Rättssäkerheten i brottmål
Rättssäkerheten i brottmål: Rapport från JK:s andra rättssäkerhetsprojekt is a 2009 follow-up report. The local PDF is the official 266-page edition.
The report defines legal certainty in this setting as confidence that nobody is convicted unless guilt has been proved beyond reasonable doubt. It warns that judges may confuse a personal belief in guilt with satisfaction of an objective proof threshold, that generalized credibility impressions are shallow, and that courts must test realistic alternative hypotheses rather than rely on a total impression.
Its chapter on statement evidence traces error from perception and interpretation through memory, retrieval, communication, and the court’s own interpretation. It describes memory as reconstructive rather than reproductive, calls identification evidence notoriously unreliable, and warns that legal actors often lack relevant memory-psychology knowledge. The chapter supports Witness reports are not ground truth: a sincere or independent witness is not an independently verified truth source.
The report’s later empirical work was deliberately limited. Its sample of narcotics judgments and post-2006 reopening cases did not support a claim of pervasive failure across all criminal cases. The project also ended before several planned investigations were completed. It nevertheless found serious individual defects and described the correction system as giving finality too much weight when credible evidence of innocence emerges.
The report is therefore strongest as evidence about standards, known failure modes, and institutional design. It cannot establish a population rate of wrongful conviction. It supports Rättssäkerhet in Swedish criminal cases and False accusations as an epistemic risk.