Multnomah County’s
Sentencing Support Project
Multnomah County’s sentencing support program is now running to
query information in a data warehouse presently including data extracted
from Portland Police, Multnomah County Sheriff, OJIN (Oregon Judicial Information
Network), and Multnomah County District Attorney databases. Oregon
Department of Corrections data will soon be included, with drug and alcohol
data and LEDS (Law Enforcement Data System) expected to follow close behind.
Juvenile Justice and other sources should join the warehouse in due course.
When the program first ran against real data, I sent an e-mail
to judges on an international mailing list analogizing this event to the
Wright Brothers’ first flight at Kitty Hawk: we have a long way to
go, but we’ve taken the first steps in a major transformation in decision-making
in courtrooms and, eventually, by policy makers.
The main objective of this technology is to equip all involved
in sentencing decisions with better information to support sentencing choices
most likely to reduce the criminal behavior of offenders. As
illustrated elsewhere on this site, a user identifies the offender
by case number, selects the charge of conviction, and views a bar chart
showing how sentencing choices for similar offenders sentenced in the past
for similar charges have correlated with reduced criminal behavior.
Tabs allow users to modify the defaults constructed by the program to match
the offender’s characteristics (employing age, criminal history, and other
variables), the crime for which sentences have been imposed, and the outcome
measure. For example, if the offender is being sentenced for Theft
II, the program will show how available sentencing elements (jail, programs,
probation, etc.) have “worked” when imposed on property offenders of similar
age, ethnicity, gender, and criminal history -- measured by five levels
in each of six categories of crime: violent, sex, property, drug, DUII
and major traffic, and domestic violence. By default, “worked” means
no new conviction within three years for an offense of the same category.
But users can alter all of this -- the category of crime for
which sentences are compared, the scope of recidivism (such as any conviction
or any police contact for any crime within five years), and all of the
defaults reflecting the characteristics of the offender. Part of
the power of this tool is that users can “recalculate” to produce new results
whenever a user changes any or all of the defaults.
This tool will enable advocates to address their arguments to
a major objective of sentencing typically largely ignored in the past,
but squarely within the objective of the Judicial Conference Resolution
from which this newsletter draws its title: reducing future criminal conduct.
Since the last issue, we’ve accomplished a great deal.
With the guidance of the focus group, program developers have added the
domestic violence “crime theme” to the profile, and have expanded the data
used to reach all OJIN data instead of merely bench probations. In
the near future, we hope to enhance the domestic violence criminal history
measure to include (optionally, and by default) police contacts recorded
as domestic calls with the defendant as the subject, and to add police
contacts to the options available to users in constructing an outcome measure.
In the longer run, we hope to extend the tool to probation revocation decisions.
The Multnomah County Department of Community Justice has received
a technical assistance grant under a cooperative agreement between the
National Institute of Corrections and the Center for Effective Public Policy
for improvement of pretrial release decision-making. Multnomah County’s
Sentencing Support tool was a major component of the grant application,
and the application of this tool to pretrial release has long been an articulated
goal of the project. We hope by this grant to increase support for
this approach to criminal justice decision making as well as support for
the tool itself.
At this point, we are working on several issues of data integrity,
linkage, and usability to get the product ready for routine use in courtrooms.
Several judges and several defense attorneys have been trained on the tool,
and the first prosecutor training is expected in the near future.
State-Level Developments
The Department of Oregon State Police has concluded its lengthy
negotiations with the Information Resources Management Division of the
Department of Administrative Services for management of the Public Safety
Data Warehouse, the project that will serve sentencing support queries
from courts on a state-wide basis. LEDS director David Yandell is
head of the project, and is moving it along a projected 33 month development
plan. The difficult hiring stage should complete soon, and actual
construction of the warehouse should be underway. In discussions
with Multnomah County’s DSS-Justice project, all involved have found Multnomah
County’s experiences extremely useful, as Multnomah County has encountered
a rich variety of issues sure to confront PSDW as it develops. Currently,
Multnomah County’s request for LEDS data has generated months of discussion
about security, which has involved Attorney General Hardy Myers and the
Assistant Attorney General assigned to the State Police, Keith Kutler.
David Yandell, Keith Kutler, Nancy Arnot (head of the Multnomah County
project), and I have met concerning these issues, which all recognize as
precursors to the PSDW itself. I’ve suggested an intergovernmental agreement
blueprint to facilitate brining new data providers on line, since each
new player has required examination anew of the same issues of security
and confidentiality.
By agreement, Keith and I are working on proposed legislation
that would appropriately position such data warehouse efforts as Multnomah
County’s and those of state agencies (and other counties and regions) with
respect to the Public Records Law. The present intent is to direct
PRL requests to public providers of information to data warehouses and
to public agency end users, while leaving warehouse administrators out
of the public disclosure business. The rationale is that data warehouses
are essentially conduits; information is copied for inclusion in the warehouse
without modification at the source, and gains any additional meaning only
when queried by end-users for their operational or analytical purposes.
Local and state warehouse managers are convinced that such legislation
can resolve many security and confidentiality issues.
Meanwhile, the Oregon Judicial Department remains committed to
providing sentencing support tools as part of the “integrated justice environment,”
based on its own data warehouse and, eventually, the Public Safety Data
Warehouse managed by the Oregon State Police. Although the OJD warehouse
is well underway, it has a long way to go before it can deliver sentencing
support tools to judges state-wide. The OJD warehouse as now operational
is designed primarily to allow efficient query of OJIN data [no small feat!]
to provide legislative support -- to answer questions and support explanations
to legislators interested in how efficiently we do whatever it is we do.
The warehouse will need to be “redesigned” to accomplish sentencing support
and other functions.
A central piece both to use of OJD’s warehouse and delivery of
sentencing support tools state-wide is completion of the long-delayed automated
uniform criminal judgment. Although that product has long been requested
by the Department of Corrections for its own good reasons, the automated
judgment will go a long way toward ensuring application of uniform data
entry rules, important to any analysis based on that data. OJD ISD
Director Carl Ward explains why the UCJ is taking so long: “the UCJ requires
data that can only be entered by GUI [graphic user interface] Calendaring,
Hold a Proceeding, and Add Criminal Case to work.” And all of this
requires an upgrade to 128M of RAM in our computers, according to our Trial
Court Administrator, Doug Bray.
Finally, the Department of Corrections has its own technology
issues, surrounding the IITP (Inmate Incarceration and Transition Plan).
As you may recall, many months ago, all judges received a glossy magazine-sized
pamphlet extolling the virtues of this approach: inmates would be assessed
at intake in terms of needs along several important lines -- substance
abuse, physical health, vocational training, mental health, educational
achievement, and so on -- and would be subjected to planned training, treatment,
and counseling during their projected incarceration to the end that they
would re-enter (or enter) the work force as productive citizens. IITP is
designed to track offenders on post-prison supervision as well as in custody,
and should supplant the old ISIS module that was virtually unused by Adult
Community Corrections in Multnomah County (which is developing its own
replacement). The IITP pamphlet said that judges and prosecutors
could access the database of assessments and progress for each inmate “now,”
but this turned out to be a bit optimistic. It took months to make
access even theoretically possible; it requires a separate security fob;
and if and when access actually works (it never has for me), you get to
a green screen about as user-friendly as old OJIN. But the IITP design
team (of which I am a member) is meeting monthly, and IITP is making progress,
albeit at the same rate as the automated criminal judgment which DOC has
been asking of us for years.
My primary focus on IITP is the information it can provide for
sentencing support. Ideally, when we select offenders to compare
for their susceptibility or not to improvement by incarceration or programs
(in or out of custody), we can include the kind of assessments upon which
IITP depends and program participation which it records.
Prognosis
Although government institutions cling with traditional tenacity to glacial pace, sentencing support progress and eventual implementation seems inevitable. The recognition of policy commitment to this tool is growing in the criminal justice system. In addition to spreading interest from other Oregon counties than Multnomah, criminal justice actors from the National Center for State Courts and several foreign jurisdictions (Nigeria, Canada, Scotland, and Australia) have shown substantial interest in our work. I will try to keep you informed and will probably be seeking your assistance in one way or another in the future. I welcome your comments and suggestions.