Statistics are among the most powerful instruments in democratic governance. Deployed with precision, they illuminate inequality and compel reform. Deployed carelessly, they distort public understanding, burden entire communities with collective stigma, and desensitise the audiences whose engagement is most needed. Since 2024, Minister for Women, Children and Social Protection Sashi Kiran has delivered a series of parliamentary statements on gender-based violence that command national attention. Her concern for vulnerable women and children is not in question. What is in question is the statistical discipline underpinning her claims: the age of her data, the adequacy of her sampling, the representativeness of her sources, and the absence — across three years in office — of any measurable policy outcome to accompany the headlines. Good intentions unaccompanied by rigorous evidence do not protect victims. They produce political theatre.
When the data was gathered
THE most frequently cited figure in Kiran’s statements — that 64 per cent of women in intimate relationships have experienced physical or sexual violence — comes from the Fiji Women’s Crisis Centre survey “Somebody’s Life, Everybody’s Business.” The fieldwork was conducted in 2010 and 2011 under the Bainimarama military government, with the report published in 2013. That is 13 to 15 years ago. Fiji in 2010 was a closed authoritarian state: media censored, civil society constrained, and domestic violence reporting structurally suppressed. The democratic, post-MIDA Fiji of 2026 is a profoundly different society. Social conditions, family structures, employment patterns, and public attitudes toward reporting have all shifted materially. Citing a 2011 survey in 2026 without stating its vintage is not an oversight. Responsible ministerial practice requires the simple discipline of saying: “In 2011, this was the position.” Parliament and the public deserve that clarity.
A clinic is not a country
The 72 per cent figure Ms Kiran cited in May 2026 Parliament — that 72 per cent of Fijian women aged 18 to 64 have experienced lifetime violence — requires particular scrutiny. It derives from a study by Dr Lice Soraurau Vaniqi conducted at the Nadi Maternity Antenatal Clinic between March and June 2024. That study screened 345 women attending a single maternity facility over four months. Of those 345 women, 247, or 71.6 per cent, disclosed experiencing domestic violence in their lifetime. This is a clinical convenience sample of pregnant women at one urban Western Division facility — not a nationally representative random-sample survey of all Fijian women aged 18 to 64. Clinical samples systematically over-represent urban women and women in crisis. Presenting this as a national prevalence figure before Parliament, without qualification, is a fundamental breach of statistical responsibility. A sample of 345 from a single clinic cannot carry the weight of a national headline.
What police records can and cannot tell
Ms Kiran’s supplementary sources — police records, DPP case logs, helpline volumes — carry well-documented limitations which Ms Kiran herself acknowledged when she noted that increased case numbers “may reflect increased reporting and growing trust in the justice system”. This caveat, offered almost in parenthesis, is in fact the most important statistical observation in her entire statement. More reports do not equal more violence. They may reflect improved public confidence, expanded helpline access, and greater legal awareness — each a genuine policy success. The Western Division alone accounted for 37 to 40 per cent of all reported domestic violence cases between 2020 and 2024, reflecting population density and service availability rather than necessarily a higher incidence of violence. Rural and maritime reporting remains structurally incomplete. A figure drawn from partial geographic coverage cannot be presented as a national statistic without explicit qualification as to what it does and does not represent.
The root causes Kiran does not name
Conspicuously absent from three years of ministerial statements is any sustained engagement with root causes. Police Commissioner Rusiate Tudravu disclosed in December 2025 that of 37,215 criminal offenders recorded between 2023 and July 2025, 11,270 were unemployed, with financial stress identified as the principal trigger for property crime and assault. Drug-related offences surged from 1650 in 2023 to more than 20,000 in 2024, driven by Fiji’s documented role as a narcotics transit node. The criminological link between narcotics, economic marginalisation, and domestic violence is internationally established. Women walking home in suburbs like Tomuka face bag snatchers hiding in roadside undergrowth. Gold chains are snatched in broad daylight. These are symptoms of a society under serious drug and poverty pressure. A minister with genuine policy intent would name these drivers explicitly and co-ordinate cross-ministerial responses. That co-ordination has been absent from the public record.
How Australia and New Zealand handle this
Australia’s Personal Safety Survey — its primary national violence instrument — was conducted in 2005, 2012, 2016, and 2021-22. Every ministerial communication specifies the survey year and presents data as a trend line, not a single headline figure. The Australian Institute of Health and Welfare explicitly distinguishes improved reporting rates from actual incidence changes. New Zealand, which records one of the highest rates of intimate partner violence in the OECD — one in three people affected — frames all data within causal analysis naming methamphetamine, alcohol, housing insecurity, and intergenerational trauma as structural drivers. Neither country’s minister would cite a 15-year-old survey or a 345-patient clinic sample as a current national prevalence figure without qualification. Temporal labelling, disclosed sampling, trend analysis, and causal framing are not technically demanding. They are the minimum discipline required of a minister placing data before Parliament.
Three years, same statistics
Ms Kiran has now served three years as minister. Across statements in 2024, October 2025, November 2025, and May 2026, she has cited variants of the same figures: 60 per cent, 64 per cent, 72 per cent, $F300 million GDP cost, 4159 child sexual offence cases. The figures shift between statements without any explanation of methodological change. The $F300 million cost estimate has appeared in at least four separate statements with no source ever disclosed. Not once has the Minister announced the commissioning of a new national prevalence survey to replace 2011 data. Not once has she presented measurable outcomes: shelter capacity increased by a specific number, conviction rates improved by a documented percentage, police response times reduced to a target figure. What Parliament receives are process outputs — plans launched, workshops convened — with no impact metrics attached. A doorstop with statistics is not a policy. It is a press release.
The cost to social cohesion
Fiji is a nation of 900,000 people — grandfathers, fathers, teachers, and professionals who have built their lives on respect, hard work, and care for their families, men who have never raised a hand in anger. When a minister repeatedly presents statistics suggesting that 60 to 72 per cent of Fijian women have experienced violence, without temporal context, without causal nuance, and without distinguishing a 2011 snapshot from a 2026 reality, the social effect is not mobilisation. It is desensitisation and resentment. Men who are good husbands and citizens hear themselves absorbed into a statistical indictment that carries no qualification. Community leaders who might be powerful allies in genuine violence prevention disengage from a discourse that feels designed for the camera rather than for change. In a plural society that has lived through coups and economic hardship, social cohesion is a fragile achievement. Statistics without context do not protect women. They fracture the solidarity that lasting reform requires.
A call for rigorous governance
Ms Kiran has genuine authority and a genuine mandate. The National Action Plan 2023-2028 exists. The Domestic Violence Act 2009 is in force. What is missing is the statistical and policy discipline to match the legal architecture. Four steps are achievable and necessary.
First, commission a fully funded national prevalence survey with disclosed methodology and stratified random sampling across all divisions and ethnicities to replace data from the Bainimarama era.
Second, every ministerial statement must specify the date, source, and sample size of every figure cited.
Third, root causes — narcotics, unemployment, poverty — must be named, and documented cross-ministerial co-ordination must be publicly reported.
Fourth, present annual outcome metrics: not plans launched, but lives measurably improved. The minister represents all 900,000 Fijians — victims and perpetrators alike — and her duty of care extends to understanding why violence occurs, not only counting it. The women of Fiji deserve that rigour. So do the men who love them.
DR SUSHIL K. SHARMA is a former Associate Professor of Meteorology, Fiji National University, and Operational Meteorologist and Manager, Climate Research and Services Division, Fiji Meteorological Services. The views expressed are those of the author alone and do not represent the views of this newspaper.
The writer recommends the commissioning of a fully funded national prevalence survey with disclosed methodology and stratified random sampling across all divisions and ethnicities to replace data from the Bainimarama era. Picture: SUPPLIED
