Select Page

Rigorous research on gender issues

Rigorous research on gender issues

Rigorous research on gender issues

Sexual Assault Allegations

Sexual Assault

Launch Topic

Research Paper

Reviewing estimates of false rape claims

The sexual violence literature understates the prevalence rate

While there is little doubt that the majority of sexual assault allegations are genuine, historically estimates of the prevalence of false allegations have varied widely.

However, prominent sexual violence academics now agree that the prevalence rate is estimable within tight bounds and very low — with the “credible” estimates said to be “around 5%” or “between 2% and 10%” of reports being false.

These consensus estimates have been relied on in official publications and the media, and used to influence government policies.

In this research paper, we identify several shortcomings in the empirical studies on which the consensus is based. The specialised police rules they follow to determine whether to classify a report as “false” exclude many false and potentially false reports. Some of the studies also suffer from incomplete or poor-quality data, limited interview response rates and mathematical errors.

The studies’ prevalence rates are better viewed as lower bound estimates, and we show that the actual prevalence rate could be materially higher.

Click below to read the paper.

Academic endorsement for the paper

I appreciate the care and meticulous attention to detail shown in this manuscript. It makes a valuable contribution to the field, enhancing the credibility of research on false reporting with a methodical, transparent, and eyes-wide-open assessment of this most complex and challenging issue.

Dr Kim Lonsway
Director of Research
End Violence Against Women International
(Lead researcher for the 2009 US high-quality prevalence study on false sexual assault reports)


This is an important and overdue study. Given the paucity of research on false allegations of sexual offences, but also the valid concerns about giving disproportionate attention to them, this is just what is needed: a quantitative analysis which raises pertinent issues regarding estimations of prevalence, definitions, and methodologies, but without, it seems, taking any side or being polemical.

Dr Ros Burnett
Centre for Criminology
University of Oxford
(Editor of ‘Wrongful allegations of sexual and child abuse’, OUP 2016)

“Keeping mum” about false allegations

Are some women’s advocacy bodies, sexual violence academics and the ABC deliberately obscuring the truth about the prevalence rate?

Government bodies concerned with women’s safety, most notably the Australian Institute of Family Studies (AIFS) and Australia’s National Research Organisation for Women’s Safety (ANROWS), along with the national broadcaster, have cited the consensus prevalence estimates to argue that sexual assault allegations are “overwhelmingly true”.

We first alerted the AIFS and the ABC to the errors in their publications more than 18 months ago, but neither body has raced to correct the record.

  • The AIFS delayed withdrawing its offending publication until December last year, and did so only quietly — without issuing a formal correction.
  • The ABC at least issued a low-key, partial correction to some of its articles, but its reporters have continued to promote the narrative that false allegations are rare.

We met similar behaviour from some (though not all) sexual violence academics when we sought their input and feedback on our research. Alas, not all welcomed our scrutiny or seemed keen to come clean about their false allegations work.

This raises the question of whether the people concerned have allowed allegiance to a partisan viewpoint — that by default we should believe women who allege sexual assault — to trump the objectivity one would usually expect or hope for.

Whatever the cause, the upshot is that these bodies have misled the public, people in the criminal justice system and policy-makers.

Click left to read our full exposé on this saga, and below for our exchanges with the ABC and AIFS.