The Lethality Threshold
Domestic Homicides Escalating
TL;DR (The Quick Version)
In 2019, the FBI recorded 1,065 domestic homicides in the United States.
In 2024, that number was 2,339.
More than doubled. In five years. According to the same agency, using the same methodology.
The standard explanation is the COVID-19 pandemic: lockdowns trapped victims with abusers, economic stress elevated violence, social support networks were severed. This explanation is real. It is documented. And for 2020 and 2021, it is partially sufficient.
It is not sufficient for 2022, 2023, and 2024 - years when lockdowns had ended, pandemic emergency conditions had resolved, economies had largely recovered, and domestic violence services had reopened.
The homicides kept climbing.
This signal is not an argument that the pandemic caused nothing. It caused a great deal. The argument is more specific: the pandemic explanation has a logical endpoint that the data does not respect. Pandemic conditions ended. The homicide trajectory did not reverse.
Something changed in the population of perpetrators - in the neurological systems governing impulse control, threat calibration, and the capacity to stop before the irreversible act - that is not explained by lockdown conditions that no longer exist.
The prior signal in this series asked: if something is degrading impulse control in dogs, what is it doing in humans?
This signal is the answer.
THE DATA
SOURCE: FBI, as reported by CBS News (December 2025) and corroborated by multiple statistical compilations:
Domestic homicides, United States:
2019: 1,065
2024: 2,339
Change: +119.6% (more than doubled)
For context on the longer trend:
Over the decade 2014-2023, intimate partner homicides of women increased 22%, driven primarily by firearm homicides, which rose 36% over the same period (Everytown Research, CDC data).
Female intimate partner homicides by all other means (non-firearm) increased only 3% over the same decade. The surge is concentrated in firearm lethality.
Nearly 70% of intimate partner homicides in the US are committed with a firearm. More than 70 women are shot and killed by an intimate partner every month, on average (2020-2023).
In 2021, 34% of all female murder and non-negligent manslaughter victims were killed by an intimate partner. Intimate partners account for approximately 50% of female homicides.
A study of intimate partner homicides found that 20% of victims were not the intimate partners themselves, but family members, friends, neighbors, bystanders, or law enforcement who intervened.
The Bureau of Justice Statistics documented more than 1.8 million domestic violence victimizations in 2024 alone.
The economic cost: approximately $3.6 trillion annually in medical costs, lost productivity, and criminal justice system expenses - a figure that reflects chronic, deeply embedded patterns of violence, not episodic pandemic spikes.
THE PANDEMIC EXPLANATION: WHAT IT GETS RIGHT
The COVID-19 pandemic did increase domestic violence. The evidence is clear and consistent across multiple countries and methodologies. The mechanisms are straightforward:
PROXIMITY AND ISOLATION: Lockdowns placed victims and perpetrators in continuous, inescapable proximity. The escape valve of work, school, and social contact that normally interrupts the cycle of violence was removed. Victims could not leave without violating public health orders. Perpetrators could not be removed to the street.
ECONOMIC STRESS: Job losses, income instability, and financial uncertainty are well-documented risk factors for intimate partner violence. The pandemic produced all three at scale. Research consistently finds that when a perpetrator loses income and the victim becomes the sole provider, violence risk increases as abusers attempt to reassert control.
SEVERED SUPPORT NETWORKS: Domestic violence services saw surges in demand precisely as their capacity was reduced. Shelters limited intake for infection control. Hotlines were overwhelmed. The social detection network - neighbors, teachers, coworkers, extended family who might notice and intervene - was disconnected.
SUBSTANCE USE: Alcohol and drug use increased substantially during lockdowns. Substance use is a documented amplifier of violence risk in relationships where it is already present.
The pandemic explanation is real. For 2020, it is compelling. Early 2020 data from multiple countries showed immediate increases in domestic violence calls, service demand, and reported incidents within weeks of lockdowns beginning.
THE PANDEMIC EXPLANATION: WHERE IT FAILS
Here’s the problem.
Pandemic conditions have a timeline. Lockdowns ended. Vaccines were deployed. Economies reopened. Schools and workplaces resumed. Social support networks reconnected. Domestic violence services returned to full operation.
If the pandemic were the cause of the domestic homicide surge, the trajectory should have reversed - or at minimum stabilized - as pandemic conditions resolved. In prior research on violence following natural disasters and emergencies, elevated rates typically return toward baseline as the emergency conditions that produced them lift.
The domestic homicide data does not show a reversal.
The FBI data runs from 1,065 (2019) to 2,339 (2024). The pandemic’s acute phase effectively ended in 2021-2022. The numbers continued rising through 2022, 2023, and into 2024.
The pandemic explanation predicts a spike followed by a return toward baseline.
The data shows a step change that is maintained and extended.
This is the critical distinction. A stress response returns to baseline when the stressor is removed. A threshold crossing - a change in the underlying system that produces the behavior - does not reverse when environmental conditions normalize.
The domestic homicide data is consistent with a threshold crossing. It is not consistent with a stress response to pandemic conditions.
What kind of threshold crossing would produce this pattern?
The answer from the rest of this signal series: neurological.
THE IMPULSE CONTROL FRAMEWORK
Domestic homicide is not simply domestic violence made lethal by circumstance. It is a specific failure of the inhibitory systems that, in most violent episodes, stop the violence before death.
The research literature on intimate partner homicide consistently identifies a specific psychological profile of perpetrators: not simply individuals who are more violent, but individuals who escalate past the stopping point that most perpetrators - even those who commit serious violence - do not cross.
The stopping point is a neurological phenomenon. It involves the prefrontal cortex - specifically the orbitofrontal and ventromedial PFC regions that govern impulse inhibition, threat assessment, and the rapid evaluation of consequences. In most violent episodes, even among habitual domestic abusers, these systems activate to interrupt escalating violence before lethality. They are the brakes.
What the MNP neurotoxicity literature documents - directly, in peer-reviewed studies across multiple vertebrate animal models - is degradation of exactly these systems. Reduced acetylcholinesterase activity. Disrupted dopamine signaling. Neuroinflammation in the PFC specifically. Impaired impulse inhibition. Social behavior deficits.
The prior signal in this series (Signal #19, The Double Invasion) identified PFAS and MNP co-exposure as producing synergistic neurotoxicity via the PINK1/Parkin mitophagy pathway - the cellular cleanup mechanism that removes damaged mitochondria from neurons. Neurons in the PFC are among the most metabolically demanding in the brain. When mitochondrial repair fails in PFC tissue, the inhibitory circuits degrade first.
The hypothesis is not complex:
If MNP/PFAS exposure is progressively degrading prefrontal inhibitory function at population level -
And domestic violence perpetrators are a subpopulation with existing neurological risk factors (trauma history, substance use, prior violence patterns) that would make them more sensitive to additional PFC impairment -
Then we would expect to see not just more domestic violence, but specifically more lethal domestic violence: more episodes where the brakes fail to engage before the irreversible act.
Which is exactly what the FBI data shows.
Not more violence episodes with the same lethality rate. A change in lethality specifically - incidents crossing the threshold from assault to homicide at a higher rate.
The firearm finding is consistent with this framework. Firearm homicides drove the increase; non-firearm homicides barely changed. This is what degraded impulse inhibition looks like in a heavily armed population: not more stabbings or beatings, but more trigger pulls that would previously not have occurred because the inhibitory system intervened in the fraction of a second between intention and action.
The gun does not cause the failure to stop.
It converts the failure to stop into a homicide.
THE CONVENTIONAL ALTERNATIVES: EXAMINED
Several alternative explanations deserve honest consideration.
GUNS: The increase in firearm homicides could reflect increased civilian firearms ownership (which did surge during the pandemic) rather than any change in perpetrator behavior. More guns in homes with domestic violence = more lethal outcomes from the same episodes.
Assessment: Partial contributor. US civilian gun ownership did increase significantly in 2020-2021. This likely contributed to elevated lethality. However, gun ownership does not explain why the rates continued rising in 2022-2024 as the ownership surge plateaued, nor does it explain the change in escalation behavior documented in qualitative DV research.
REPORTING IMPROVEMENTS: Better tracking, improved FBI data collection methodology, and increased willingness of communities to classify deaths as DV-related could artifactually inflate apparent increases.
Assessment: Real but insufficient. FBI methodology does evolve. However, a doubling of recorded homicides-- not incidents, but deaths - is difficult to attribute primarily to counting changes.
Dead bodies are reliably counted. The murder rate and its relationship to intimate partner violence has been tracked for decades. A 119% increase in five years is not a statistical artifact.
ECONOMIC INEQUALITY: Persistent economic stress post-pandemic, concentrated in communities already at highest DV risk, could sustain elevated violence even after acute pandemic conditions resolved.
Assessment: Real partial contributor. Economic stress is a genuine ongoing risk factor. However, economic conditions broadly improved in 2022-2023 in most metrics, while homicides continued climbing. Economic explanation faces the same timeline problem as the pandemic explanation.
CULTURAL/SOCIAL FACTORS: Political polarization, erosion of institutional norms, social fragmentation, and media environment changes could elevate violence risk generally.
Assessment: Possible contributor of unknown magnitude. This explanation is difficult to falsify and difficult to quantify. It is also the explanation that would not implicate any environmental factor and would therefore be strongly preferred by those disinclined to examine environmental causation.
None of these alternatives are wrong. All are likely partial contributors. None provides a mechanistic explanation for why lethality specifically - the failure of inhibitory systems at the moment of irreversible escalation - would increase at this magnitude and persist after acute stressors resolved.
THE POPULATION-LEVEL LENS
This signal, like every signal in this series, is not a claim about individuals. It is a claim about population distributions.
Most people, even under severe stress, do not commit domestic homicide. The population of people who do is small. What matters for the signal is not any individual’s behavior, but the shape of the distribution: how many people are operating close enough to the lethality threshold that small changes in inhibitory capacity tip them across it.
If MNP/PFAS neurotoxicity is shifting the population distribution of impulse inhibition - slightly degrading PFC function across a large number of people - it does not cause most people to become violent. It shifts the tail of the distribution: the subpopulation already at the margin of lethality becomes somewhat larger, and the subpopulation that would have stopped short of killing is somewhat smaller.
At population scale, a small shift in the distribution of inhibitory function produces a measurable change in the homicide rate.
This is not speculation about whether specific perpetrators had elevated MNP brain loads. It is a claim about what population-level neurological change looks like when expressed through pre-existing patterns of intimate partner violence.
The perpetrators who crossed the threshold in 2022, 2023, and 2024 - after the pandemic excuses were no longer operative - are not a mystery that requires new criminals. They are the same population of domestic abusers that existed before, now including a larger fraction for whom the brakes did not hold.
THE SIGNAL WITHIN THE SIGNAL
There is a specific pattern in the intimate partner homicide data worth examining closely.
The 10-year trend from 2014-2023 shows a 22% overall increase, but with firearm homicides rising 36% and non-firearm homicides rising only 3%.
This divergence matters. Non-firearm domestic homicides require sustained physical action - strangulation, beating, stabbing - that gives the inhibitory system multiple opportunities to interrupt. The act takes time. The brakes have more chances to engage.
Firearm homicides can complete in a fraction of a second. If the brakes are degraded but still partially functional, the effect would be exactly this: a larger increase in fast, irreversible acts (firearms) than in slow, interruptible acts (non-firearms).
The data shows this pattern precisely.
This is not proof of the MNP hypothesis. It is consistency with it.
THE CONNECTIONS
This signal is part of a cluster. The same 2020 threshold appears across multiple independent behavioral signals:
Signal #20 (The Dog Signal): Fatal dog attacks roughly doubled from 2019 to 2022, with the increase appearing in stray dog populations that cannot be explained by human behavioral factors. The parallel to the domestic homicide data is direct: both show threshold crossings at the same period, in behavioral domains governed by the same neurological systems.
The cognitive decline signals document the mechanism: progressive MNP/PFAS accumulation in PFC tissue, disrupting the inhibitory circuits that govern impulse control and behavioral regulation.
The violence signals are not a separate phenomenon. They are the behavioral expression of the cognitive decline - the observable downstream consequence of what happens when the prefrontal brakes degrade at population scale, expressed through the situations where those brakes matter most.
In the overcrowded apartment.
In the argument that, once, would have stopped short.
THE HONEST LIMITS
The causal claim here - that MNP/PFAS neurotoxicity is a contributing factor in the domestic homicide increase - is not established in the peer-reviewed literature. Yet. No study has measured MNP body burden in domestic homicide perpetrators and compared it to controls. No study has directly linked population-level MNP exposure trajectories to intimate partner homicide rates.
What exists is:
1. A documented, robust increase in domestic homicides that persists after pandemic explanations expire.
2. A documented mechanism (MNP/PFAS PFC neurotoxicity) that would produce exactly this kind of lethality increase.
3. A convergent pattern across multiple behavioral signals (dogs, traffic, cognitive disability self-reporting) at the same 2020 threshold.
4. No competing mechanistic explanation that accounts for the post-pandemic persistence.
The honest position: this is a hypothesis with consistent supporting evidence and no available means of direct confirmation in current published research. It is the kind of hypothesis that the medical and epidemiological establishment should be designing studies to test.
They are not.
That gap - between the magnitude of the data and the absence of institutional investigation into environmental neurotoxicant contributions to violence - is itself a signal.
SOURCES
CBS News, December 2025: FBI domestic homicide data 2019-2024 (1,065 to 2,339)
TheWorldData.com compilation (December 2025): same FBI data corroborated
Everytown Research: 10-year trend 2014-2023, firearm vs. non-firearm intimate partner homicide breakdown
CDC / National Violent Death Reporting System: intimate partner firearm homicide data 2020-2023
PMC reviews (multiple): COVID-19 and domestic violence mechanisms, documented increases 2020
CDC NISVS: 1.8 million DV victimizations 2024
ScienceDirect (2024): MNP neurotoxicity review, PFC disruption, impulse control impairment in animal models
Frontiers in Neuroscience (2025): MNP behavioral effects, anxiety-like behaviors, social impairments
Signal #19 (The Double Invasion): PINK1/Parkin mitophagy pathway, PFAS/MNP synergistic neurotoxicity