The unlockable prison gate

However, the resentencing did not merely dismiss the firearm enhancement. It downgraded the murders from first degree to second degree, even though the court did not hear any evidence about the murders or find any errors regarding the convictions, and there was no relevant change in special circumstance sentencing law. Had the jury not found firearm use, there would have been no ground for resentencing. But because it did, parole for Frederickson now is not prohibited but presumptive. He can be released--without delay.

The court explained it rescinded the LWOP sentence because it "continues to believe that a man can turn himself around. And just try to better himself and better those around him." But the purpose of the initiative was to implement the public's belief that the "worst of the worst" deserved LWOP (or death), and to prevent a judge's personal belief that a double murderer can "try to better himself" from overriding the public's unambiguous conclusion that someone who has murdered twice should never get a chance to notch number three. Through resentencing, the judge unilaterally subverted the state law excluding multiple murderers from parole eligibility.

The disfavored sentence

Frederickson accords with judicial professionals' distaste for both lay decision-making (as voters and jurors) and the LWOP sentence itself. To evade public demand for serious punishment (and provide "expert" authority for sentence dilution), California created the Committee on Revision of the Penal Code (CRPC). Its 2021 report cited a study asserting "life without parole sentences do not result in any greater public safety benefits than life with parole sentences." Kleinstuber & Coldsmith, "Is Life Without Parole an Effective Way to Reduce Violent Crime? An Empirical Assessment," Crim. & Publ. Pol'y 19(2), 625 (2000). It is counterintuitive to conclude a community is just as safe when murderers are released as when they are confined. It is also wrong.

The study's conclusion derived from correlation: states with more LWOP-serving inmates do not tend to have significantly lower rates of violence. Yet the study itself recognized there could be reverse causation: "a state may have a high LWOP population ... because the state had a high rate of violent crime." In other words, that more people lock their doors and use burglar alarms in high-crime areas does not prove that locks and alarms have no effect on crime.

Another objection is the supposed arbitrariness of the LWOP penalty, as nearly all first-degree murders could be charged as special-circumstance murder (just as nearly all murders could be charged as first degree). But this flexibility is a feature, not a bug. It ensures the measure of punishment rests with jurors and elected prosecutors, based on the crime. By contrast, the CRPC recommends sentence review for all LWOP inmates, ensuring punishment depends on unelected parole administrators and judges, based on inmates' expressions of remorse.

Of course, the best way to eliminate arbitrary sentencing is to follow the federal rule and deny parole eligibility to all first-degree murderers. United States v. Dencklau, 160 F.4th 1046, 1062 (9th Cir. 2025).

The most explosive objection is that LWOP sentencing is racially biased. Justice Kelli Evans's dissent in People v. Hardin, 15 Cal.5th 834 (2024), cited a Human Rights Watch report contending African Americans are sentenced to LWOP at 18 times the rate for Whites. But footnote 38 of that report derived that "rate" by dividing the LWOP population of each race by its total population in the state. But the relevant comparator to establish sentencing bias is not the statewide population of each racial group but the population convicted of first-degree murder (without a special circumstance).

And CRPC's own data support the opposite conclusion. Its 2021 report lamented that "79 percent of people serving [LWOP] are people of color" but just three pages later recounted the overall first-degree murder numbers: 34% Black, 38% Latin, and 9% other, a total of 81%. Whites thus accounted for 19% of all first-degree murder sentences but 21% of those denied parole eligibility.

Of course, if the Supreme Court wants to examine bias, it should collect and publicize all data, rather than let advocates cherry pick their preferred numbers.

Justice Evans also objected to the effect of the felony-murder rule, citing a report co-sponsored by the Felony Murder Elimination Project, which found "Black individuals account for 37.2% of California's overall LWOP population [while] 42.7% of Black people [receive LWOP for] a felony murder special circumstance." Felony murders evoke special concern among the public because they tend to kill strangers; these killings are harder to solve and to protect against, and often endanger bystanders. Moreover, the willingness to kill a stranger predicts a greater willingness to re-offend than exists among killers who bear animus against one specific person. The felony-murder special circumstance might be wise or unwise, but it does not prove racial bias.

In any event, sentencing leniency does not advance racial fairness. Bias can more easily corrupt justice if parole is available (as in Frederickson) because judges and parole boards exercise subjective discretion about who deserves a third chance than it can if parole is unavailable because a unanimous jury objectively determined there were two murders rather than one.

Kleinstuber & Coldsmith revealed the real objection to permanent punishment: "These sentences are extraordinarily expensive." But how much should we pay to prevent homicides? Does the answer depend on whose murder will be prevented, and who's paying to prevent it? The fiscal costs of confining offenders are borne by the taxpaying rich, but the physical costs of releasing them are borne by the vulnerable poor.

Accordingly, the more democratic and inclusive the decision-making process, the longer the sentences tend to be. The electorate has repeatedly voted for permanent punishment, while unelected commissioners and judges repeatedly undermine it.

Federal constitutional principles

Because Frederickson involved California law, there is likely no recourse in federal court. But there should be. Just as voting "ensures the people's ultimate control in the legislative and executive branches, jury trial is meant to ensure their control in the judiciary." Blakely v. Washington, 542 U.S. 296, 306 (2004). Californians directly voted to punish multiple murderers with LWOP or death, and the jury found Frederickson was a multiple murderer. As Justice Gorsuch recently explained, the jury has the "constitutional authority to set the metes and bounds of judicially administered criminal punishments." United States v. Haymond, 588 U.S. 634, 646 (2019).

That authority is usually considered a constitutional ceiling on punishment, not a floor. But it is not clear why this authority should be a one-way ratchet on punishment--downward. Criminal justice rules do exist to protect criminal defendants, but also the process of justice itself.

In Batson v. Kentucky, 476 U.S. 79, 87 (1986), the Supreme Court barred racial discrimination in jury selection, finding it harmed the accused--and the jurors illegitimately excluded. And the harm extended beyond excluded jurors; it also "undermine[d] public confidence in the fairness of our system of justice." Because the fairness imperative protects not just the defendant but also the public, the Supreme Court later held the rule was reciprocal, restraining both prosecution and defense from dismissing jurors on discriminatory grounds. Georgia v. McCollum, 505 U.S. 42 (1992).

A juror's right to serve was not an asymmetric weapon for the defense. Jurors' right to sentence should not be one either.

When the Supreme Court outlawed the death penalty for juveniles, it linked citizens' capacity for civic self-government with defendants' capacity for personal self-government; juveniles' irresponsibility precluded them from voting and jury service, so it could also exempt them from maximum punishment. Roper v. Simmons, 543 U.S. 551 (2005). The reverse of this logic now extends to adults: they are shielded from permanent punishment as defendants, so courts need not respect their decisions as voters or jurors. They still formally vote and serve on juries, but only for show: their decisions may be evaded by judges who express optimism about convicts' prospects for bettering themselves--knowing that if they guess wrong, someone else's children will pay the price.

We have a right to vote on statewide initiatives and jury verdicts. But do we have the right for those votes to matter?

"Forget it Jake, it's California."