When Justice Hits a Dead End: The Mary Fong Lau Case and America's Vehicular Homicide Problem

On March 16, 2024, Mary Fong Lau drove her Mercedes SUV at 70 miles per hour—the wrong way down a residential San Francisco street. She killed an entire family: Diego Cardoso de Oliveira (40), his wife Matilde Moncada Ramos Pinto (38), and their two infant sons, one-year-old Joaquim and three-month-old Cauê. They were waiting for a bus to the zoo.

On February 13, Lau pleaded no contest. The judge indicated she’ll receive probation. No prison time.

The Legal Gymnastics

Lau’s strategy was textbook asset protection:

1. Immediately after the crash: Transfer millions in real estate into a new LLC; sell other properties to family members
2. Plead no contest instead of guilty: Avoid creating evidence for civil lawsuits
3. Accept conviction without admission: Get sentenced while preserving deniability
4. Let age and “remorse” do the work: Judge decides 80-year-old dying in prison isn’t proportional

The prosecutor objected. Victims’ families are “appalled.” But it worked.

Why No Contest Matters

When you plead guilty, you admit you committed the crime. That admission becomes evidence in civil court—devastating for wrongful death lawsuits.

When you plead no contest, you accept punishment without admitting anything. The criminal case resolves, but victims’ families must independently prove every element of negligence in civil court. No shortcuts. No smoking gun admission.

For someone who transferred assets into LLCs immediately after killing four people, this isn’t about remorse. It’s about limiting liability.

The Intent Gatekeeping Problem

Prosecutors defend light sentences for vehicular homicide with a consistent refrain: the driver didn’t intend to kill anyone. Without intent, they argue, murder charges are off the table. Life sentences are off the table. The defendant is a negligent person, not a murderer.¹

In legal terms, this is an argument about mens rea — the guilty mind, the mental state the prosecution must prove to secure a conviction. Murder requires malice aforethought. Vehicular manslaughter requires only criminal negligence or gross negligence. Negligence is a far lower mens rea, which is why it carries a far lower sentence. The prosecutor’s implicit claim is that the defendant’s mental state at the time of the act reaches negligence but not malice — so manslaughter is the right charge, murder is off the table, and a decade in prison (if that) is the appropriate consequence.

This argument has one problem: the law already rejected it.

Depraved indifference murder — recognized in most U.S. jurisdictions and codified in the Model Penal Code — exists precisely to handle cases where the defendant had no intent to kill but whose conduct was so reckless it satisfies the malice mens rea anyway.² The doctrine holds that implied malice — conscious disregard for human life — is a fully valid mental state for second-degree murder. Courts developed it because they recognized that some conduct is so obviously lethal that classifying it as mere negligence becomes a legal fiction. Firing a gun into a crowd. Dropping boulders from an overpass. Driving 120 mph through a populated area while high on methamphetamine.

The doctrine doesn’t ask whether you wanted anyone dead. It asks whether your conduct demonstrated that you simply didn’t care if someone died. That is a mens rea finding — implied malice — and it supports a murder charge.

Driving 70 mph the wrong way down a residential street satisfies that standard. Driving with a BAC of 0.24 at highway speeds satisfies that standard. The prosecution’s mens rea problem isn’t that implied malice is absent — it’s that proving it requires more work than charging manslaughter.

The “no intent” defense isn’t a legal barrier. It’s a rhetorical habit that prosecutors allow to go unchallenged, judges validate through lenient sentencing, and the public accepts because we’ve been conditioned to treat cars as a special category of risk, where death is tragedy rather than consequence.

The Exception That Proves the Rule

There is one vehicular context where the system does construct implied malice: repeat DUI. California’s Watson doctrine, established in 1981, holds that a driver convicted of DUI and given the formal advisement that drunk driving can kill — then kills someone in a subsequent DUI — can be charged with second-degree murder.³ The prior conviction and advisement together establish the mens rea directly: you had documented knowledge that this conduct was lethal, you chose to do it anyway, someone died. The system calls that conscious disregard for human life. It calls it implied malice. It calls it murder.

Notice what Watson actually proves. The underlying principle isn’t that alcohol is uniquely evil or that DUI is categorically different from other reckless driving. The principle is: documented prior knowledge of a lethal risk + deliberate choice to create that risk = implied malice mens rea. The prior conviction is just the mechanism the system uses to establish that knowledge on the record.

But that mechanism is unnecessary when the knowledge is intrinsic to the act itself. Nobody needs a prior conviction or a formal advisement to know that driving 70 mph the wrong way into oncoming traffic will kill people. The lethality is self-evident — which means the mens rea is self-evident — in a way that’s arguably stronger than the Watson framework, where a defendant at least had to be formally told that DUI is dangerous.

The logic the system used to satisfy the mens rea requirement for repeat drunk drivers applies with equal or greater force to sober drivers who commit acts of equal or greater obvious lethality. The question isn’t why DUI gets treated as potential murder. It’s why that reasoning stops at alcohol.

The Constitutional Barrier — and the Outcome-Based Reality

When prosecutors resist life sentences for extreme vehicular negligence causing mass death, they invoke proportionality: the Supreme Court has held that punishment must be proportionate to culpability (mental state), not just harm (outcomes).⁴ The death penalty, they note, requires intent to kill or depraved indifference so extreme it approximates intent — gross negligence alone doesn’t qualify. This sounds like a principled constitutional limitation.

It’s also a lie about how the system actually works.

The criminal justice system already runs on outcome-based sentencing for vehicular crimes. Consider DUI in California:

• First DUI (no injury): Misdemeanor. Up to 6 months in county jail, typically probation.⁵
• Same DUI + someone dies: Felony. 4-10 years for gross vehicular manslaughter, or 15 years to life for Watson murder.

Same driver. Same blood alcohol level. Same choice to drive drunk. Same mens rea. The only difference is whether someone happened to be in the crosswalk when the car ran the red light. That’s pure outcome-based sentencing — a variance of up to 30x or more in prison time based solely on whether death occurred, not on any difference in mental state or moral culpability.

If the system already punishes outcomes this aggressively in vehicular contexts — if killing someone while drunk can multiply your sentence by decades compared to identical conduct that merely endangered people — then the proportionality objection to outcome enhancements for multiple deaths is simply incoherent. The system has already decided that outcomes matter enormously when cars kill people. The only question is whether that principle should scale with body count.

Life without parole isn’t constitutionally barred for negligence causing mass death. The barrier is legislative, not constitutional. States could enact outcome-enhanced sentencing for extreme recklessness:

• Extreme negligence (wrong-way, high speed) + 1 death = 15-25 years
• Same negligence + 2+ deaths = 30 years to life
• Same negligence + 4+ deaths = life without parole

This is constitutionally sound. It mirrors the outcome-based framework the system already applies to DUI. It just requires political will.

What Needs to Change

1. Mandatory Charging Standards

Specific scenarios should trigger mandatory depraved indifference charges:

• Wrong-way highway driving above 60 mph
• Street racing resulting in death
• DUI with prior convictions resulting in death

Remove prosecutorial discretion to plea down to manslaughter.

2. Catastrophic Outcome Enhancements

Multiple deaths should dramatically increase sentencing:

• Current: Most states treat 1 death vs. 4 deaths identically (concurrent sentences)
• Proposed: 2+ deaths triggers mandatory 30-years-to-life; 4+ deaths triggers life without parole

3. Close the No Contest Loophole

For vehicular homicide cases, no contest pleas should require:

• Judicial finding that defendant lacks resources warranting civil protection, OR
• Waiver of statute of limitations on civil claims, OR
• Prohibition entirely for cases involving 2+ deaths

4. Age Shouldn’t Trump Body Count

“She’d die in prison” is a reason to avoid incarcerating 80-year-old shoplifters. It’s not a reason to give probation to someone who killed four people through extreme recklessness.

If you’re too old to survive prison, you’re too old to drive 70 mph the wrong way.

The Broader Pattern

The Lau case isn’t an outlier. It’s emblematic of a two-tier justice system where wealth fundamentally alters outcomes in vehicular homicide cases.

Case Study: The “Affluenza Teen”

Ethan Couch killed four people in 2013 while driving drunk at age 16, with a blood alcohol level three times the legal limit.⁶ Prosecutors sought 20 years. His defense introduced the term “affluenza” — arguing his wealthy upbringing left him unable to understand consequences. Judge Jean Hudson Boyd sentenced him to 10 years probation and therapy at an upscale facility.

When he violated probation by fleeing to Mexico, he was sentenced to 720 days in jail — 180 days for each victim. Total time served for killing four people and paralyzing another: under two years.

His probation ended in December 2023. Ethan Couch is a free man.

Case Study: Rebecca Grossman

Rebecca Grossman, co-founder of the Grossman Burn Foundation, drove up to 81 mph through a marked crosswalk in Westlake Village in 2020, killing brothers Mark (11) and Jacob (8) Iskander.⁷ She fled the scene.

Unlike Lau and Couch, Grossman was actually convicted of second-degree murder — a rarity in vehicular homicide cases. Prosecutors sought 34 years to life. The judge sentenced her to 15 years to life, citing her lack of criminal record and philanthropic history.

She’s eligible for parole as early as 2033 — potentially serving nine years for two murders.

The judge’s reasoning? Her wealth-funded philanthropy and clean record warranted leniency.

The Mechanics of Inequality

This pattern repeats across jurisdictions:

Immediate asset protection: Wealthy defendants transfer property into LLCs, sell to family members, restructure holdings — sometimes within days of crashes. This isn’t criminal; it’s estate planning expertise that costs six figures and requires sophisticated legal teams most defendants can’t afford.

Elite legal representation: Couch’s family paid for expert psychologists who invented novel defenses. Grossman’s defense team blamed her ex-boyfriend and argued the victims were “out of the crosswalk.” These strategies require resources: expert witnesses ($10,000-50,000 per trial), private investigators, jury consultants, media management.

Sentencing theater: Judges consistently cite “lack of criminal record” and “community ties” as mitigating factors. Judge Brandolino noted Grossman’s philanthropic history, as if charitable foundations funded by family wealth should offset killing two children.

The plea bargain advantage: No contest pleas preserve civil defense options. Wealthy defendants can afford the parallel civil litigation; poor defendants take guilty pleas for immediate sentence reductions because they can’t fight on two fronts.

The data bears this out: Studies show vehicular homicide sentencing varies wildly based on defendant resources, not harm caused.⁸ Sentences for killing one person range from probation to 20+ years, with wealth and race as stronger predictors than blood alcohol level or speed.

The Legal Framework Already Exists

The doctrine isn’t theoretical. Courts have already established every piece needed to charge extreme vehicular recklessness as murder.

On cars as deadly weapons: California’s People v. Aguilar (1997) established that a vehicle qualifies as a deadly weapon when “used in a way to cause substantial, significant or great bodily injury or death.”⁹ Minnesota’s Supreme Court ruled in 2024 that recklessly driven cars are “dangerous weapons” when operated in a manner “likely to produce death or great bodily harm.”¹⁰ These aren’t edge cases — courts routinely uphold assault with a deadly weapon charges for vehicle attacks.

On implied malice without intent to kill: already established, codified, and applied in vehicular contexts through the Watson doctrine.²³

Then look at what prosecutors actually charged in three cases involving comparable conduct — high speed, multiple deaths, no meaningful intent to kill:

• Rebecca Grossman: Second-degree murder (rare). Convicted. Sentenced to 15 years to life — less than half the 34 years prosecutors sought. Eligible for parole as early as 2033.
• Mary Fong Lau: Vehicular manslaughter. 70 mph wrong way, 4 deaths. Judge indicated probation.
• Ethan Couch: Intoxication manslaughter. BAC three times the legal limit, 70 mph in a 40 zone, 4 deaths. Probation.

Grossman’s case shows the framework works when prosecutors use it. Lau and Couch show what happens when they don’t. The variance isn’t in the law — it’s in the will to apply it.

The Sister’s Words

Denise Oliveira, sister of Diego Cardoso de Oliveira, said it best: “It feels like we have no rights. I feel deeply disrespected by this process. It doesn’t feel like justice.”

She’s right. It isn’t.

Update: Lau’s formal sentencing is scheduled for March 20, 2026 — the second anniversary of baby Cauê’s death. Victims’ families continue to pursue civil litigation against shielded assets.