Julie Aunger bustles around her cafe near the Paignton seafront. It’s summer 2020, lockdown has eased, and the customers are coming back. The cafe is called “Our Katie’s Tea and Coffee House”. It’s named after her youngest daughter. 

I’m sitting at a table in the back, with a pink ring-binder file, compiled by Julie, in front of me. It’s packed with official documentation relating to Katie’s death, nearly four years ago, at the age of 21. She died of a drug overdose, with her abusive ex-boyfriend, Mitchell Richardson. They were found dead on 14 November 2016, by Richardson’s mother.

When the lunch rush is over, Julie sits down and starts to talk. She tells me about Katie’s life, and her relationship with Richardson, who was 12 years older than her daughter. She tells me about Katie’s sudden death, and what happened after the bodies were found. And she recounts, in forensic detail, the many failings in the police investigation that followed. 

She believed then, and believes now, that Mitchell Richardson planned Katie’s death. This wasn’t an accidental overdose, she says. She’s sure: the police failed her daughter. Inside the pink folder is a stack of documentation that gives credence to her claims.

Julie isn’t the only mother I’ve spoken to who is convinced that her daughter’s death is a homicide missed by police. In the UK, we’re told that two women a week are killed by their current or former partner. But the truth is, the number is undoubtedly higher than that. Over months of investigation, it has become clear that official numbers do not reflect the true level of lethality of domestic abuse. Deaths are being missed. Julie is not alone. 

For a decade now I’ve reported on these issues, and about a year ago, I began to hear about so-called hidden homicides – unseen, uncounted deaths. These are cases where, despite a known history of serious domestic abuse, police don’t investigate a possible crime. After arriving at the scene, decisions are made to accept events as they appear – a suicide, or an overdose, or an accident. Where families, keen to share their suspicions with police, are ignored. 

Together, these cases reveal a systemic problem at the heart of how we police domestic abuse. They are also a window into what the state cares about – and what it doesn’t. They matter at an individual level, but they are a snapshot of a poorly documented problem. While the Office for National Statistics collects data on homicides, it does not offer important context to the numbers, such as any background information on the abusive relationship that led up to the homicide, or even the sex of the partner or ex-partner responsible. Put simply, the state does not count how many women are being killed by domestic abuse. This job is done, instead, by a volunteer effort – the Femicide Census. 

In 2018, there were 91 killings of women by a man with whom they were, or had been, in an intimate relationship. But these are numbers reflecting ‘recognised’ murders or manslaughter – they have been investigated as killings by police. They’re the deaths we know about. Nobody has attempted to grasp the scale of hidden homicides; sudden, unexplained deaths with a background of domestic abuse where obvious explanations are accepted by police and coroners, and never fully probed.

As a historic piece of legislation is set to be enacted that will better protect victims – the Domestic Abuse Bill – this investigation reveals the near total lack of national data on deaths by domestic abuse. It reveals a clear pattern of poor decision-making that allows abuse to escalate, where obvious red flags are discounted, families’ concerns are ignored, and abusers’ narratives are all too often believed. It is an investigation that reveals a national scandal, hidden from view. 

I’m going to return to Julie Aunger, and her daughter Katie. But first, I’d like to introduce you to the Skeltons. 

In October last year, Peter and Elizabeth Skelton, a couple in their 80s, demanded that the state give them answers about the death of their daughter, Susan Nicholson, who was found unresponsive by her partner, a man called Robert Trigg, in 2011.

As soon as Susan’s body was found, a narrative took hold. This was an accidental death, Trigg said. She had accidentally been suffocated on the sofa, as they slept after an evening of heavy drinking. The police believed Trigg.

But to the Skeltons, the story didn’t add up. They urged the police to investigate. But Sussex police refused, despite the fact that they had Trigg on record as a serious domestic abuser – and despite the fact that his previous partner, a woman called Caroline Devlin, had also died suddenly. 

In the teeth of the authorities’ denial and delay, Trigg was only finally brought to justice because the Skeltons undertook their own investigation. They hired a solicitor, an experienced criminal barrister, and a Home Office forensic pathologist to look again at their daughter’s post-mortem. The evidence uncovered by this determined couple, then in their late 70s, forced Sussex police to return to the case and charge Trigg with murder. But not before he’d seriously abused two more women, after Susan’s death. In total, it took six years before Trigg was convicted of the murder of Susan Nicholson, which led to a second conviction of manslaughter for Caroline Devlin’s death, too. 

This is the risk when men who kill women are allowed to walk free. 

Professor Jane Monckton Smith is a forensic criminologist who studies patterns of domestic abuse. A former police officer, later a heavy metal band singer, later an academic and now a renowned authority on domestic homicide, I met her in her office at the University of Gloucestershire. She says that when police and coroners fail to properly investigate abusers who may have killed, there are inevitable consequences for future victims. These abusers “think they’re bulletproof,” she says.

But the Skeltons’ fight didn’t end with Trigg’s murder conviction. The coroner in Susan’s case originally gave a verdict of “accidental death”, but that was overturned after the trial. Now, the couple wanted police to accept responsibility, to hold Sussex Police accountable for their failure to protect their daughter. They asked for an Article 2 compliant inquest – only granted in cases where the state has sufficient knowledge of a risk to activate its duty to protect someone’s right to life. The Skeltons believed the authorities had plenty of evidence of Trigg’s history of violence to trigger that duty to protect Susan. The coroner disagreed. So did the police. 

And so, once again, they had to fight.

You have to be brave to take on the state. Things can turn ugly. And here, they did. Having for years ignored the Skeltons’ pleas, and having finally been proved completely wrong in their decision not to scrutinise the deaths of not just one, but two women, Sussex Police opted to double down. In a particularly graceless move, the force issued the Skeltons, a bereaved, elderly couple, with a threat. If they chose to contest the coroner’s decision to refuse an Article 2 inquest, and lost, they would be hit with the police’s legal costs. Thousands of pounds. After they’d already spent their life savings on the investigation that brought Trigg to justice. 

Peter Skelton, sitting beside his wife in their home in Worthing, told me: “We can’t understand why the police covered up for Susan’s murder when it was an obvious murder… You can understand perhaps, them making a mistake if it’s a borderline sort of thing. But this was obviously a murder. But why the police covered up for it, we can never understand.”

The Femicide Census, released in November 2020, showed that 27 of the men found to have murdered women over the past decade had killed before. Susan Nicholson’s case shows exactly what a hidden homicide can look like. When police fail to properly investigate a possible homicide after serious domestic abuse, there’s a risk that that other women will die. 

And these cases keep coming.

Renowned children’s author Helen Bailey was killed by her fiancé, Ian Stewart, in 2016. Stewart denied knowing anything about her disappearance, and was only arrested and charged three months later, when her body was found in a cesspit on her own property. He was shown to have drugged her into submission with sleeping medication, over months, and was convicted and jailed for her murder.

But then, questions began to be asked. Six years earlier, it emerged, his wife Diane Stewart, the mother of his children, had died suddenly and unexpectedly in the garden of their home. Stewart was the only other person on the premises at the time. Her family had been suspicious – but police had not followed up on their concerns. Belatedly, police began to investigate. In July 2020, Stewart was arrested and charged with her murder, too. 

We can’t report more about this case, because doing so could prejudice the upcoming trial. At his plea hearing in November, Stewart said he was not guilty.

We spoke to 14 family members and their lawyers about the deaths of a loved one. They told of police and coroner rebuffs, investigations that were lacking or absent, evidence not gathered or lost, mistakes, delays, and a culture of defensiveness that frequently manifested as aggression. These cases all share a clear pattern:

• Police fail to grasp the true danger of the risk repeat abusers pose to the women they live with.
• When police arrive at the scene of a sudden, or unexplained death, judgments are made too fast, often by officers not trained in domestic abuse or coercive control
• A systemic refusal by the authorities to take seriously and act on families’ concerns about domestic abuse.
• A failure to secure a scene when a death initially appears as a suicide or overdose meaning vital evidence is lost 
• Forensic post mortems are not ordered
• Witnesses are not interviewed 
• Prolonged battles by traumatised families to persuade authorities to reconsider their initial refusal to investigate a loved one’s death as a potential homicide.

It’s a pattern that became clearer as Julie Aunger told me more about Katie. Aged, 21, Katie was found dead in a flat alongside the body of the man who had abused her for 18 months. He’d spat on her, publicly humiliated her, beaten her and strangled her. After one particularly violent incident, the first time the police were called, Katie had refused to press charges. She had believed Richardson’s pleas that he wouldn’t hurt her again. 

But a second incident changed things for Katie. She had to be rescued from her flat after Richardson held her hostage for 36 hours, beating her and threatening to kill her, and to rape and kill her mum Julie. She had feared she wasn’t going to get out alive. This time, she bravely pursued a prosecution against him, and Richardson was charged with two counts of assault by beating, resisting arrest and criminal damage. He admitted the charges. This was the moment she left him. 

Katie was referred to the local MARAC – the multi-agency risk assessment conference – which only deals with the highest risk cases of domestic abuse. In her police statement, captured on video, she told special officers in detail about the abuse she’d suffered at Richardson’s hands.

Research shows that there are particular moments in an abusive relationship that are red flags. Non-fatal strangulation is one of them. The moment when a victim decides to leave her partner is another. But police are not always willing to take those red flags seriously. Nor do many officers, it seems, properly grasp the dangers of coercive and controlling behaviour, which can escalate when an abuser feels they are losing all means of control.

A month after giving that detailed statement, Katie was found dead. Given the two police rescues, the documented threats to kill, Katie’s mother can still hardly believe that the police decided it was ‘just’ a drug overdose. All the warning signs for homicide were there. “I’ve always believed that Mitchell planned her death,” Julie told me.

But Devon and Cornwall CID didn’t see things the same way.

Julie believes that the force had written her daughter off as a drug addict from the start, and says their approach to the scene of her death proved they had already decided that this death was not suspicious.

The result was a substandard investigation into the cause of death. Richardson’s mother, for example, by her own admission, had cleaned up the scene after finding their bodies, before police were called – but no action was ever taken against her. What forensic evidence remained was not properly stored by police, and by the time Julie asked for it to be readied for Katie’s inquest, the contents of the evidence bags were crawling with flies. A police report that was meant to describe Katie’s interview detailing Richardson’s abuse was written for the coroner, but it barely covered two sides of A4 paper, and offered a sketchy account of what she had said.

A hair strand test proved that Katie wasn’t a drug addict. That narrative does not hold up. And so Julie has fought, for four years, to get police to answer serious questions about how they treated her death. This is no longer just about her daughter, but about other women. She told me, she never wants another family to experience the shoddy work she says was carried out by Devon and Cornwall police, or go through an inquest feeling so bewildered and disempowered.

Sophie Naftalin is a solicitor who represents bereaved families at inquests, and specialises in cases where relatives suspect a death has been caused by domestic abuse. “It’s very traumatic for families. I think they feel very, very, very let down by that process,” she says. “And then I think it becomes almost a rolling nightmare because after the police investigation has let them down, they have some idea that the coroner will then be looking at the border circumstances.” But in reality, a limited police investigation begets a limited coroners’ investigation “because that evidence simply isn’t there. So it’s essential that the police in the first instance record information and ensure that it’s there. It may be that they don’t uncover any criminality, but if they don’t do that work in the first instance, they will miss that opportunity. And those families will be forever in that situation of not knowing.”

A situation of not knowing is exactly what Caramella Brennan is facing. 

In the middle of the Covid pandemic, in a special Nightingale court set up in Leeds, Caramella’s bereaved family brought a civil court case against a hospital for failing to properly store their daughter Emily Whelan’s body. Extensive decomposition of the body, which was not frozen for 10 months after Emily’s sudden death, meant a forensic post mortem was seriously limited at the point when the family’s concerns about domestic abuse were finally being taken seriously.

What happened to Emily after she died is complex and upsetting. It demonstrates the struggles that families go through to be heard in a system that doesn’t seem to pay attention. It underlines how crucial securing a scene and collecting evidence is. Because with some evidence… once it’s gone, it’s gone.

To understand, we have to go back to November 2016. Emily was found, early in the morning, by someone known to her. For legal reasons, we cannot name them. Her upper body was partially stuck in a gap down the side of the bed, her neck extended backwards. She was unresponsive and blue, and an ambulance was called. Rushed to hospital, she died the next day.

According to her family, Emily was just about to move out. Her bags were packed. Some of her children’s furniture had been taken down. Emily had never reported domestic abuse to the police. Few do – only an estimated 18 per cent of women. But straight away, her family were suspicious about her death. They allege that someone known to Emily was isolating and controlling her. It is evident from Emily’s diaries that she was struggling with her mental health. To her family, she seemed to have become insecure and withdrawn. But despite raising their concerns, the family struggled for months to get the police to take any notice.

And very likely, nothing would have been done if one night, Emily’s parents hadn’t sent a pleading message to Professor Jane Monckton Smith. Caramella remembers writing the subject line as if it was from Emily’s children. “We needed to find something to grab this lady’s attention because she probably gets hundreds of emails every week,” Caramella says.  “And we decided on ‘My lovely mummy age 25 is dead. Please help.’ And she got back to us, and thank God she did.”

Monckton Smith met the family, and talked to them a lot about Emily and about her life. With her experience of looking at domestic abuse and analysing risk, the forensic criminologist’s assessment was that there was high probability Emily was a subject of coercive control and domestic abuse. Her report concludes that had Emily not died, she was in danger of coming to further harm. But she also used her influence to get West Yorkshire Police to return to the case. 

A police investigation into the circumstances leading up to and including Emily’s death began in the summer of 2017 – nearly eight months after she died. It was everything the family had been pushing for. Eventually, an individual was arrested and questioned on suspicion of coercive and controlling behaviour. And to the relief of the family, a forensic post mortem was finally ordered. It was only at this point that Emily’s family were told the news that her body had been left to decompose so badly it was possible nothing useful could be learned. The news was traumatising.

“It was too late. Any chance we had to prove it, everything had already gone,” says Caramella. “I felt robbed. I was angry. I cried. I screamed. It was all to no avail. What can you do when the evidence has gone away? It’s just gone.” As she points out, nobody believes the authorities could make this type of gross error. “You don’t think the coroner makes mistakes. You don’t think the mortuary makes mistakes or the police. They absolutely do. They do. And I have no words for how I felt that day when they told me. I was so elated when they said, we’re going to do a forensic post mortem I thought, yes. Maybe we’ll get some answers now. And then when they said, well, I’m sorry, Emily’s body’s never been frozen.”

Although a pathologist concluded that Emily died of a lack of oxygen to the brain after a cardiac arrest, the underlying cause couldn’t be determined – and third party involvement could not be completely ruled out. But nothing can be proved definitively – because the crucial piece of evidence, Emily herself, can no longer be used. No other case I have reported on demonstrates as clearly the jeopardy that if evidence is not retained, and families are not listened to, everyone is left in the worst of all possible worlds. Without answers, and with suspicions left hanging – on all sides. 

The Senior Investigating Officers Handbook is a thick, weighty tome. It details the protocols, the approach, and the investigatory mindset that a senior officer should bring to any crime scene. Three basic first responder principles are hammered home:

A – Assume nothing.

B- Believe nothing.

C – Challenge and check everything.

Time and time again during this investigation, families have said they are shouting for someone to act – to forensically examine a scene, or a body, or the dynamics of a toxic relationship. They want police, coroners and pathologists to dig deeper, look harder.

But what if there is no potential killer? No smoking gun? What if there is no question that a woman who was being domestically abused has taken her own life. What then?

In 2006 a man escaped prosecution after he was accused of the manslaughter of his wife, Gurjeet Dhaliwal. Gurjeet was a mother of two sons, described after her death as bubbly, hardworking and loving. She had taken her own life the year before, in 2005. 

But, she’d kept a diary. And in that diary she’d detailed multiple incidents of psychological abuse and violence.

The Crown Prosecution Service tried to prosecute her husband with manslaughter, attempting to criminally connect his abusive behaviour with her suicide. But a judge at the Old Bailey ruled, no, the man could not be held criminally liable. And the court of appeal upheld that judgment.

Just over ten years later, in early 2017, a 46 year-old woman called Justene Reece took her life after suffering extreme domestic abuse, stalking and harassment by her former partner, Nicholas Allen. 

Before her death, Justene had reported 34 incidents of abusive behaviour to the police. She had taken sanctuary in a refuge, and had applied to a family court for a non-molestation order to prevent Allen from contacting her or coming to her house. Justene didn’t have a lawyer, so she represented herself in that court hearing – a daunting task. But Allen took no notice. He soon breached the order. In fact, he was later shown to have breached it seven times. And though the police were told, they didn’t arrest him.

The abuse Allen inflicted on Justene was unrelenting. In February 2017, she killed herself. 

It was only after her death that Staffordshire Police began to join the dots, and realised the enormity of what Allen had done. With evidence of stalking and harassment, officers approached the Crown Prosecution Service to ask if there  could be a ‘holding charge’ so he could be held on remand. This would give them time to search for more evidence, not just Justene’s phone and her social media, but from family and friends too – while protecting the public from this man.

Hannah Sidaway, a special prosecutor for the Crown Prosecution Service in the West Midlands was given the file. She could tell the police were shaken, as more and more disturbing evidence was being revealed. She wondered: could Allen be charged with manslaughter? 

She researched the law, and, undeterred by the appeal court’s 2006 refusal to allow a manslaughter prosecution of Gurjeet Dhaliwal’s partner, she read the comments made by the appeal judges. These comments – effectively guidance – told her precisely what she would need to build a successful manslaughter charge.

Speaking to me over Zoom in lockdown, Sidaway told me: “There were a few cases, not many, where the possibility of charging manslaughter, where a victim had taken their own life had been discussed, and the judges had made comments as to certain criteria which would apply, that may make such a charge possible. And it was a very tough set of circumstances.”

She had to be able to show that Nicholas Allen’s actions were the sole cause of Justene’s decision to take her own life. And with guidance from a senior criminal barrister, and after commissioning a psychological report from an expert who drew on the copious evidence now being secured by the police, she got what she needed to give police the nod. Allen could be charged with manslaughter.

Sidaway was in court the day he was brought in from prison to enter his plea. Despite having denied everything up until that point, Allen admitted manslaughter. There was never a trial. He was jailed for 15 years. The judge underscored his responsibility for Justene’s death, saying: “You clearly caused her to lose her life and before that to experience, over a protracted period of time, what must have been a living nightmare. She committed suicide as a direct result of your sustained and determined criminal actions, actions which you clearly knew were having a profound effect upon her.”

It was a legal first, and it has set a precedent. The case showed that an abuser could be successfully charged with manslaughter when their victim was unable to withstand their sustained battery of verbal, psychological and emotional abuse.

For Sidaway, this case may have been successful professionally, but personally, she pays a price for the job she does. “It’s very difficult as a prosecutor to keep emotionally detached,” she told me. “We have to maintain our objectivity because otherwise we’d make the wrong decisions on charges. And we have to make decisions on the basis of the evidence with the head, not with the heart. However… there are clearly some cases and some circumstances where you feel it is imperative that you get justice for the victim. I know members of the justice system who make sure no stones go unturned to get the best evidence, to ask the right questions, so that you get the best possible outcome for the family. But yes, it is traumatic.”

Reporting on suicide is ethically fraught, and ascribing single causes to someone’s decision to take their life is likely to be inaccurate. But speaking to Sidaway, reading the judge’s remarks, it’s clear: suicides can be legally, criminally, linked to domestic abuse. 

Which means we must now ask the question: how many suicides caused by domestic abuse are going uninvestigated, and uncounted, too? 

Jane Monckton Smith, the forensic criminologist I have relied on so much in shaping and reporting this issue, says the number could be frighteningly high… tripling, quadrupling the number of domestic abuse killings. 

If families seeking justice know that criminal charges can be brought if a loved one dies following sustained domestic abuse, there is likely to be increased pressure on the authorities to investigate these deaths. That pressure could gain momentum from a recent judgment handed down by the Supreme Court, which ruled that coroners only need to find that someone has been unlawfully killed on the balance of probabilities, rather than the former, more stringent standard of proof, “beyond reasonable doubt.”

Frank Mullane, chief executive of the charity Advocacy After Fatal Domestic Abuse says that the effect of this ruling will not be known immediately, “but what it may well mean is if a family has lost a loved one to a suicide following domestic abuse, and has good reason to believe that person was unlawfully driven to that suicide, then they may believe it was an unlawful killing.. and that is criminal.” A family would then, Mullane points out, “have good reason to go to police, or a court of law, and say this is what an inquest has determined on the balance of probability was an unlawful killing. Would you please apply resources and investigate this?”

As the story of Nicholas Allen’s prosecution for manslaughter shows, there are chinks of hope. As I write this piece, the Domestic Abuse Bill is in the House of Lords for its final five day debate before it is voted on, and becomes law. The Bill is a crucial step in protecting victims, because it cements the first ever statutory definition of what domestic abuse actually is. A range of newly recognised types of abuse are included in the definition, which extend our understanding beyond the notion of violence, or even coercive and controlling behaviour, to emotional abuse, and economic abuse – acts such as restricting someone’s access to money, or exploiting them, or sabotaging their ability to make a living.

This has far-reaching implications; if our understanding of domestic abuse is becoming more sophisticated, and enshrined in our laws, more victims can, theoretically, be protected by police and other public authorities. 

At the extreme end, it could mean that the risk of domestic abuse killings – like those of Caroline Devlin, or Susan Nicholson – could be recognised before someone loses their life. 

Without counting how many women in abusive relationships are dying suddenly, or in unexplained circumstances however, none of this resolves the issues uncovered in this investigation: the missed opportunities, failed or shoddy investigations, and poor police understanding of the toxic dynamics and risk inherent in abusive relationships. 

Homicides will stay hidden until there is a fundamental change in how police and coroners understand domestic abuse – and until we start, officially, to find a way of counting those who may have been killed because of it. 

Portraits by Tom Pilston for Tortoise