Ta uni

POLICY BRIEF
June 2018
Alena Yarmosky

The Impact of Early Representation: An Analysis of the
San Francisco Public Defender’s Pre-Trial Release Unit
In October 2017, the San Francisco Public Defender’s Office piloted the Pre-Trial Release Unit (PRU) to
enhance access to pre-arraignment legal representation for indigent arrestees. Using data provided by the
Office, this study finds the pilot program doubled the likelihood of release at arraignment – from 14% to 28% –
for arrestees who received arrest-responsive interventions from the PRU. The intervention is projected to
save approximately 11,200 jail bed-days per year at an annual cost of approximately $335,000. i Furthermore,
the PRU’s efforts to advocate for the dismissal of parole holds reduced pre-trial incarceration by 44%, or an
average of 9.5 days, among eligible parolees who were held in custody for violation of their parole orders.

Context
When individuals are arrested, they are often held
in jail until their arraignment hearing (the first
time a defendant is brought before a judge). At
arraignment, the court decides whether an
individual should be held in custody or released
pre-trial, with or without court supervision. Public
defenders traditionally provide representation for
indigent defendants starting at arraignment.
However, the pre-arraignment period is critical
for a number of reasons: bail is set, formal
charges are filed, and case investigation begins.
Individuals who can afford a private attorney
immediately after booking have access to
services that may increase the likelihood that
they will be released from jail prior to
arraignment, or that their charges will be
dropped altogether. Indigent arrestees – who are
not provided a public defender until the
arraignment hearing – do not receive these
benefits.
capolicylab.org

Pre-trial release can have tremendous impact on
defendants’ lives and later case proceedings.
Defendants who are incarcerated pre-trial plead
guilty at higher rates, are more likely to be
convicted, and face longer sentences than
similarly-situated releasees.ii Researchers have
found that even a relatively short period in jail
pre-trial – as few as two days – correlates with
negative outcomes for defendants and for public
safety when compared to those defendants
released within 24 hours.iii

Overview of the PRU
Against this backdrop, the San Francisco Public
Defender’s Office began providing prearraignment representation to a subset of
criminal defendants in October 2017, in a program
called the Pre-Trial Release Unit (PRU). San
Francisco is among the first counties in the
United States to provide pre-arraignment
representation to indigent defendants.iv

1

In addition to the PRU, San Francisco has enacted
other important policy and programmatic
changes to make pre-trial release more equitable.
For example, the San Francisco Pretrial Diversion
Project supports pre-trial release through various
programs. In 2016, San Francisco adopted the
Public Safety Assessment (PSA), which provides
judges with an evidence-based risk score to
inform release decisions and an alternative to
money bail. Initial analysis by participating
agencies suggests that the PSA has led to a
decrease in the number of pre-trial detainees and
the frequency of pre-trial detention.v
One rationale motivating these multiple
initiatives is San Francisco’s effort to avoid the
construction of a new jail by reducing the overall
county jail population by 83,0202 jail bed-days per
year.vi The county was further motivated to
establish the PRU as it believed that providing
pre-arraignment representation could reduce
wealth-based inequities in access to justice.
The PRU provides two primary interventions.
Clients arrested on new criminal activity may
receive “arrest-responsive” interventions
designed to help build their case, which include
client interviews, case investigation, notification
of a prior attorney of record, family/friend
contacts, and recruitment of community
members to attend arraignment.
The PRU also provides “parole advocacy” when
the primary reason behind the detention is
violation of one’s parole orders. Parole advocacy
involves PRU staff directly contacting agents to
advocate for dismissal of clients’ parole holds.
During our five-month study period, the PRU
provided services to a subset of indigent
defendants (1,024 unique cases). Two attorneys
and one investigator provided PRU arrestresponsive services in an average of 42 cases per
week. The cost of the program was $335,000 for
the first year. Given resource limitations, PRU
staff prioritized defendants with more serious
booking charges and more extensive criminal
histories, when possible. Parole advocacy was
capolicylab.org

provided unprioritized to every defendant for
which unit staff had time (231 out of 308 eligible
parolees, or 75 percent of cases). Figure 1
provides a breakdown of interventions provided.
Figure 1: PRU Services, by Intervention Typevii

Methodology
To quantitatively assess the impact of the PRU on
length of pre-trial incarceration, we generated a
dataset of booking, charge, and demographic
information for all arrestees booked into county
jail during our study period (October 2, 2017 February 28, 2018) from the Public Defender’s
GIDEON case management system. We then
merged this booking data with PRU treatment
information, coded by intervention type.
To compare outcomes for those who received
PRU services to those who did not, we used a
propensity score matching approach to minimize
differences between treated and non-treated
arrestees. The propensity score indicates the
likelihood that a client receives arrest-responsive
PRU treatment given: age, race, gender, severity
of booking charge, out-of-county warrants,
parole or probation holds, and criminal history.
We then used a “nearest neighbor” matching
technique to match clients treated by the PRU
with similarly-scored defendants who did not
receive treatment.
Because there was little selection bias associated
with parole advocacy, we used a regression
2

model to measure impact of parole advocacy on
eligible parolees’ length of incarceration.viii

Findings
Our findings suggest that the PRU has
demonstrated promising initial success in
decreasing the length of pre-trial detention.
Specifically, our analysis reveals that individuals
who receive arrest-responsive services are twice
as likely to be released at arraignment when
compared with similarly situated, non-treated
arrestees. Similar, not-treated arrestees are
released at arraignment 14 percent of the time,
compared to a 28 percent rate for treated
arrestees. Though results were consistent in
several robustness tests, confirmation of this
result via randomized trial would strengthen the
causal nature of the finding.

100%
44%

Release at arraignment for those receiving
arrest-responsive services
(significant with 99.9% confidence)
Length of pre-trial incarceration for clients
receiving parole advocacy services
(significant with 97.5% confidence)

approximately 4,689 jail bed-days during its initial
5 months of operation or an average of 11,253 jail
bed-days saved per year.
Finally, we found that parole advocacy as an
independent intervention significantly reduced
the length of incarceration. Among eligible
parolees, parole advocacy provided by the PRU
reduced the average length of pre-trial
incarceration by 44%, or 230 hours (approx. 9.5
days). Interviews also suggested that parole
advocacy increases the speed at which parole
holds are lifted and reduces the number of parole
petitions filed.
These promising findings suggest that other
jurisdictions may wish to experiment with early
representation. We suspect that the impact of
early representation may be even larger in a
jurisdiction that has not undertaken extensive
efforts to reduce pre-trial detention.

Further Research
Our findings indicate that pre-arraignment
representation significantly impacts the
likelihood of release at arraignment. We
recommend that the Public Defender’s Office
repeat this analysis at the PRU’s 18-month mark
to confirm our findings with a larger sample size.

Interviews with the Public Defender’s Office
suggest this result may be due to attorneys’
increased ability to argue for release at
arraignment, including increased access to client
information, early case investigation, and the
presence of community members at arraignment.

The California Policy Lab builds better lives through
data-driven policy. We are a project of the University of
California, with sites at the Berkeley and Los Angeles
campuses.
This research publication reflects the views of the
author and not necessarily the views of our funders,
our staff, our advisory board, the Regents of the
University of California, or the San Francisco Public
Defender’s Office.

Using a rough extrapolation, we estimate that
the PRU’s arrest-responsive treatment saved
i

Annual cost retrieved from: Office of the Controller, City &
County of San Francisco. (2018). Evaluation of Pilot
Programs Funded to Reduce the Jail Population.
ii
Dobbie, W. et. al. (2018). The Effects of Pre-Trial Detention
on Conviction, Future Crime, and Employment: Evidence
from Randomly Assigned Judges. American Economic
Review, 108(2), 201-240. doi:10.1257/aer.20161503
iii
Subramanian, R. et. al. Incarceration’s Front Door: The
Misuse of Jails in America. Vera Institute of Justice,
(February 2015).

capolicylab.org

iv

Miami-Dade County, FL began providing early
representation in 2013.
v
CPL is partnering with the San Francisco Sheriff’s
Department to analyze the PSA’s effect on pre-trial release.
vi
Work Group to Re-envision the Jail Replacement Project.
(2017). Final Report.
vii
The category “all other” includes the following
interventions: outside contacts (91 cases), in-person
arraignment recruitment (28 cases), in jail referrals (19
cases), and bail advocacy (4 cases).

3