Proposition 25 – referendum Allowing legislation Replacing existing cash bail system with one based on offender risk assessment to go into effect.

The first step in analyzing Proposition 25 is to understand that it is not a “normal” proposition.  It is a legislative referendum – that is a critical difference because it yields the opposite presumption (for us) from a routine proposition.  A referendum is another weird California voting mechanism that allows a group – here the bail industry – to gather voter signatures to challenge legislation that is passed by the Assembly and Senate.  Where, as here, enough signatures are gathered, a referendum stops an existing law (this one signed by our favorite Governor – Brown) from going into effect unless it is approved by a majority of voters.  So, unlike most proposition, this one was drafted by our legislature, in consultation with all of the groups that provide input for such legislation, and signed into law.  Thus, our bias favoring legislative action leads us to start with a somewhat favorable position on referenda like this one.  

The next step is to understand the concept of cash bail.  The system stems from our concepts of presumed innocence.  The idea is that one should not be incarcerated until he or she has been convicted – absent some pressing need (like danger to others, danger of flight, etc.).  Our justice system is based on the old British system – there, the concept of cash bail did not exist, though there has commonly been some way of vouching that the accused would show up for trial.  Cash bail does certainly incentivize appearance at trial, but it also creates a disparate system where persons with wealth who are accused of a crime experience much different treatment than those who are poor.  Moreover, California’s bail schedules are high – with a median bail amount of $50,000 we sit at five times the national average.  That means that with a median household income of $71,228, most Californians would be unlikely to readily afford the median amount of bail.  Worse yet, for the vast majority of offenders who need to borrow from a bail bondsperson (the folks who funded this proposition), they will commonly pay a non-refundable 10% fee every year that bail is “put up” by the bonding company.     

The data on pre-conviction incarceration suggests a problem.  For example, a small study in New Jersey that experimented with removing the cash bail system, found a 36% drop in incarceration rates.    A similar study found that 12% of New Jersey’s jail population – 1500 people – were in jail because they could not come up with even $2500 in bail money.  Id.  Nationwide, about 440,000 persons – fully 20% of those incarcerated in jails – are there because they cannot make bail.  (Id.)   There is also some statistical evidence suggesting incarceration, especially pre-trial, comes with a number of other detriments.  One study found that an offender’s pre-trial incarceration correlates to the offender being four times more likely to receive a sentence of imprisonment (State Prison vs. Jail), three times more likely to be given a lengthier sentence than an offender who was not incarcerated, 56% more likely to re-offend prior to trial, and 51% more likely to recidivate (commit more crime) after sentence completion.  Now, we should point out these figures show correlation – not causation.  Nevertheless, the figures do give us pause.  A large study of 153,000 prisoners over a long period of time, that did attempt to compensate for factors such as risk propensity, violence propensity, offense type etc., still concluded that incarcerated offenders were three to five times more likely to suffer longer sentences and were more likely to be sent to prison.  Incarceration also has tangible public costs.  Incarceration rates nationwide have been increasing for decades as have the resulting costs – between 1980 and 2010 the per capita expenditures (amounts each of us pay) dedicated to incarceration have more than tripled, from $77 per person to $260 per person annually. 

In 2016, California Supreme Court Chief Justice Cantil-Sakauye established a yearlong working group to study California’s cash bail system.  Over its one-year period, 11 California judges, one court administrator and 40 outside groups studied the system and reached this consensus conclusion: 

“After spending a year studying the issue, it became clear that the current system of money bail fails to adequately address public safety and the profound negative impacts on those individuals who should not be detained before trial,” said Judge Brian J. Back, the workgroup’s co-chair and a judge at the Ventura County Superior Court, one of the California courts that use risk assessment tools. “Thousands of Californians who pose no risk to the public are held in jail before trial, while others charged with serious or violent offenses may pose a high risk and can buy their freedom simply by bailing out…”

That strikes us as a strong indictment from the most knowledgeable, independent, Californians.  We should also note that the current COVID crisis has pushed California counties to experiment with no cash bail, with courts urged to release offenders in order to reduce dangerous jail overcrowding.  It is too soon to tell if that experiment has been a success, but it does not seem to have been disastrous.  That said, we should point out that some States – notably New York – that have experimented with different versions of removing or limiting cash bail have suffered some setbacks; with some stories of offenders repeatedly re-offending pre-trial. 

If passed, proposition 25 would allow previously enacted SB-10 to become the law of the land.  That law does away with cash bail and replaces it with a “risk algorithm.”  The algorithm is intended to classify offenders into low risk, medium risk, or high risk.  Low risk offenders would generally be released pre-trial and medium to high-risk offenders would be more likely to be incarcerated.  Both the District Attorney and the offenders would have the right to argue why incarceration should or should not be ordered if they disagreed with the assessment.  

Opposition to the proposal comes not only from those who want to maintain or increase incarceration rates (an argument we dispose of quickly) but also from those who fear that putting the decision to release into the hands of an algorithm, or even a judge is also risky.  We do not necessarily disagree; there could be racial and other bias issues that could arise with the creation of such algorithms and that problem is not necessarily solved by inserting human judges into the analysis.  Still, we have to believe that removing the “gating” factor of money is positive.  Especially with the apparent detriments we discussed above associated with pre-trial incarceration we think that using money as the determining factor is a bad call.  While there could also be problems with the new proposed technique, we would entrust the legislature to implement fixes to the system as they arise – a luxury we have here because this is not a constitutional amendment or other “normal” proposition.  

Faced with the proposal of backing our legislators and one of our favorite all-time Governor’s carefully considered effort to address a long-time problem we choose to move forward; understanding that in doing so we are engaging in something of an experiment that includes risk.  There will doubtlessly be criminals that commit more crime but we are hopeful that the vast majority of offenders – those that are still capable of rehabilitation – will benefit from changing the system to focus more on justice than expediency.  

Strong YES 

  1. #1 by Evan Hoffman on October 31, 2020 - 10:41 am

    I agree that this is a good step. However, I am interested in learning more about the mechanics and inputs into the algorithm. For example, will the algorithm omit factors such as race, sexual orientation or income?

    • #2 by politoesq on October 31, 2020 - 12:48 pm

      Great question – and the primary reason some progressive groups oppose the measure. The State legislation directs the various counties (all 58 in California) to create these algorithms. So, they do not currently exist. There is data showing that even algorithms that are intended to remove human biases, can themselves implement the implicit biases of their programmers or society at large. We think that will be something that will need to be monitored and probably will lead to adjustments over time. This relates to something else we liked (but dropped on the editing floor – though you will see it in our citations) about this legislation: The law gets automatically reviewed in 2023. While we would not be this generous if this were not a referendum, we think the imperfect here should not be the enemy of the good. Getting rid of overt discrimination on the basis of financial means is a positive, needed, step in our minds. Add’l Sources: https://calbudgetcenter.org/wp-content/uploads/2020/08/CA_Budget_Center_Money_Bail_Prop25_Aug2020.pdf
      https://calmatters.org/politics/california-election-2020/2020/10/cash-bail-justice-algorithm-risk-assessment-prop-25/

  1. Summary of Nov 2020 Proposition Analysis | PolitoMuse

Leave a comment