Social Credit Systems promise fairness and efficiency but we find that they can reinforce existing inequalities and bias our perceptions in lasting ways

Across three preregistered experiments, we find that when SCS scores are made available:

1) Overall trust behavior decreases

2) Cooperation between individuals decreases

3) And strikingly, when people with lower scores behave identically to others they are judged more harshly and rewarded less

Despite the clear privacy and moral issue, the justification for SCS systems is that transparency leads to trust and encourages interactions.

But what if the opposite is true?

It sounds like dystopian fiction but Social Credit Scores are a reality for millions of people around the world and are becoming increasingly prevalent, even in western societies.

Additionally, robustness checks suggest that relatively small samples (n<30) are able to support accurate neural forecasts

Using brain measures we are thus able to access a more universal index of preference that results in improved accuracy of market-level forecasts.

In the experiments, we vary the representativeness of the laboratory sample to the market of interest. We find that while behavioral forecasts are significantly impacted by sample/market similarity, neural forecasts are not.

We find these more widely shared signals primarily in early affective responses measured in the nucleus accumbens (NAcc).

In this paper we show that the decision-making process can be broken down into components, some of which are more widely shared across individuals then others. Neural measures of the shared components consequently lead to more accurate aggregate level forecasts.

Previous research on Neuroforecasting has shown that neural activity collected from a small sample in the laboratory can improve forecasts of real-world market preferences. But what has been less clear is how this is happening.

Link to paper: academic.oup.com/pnasnexus/ar...