Avoiding bias

We have built Oxford Analytica to avoid bias -- right down to the design of our organisation and network.

To provide the most accurate assessments of the global political economy, we have to avoid bias -- even unconscious bias. We have spent more than 40 years developing techniques to identify and mitigate bias.

Bias is one of the most fundamental of errors in qualitative analysis. Social science has shown that people often mistakenly prioritize certain pieces of evidence when drawing conclusions. This may be due to a desire to see previous findings reaffirmed or using the most accessible evidence or straining to connect unrelated data. When analysts make incorrect pronouncements on the direction of the world, bias is almost always found at the beginning of their methodological process.

We have built Oxford Analytica to avoid bias -- right down to the design of our organisation and network.

Four key ways we avoid bias

Independence prevents us from shading our analysis to suit a corporate, partisan or ideological agenda.

Studies have found that desire to protect reputations encourages established analysts to stick to conventional wisdom while the need to make a splash leads young analysts to make bold claims. The absence of bylines in our published analysis and much of our advisory work encourages our analysts and network to make the right call without thinking about career implications.

We critique our analysis using both internal and external experts. This process exposes biases, provides rigorous challenges to arguments and allows new viewpoints to fill gaps left by the initial author.

Many issues in the global economy cut across regions and topics, yet academia and business often inhabit discrete silos. By running analysis across experts from different fields, we identify the blind spots inherent in any one area.