A psychological mechanism by which individuals with higher cognitive sophistication are hypothesized to be better able to selectively interpret evidence in ways consistent with their political or social identity. Contrasts with classical dual-process reasoning, which predicts that reflective thinking reduces bias.
Definition and scope
The identity-protective cognition (IPC) account, articulated prominently by Dan Kahan, proposes that:
- More cognitively sophisticated individuals deploy their reasoning abilities to defend identity-consistent positions
- Higher education and analytical thinking may exacerbate political polarization on identity-laden topics (climate, evolution, vaccines)
- Sophisticated reasoners are better at motivated reasoning—not less biased, but better at justifying their prior beliefs
This account stands in tension with classical reasoning theories, which predict analytical thinking reduces error and bias.
Empirical status
Recent work contests the scope and strength of IPC:
- Tappin, Pennycook & Rand (2021) — Rethinking the link between cognitive sophistication and identity-protective bias — Three experiments show cognitive sophistication predicts accuracy but not reduced polarization, and that when it does associate with ideology, it's mediated by prior factual beliefs, not identity-protective reasoning per se.
- Pennycook, Bago & McPhetres (2022) — Science beliefs, political ideology, and cognitive sophistication — Experiments on climate change show cognitive sophistication is associated with decreased (not increased) polarization under partisan motivation, contradicting IPC predictions; finds knowledge and analytic thinking predict accuracy across diverse science topics regardless of ideology.
- Bago, Rand & Pennycook (2023) — Reasoning about climate change — Directly tests motivated System 2 reasoning hypothesis; finds deliberation does NOT amplify partisan bias but instead increases coherence with prior beliefs (rational Bayesian updating with different priors, not identity-protective distortion).
Related mechanisms
- Motivated reasoning — broader tendency to process information in identity-supporting ways, but not necessarily via sophisticated reasoning
- Confirmation bias — selective attention to confirming evidence
- Belief overkill — tendency to amass arguments beyond what's needed to justify a belief