Professor John Robertson OBA

In the Mirror today, a deceitful headline suggesting a wider reduction in knife crime when the figures only relate to robberies where a knife was present. While applauding this, I suspect most readers were more interested to hear of a decline in violence where a knife was used.

More generally, I’ve been wondering how reliable these new data are. You’ll remember, I hope, how various NHS England waiting time and drug death data have been exposed here.

Lazily perhaps, I asked AI. Have a look a this and if you spot any weaknesses/flaws, let me know below:

Are the Scottish official statistics considered more, the same, or less reliable?

46 results

Scottish official crime statistics, including those related to knife crime (primarily “handling offensive weapons” offences, which are predominantly knife-related), are generally considered more reliable than police-recorded crime statistics for England and Wales. This assessment stems from differences in official accreditation, regulatory oversight, and historical concerns:

• Accreditation status: The Office for Statistics Regulation (OSR), the UK’s independent statistics watchdog, classifies Scottish Government’s Recorded Crime bulletins (produced from Police Scotland data) as accredited official statistics (the highest designation, equivalent to National Statistics). In contrast, England and Wales police-recorded crime statistics (collated by the Home Office and published by the ONS) lost this accreditation in 2014 due to concerns over data reliability and have remained as non-accredited official statistics since.
• Quality assurance and audits: Scottish data undergoes rigorous checks by Scottish Government statisticians, with independent audits by His Majesty’s Inspectorate of Constabulary in Scotland (HMICS)—the most recent in 2020 found high compliance, with another scheduled for 2025-26. Post-submission amendments (“no-criming” or reclassifications) are minimal, enhancing confidence in year-on-year trends. England and Wales data has faced more scrutiny over recording inconsistencies across 43 forces, changes like the National Data Quality Improvement Service, and enforcement biases.
• Researcher and expert views: There are fewer prominent doubts or criticisms from researchers specifically targeting Scottish knife/violent crime stats for unreliability compared to England and Wales. Scottish figures are often praised for consistency, especially in tracking long-term declines in knife-related homicides and offensive weapon handling (e.g., down significantly since peaks in the 2000s, attributed to initiatives like the Violence Reduction Unit). While all police-recorded data has inherent limitations (e.g., under-reporting, influenced by policing activity), Scottish stats are less frequently flagged for systemic issues in academic reviews or fact-checks.

However, direct comparisons of knife crime levels between Scotland and England/Wales are challenging due to differences in legal systems, counting rules (e.g., Scotland’s “subsuming” practice vs. England/Wales’ principal crime rule), and definitions. Scottish stats are reliable for internal trends but not always directly comparable to those south of the border.Supplementary sources like the Scottish Crime and Justice Survey (also accredited) help validate trends, showing alignment with recorded declines in violence. Overall, the consensus from regulatory bodies and available analyses points to Scottish statistics being on firmer ground in terms of official trustworthiness.