pacific · higher risk group · 12 indicators, 2010–2022

Ranked #1 of 24 on the AWPRI, with a score of 0.703 in 2022. A higher score means heavier animal use, weaker welfare governance, or research tilted further towards automation.

Risk Profile

AWPRI 20220.703
Rank#1 / 24
L1 · Animal-use exposure0.511
L2 · Governance capacity1.000
L3 · AI amplification0.598
Range 2010–20220.672–0.745
Radar chart of China's score on each of the 12 AWPRI variables in 2022, drawn against the 24-country mean.123456789101112

Variable scores in 2022. The solid shape is China and the dashed shape is the 24-country mean. A higher score means heavier exposure, weaker governance or a stronger tilt towards automation. The range chip above gives the highest and the lowest composite score this country has recorded across 2010–2022, which describes the country's own history and is not a confidence interval. Point at any vertex to see the variable's name and scores. The numbered key is below the chart.

  1. 1Farmed animals slaughtered per capita
  2. 2Aquatic animal share of flesh supply
  3. 3Meat supply per capita
  4. 4Animal share of protein supply
  5. 5Animal welfare legislation, stock
  6. 6Rule of law
  7. 7Animal welfare legislation, five-year flow
  8. 8Civic space
  9. 9Civil liberties
  10. 10AI research intensity
  11. 11Precision agriculture research intensity
  12. 12Animal welfare research base

Variable Breakdown

L1. Animal-use exposure (four variables)
Farmed animals slaughtered per capitapanel mean 0.488 · lower third of the panel · weight 0.0833 · direct welfare pathway0.323
Aquatic animal share of flesh supplypanel mean 0.416 · upper third of the panel · weight 0.0833 · direct welfare pathway0.673
Meat supply per capitapanel mean 0.570 · middle third of the panel · weight 0.0833 · direct welfare pathway0.543
Animal share of protein supplypanel mean 0.744 · middle third of the panel · weight 0.0833 · direct welfare pathway0.504
L2. Governance capacity (five variables)
Animal welfare legislation, stockpanel mean 0.433 · upper third of the panel · weight 0.0667 · direct welfare pathway1.000
Rule of lawpanel mean 0.327 · upper third of the panel · weight 0.0667 · indirect welfare pathway0.997
Animal welfare legislation, five-year flowpanel mean 0.605 · upper third of the panel · weight 0.0667 · direct welfare pathway1.000
Civic spacepanel mean 0.222 · upper third of the panel · weight 0.0667 · indirect welfare pathway1.000
Civil libertiespanel mean 0.224 · upper third of the panel · weight 0.0667 · indirect welfare pathway1.000
L3. AI amplification (three variables)
AI research intensitypanel mean 0.501 · upper third of the panel · weight 0.1111 · assumed welfare pathway0.845
Precision agriculture research intensitypanel mean 0.369 · lower third of the panel · weight 0.1111 · assumed welfare pathway0.209
Animal welfare research basepanel mean 0.555 · upper third of the panel · weight 0.1111 · assumed welfare pathway0.739

Each bar is this country's score on the 0 to 1 scale, which is the country's position between the lowest and the highest value any of the 24 countries recorded in 2022. A score of 1 means that no country in the panel stood further along that variable in that year, and does not mean that the underlying quantity reached any absolute maximum. The welfare pathway states how the variable connects to animal welfare. A direct variable counts animals or counts the law that covers animals. An indirect variable measures a condition under which welfare law is enforced or contested. An assumed variable is a proxy whose link to welfare outcomes this project asserts and has not demonstrated. The full statement for each variable is in the codebook.

What Moves This Country's Score

Pushing the score up

  • Civic spacescore 1.000 vs panel mean 0.222 · weight 0.0667 · +0.0519 of the composite · L2. Governance capacity
  • Civil libertiesscore 1.000 vs panel mean 0.224 · weight 0.0667 · +0.0517 of the composite · L2. Governance capacity
  • Rule of lawscore 0.997 vs panel mean 0.327 · weight 0.0667 · +0.0447 of the composite · L2. Governance capacity

Pulling the score down

  • Animal share of protein supplyscore 0.504 vs panel mean 0.744 · weight 0.0833 · −0.0199 of the composite · L1. Animal-use exposure
  • Precision agriculture research intensityscore 0.209 vs panel mean 0.369 · weight 0.1111 · −0.0177 of the composite · L3. AI amplification
  • Farmed animals slaughtered per capitascore 0.323 vs panel mean 0.488 · weight 0.0833 · −0.0138 of the composite · L1. Animal-use exposure

Each figure is the variable's weight in the index multiplied by its distance from the 24-country mean in 2022. Added up over all 12 variables they come to +0.2388, which is exactly China's distance from the panel's mean composite of 0.464. Because the weights differ between layers, a variable can be further from the mean than another and still move the score less. These figures are accounting terms and not causal estimates. The figures state which components of the index this country's score is made of, and do not state what would happen to animals if the underlying quantity changed.

Historical Trajectory and 2030 Trend

The solid line is the observed AWPRI score, plotted at one point per year. The dashed line is an ordinary least squares trend fitted to those 13 points and extended to 2030. The shaded band is the 95 per cent prediction interval around that fitted line, and is not an interval on the index itself. Where the slope is not distinguishable from zero the projection carries no information about direction and the chart says so. Point anywhere on the chart for year-by-year values.