1

Access to Justice Data

Research Data Centres & Real Time Remote Access

Kathy AuCoin

Chief Data Access and Data Development

August 2013

New CANSIM Tables

March 2012 – 45 CANSIM tables As of July 2013 –100+ CANSIM tables Civil Courts data tables Police Resources data tables – municipal level Detailed Criminal Offences and Crime Severity

Index by municipality

2

Additional CANSIM Tables2013/2014

Legal Aid Data Tables

Additional Police Resources Data

Homicide by region

Intimate partner homicides

Hate Crime and Cybercrime data3

Justice Files

Files currently available in RDC & RTRA

o Uniform Crime Reporting Incident Based Survey

o Homicide Survey

Additional files for RDC/RTRA by 2014-2015

o Hate Crime Supplemento Integrated Criminal Courts Survey

4

Uniform Crime Reporting (UCR) Incident Based Survey

Measures the incidence of police reported crime in Canada

Close to 100% coverage, national, provincial, CMA and non CMA

UCR generates micro data for the purposes of examining characteristics of victims, accused and incidents

5

UCR Incident Based Survey

Captures data on more than 200 Criminal Code offences

Violent crime Violations causing death, sexual & physical assaults, robbery,

extortion, violations resulting in the deprivation of freedom

Property Crimes Arson, break & enter, shoplifting, mischief

Other Crimes Impaired driving, prostitution, possession of weapons,

counterfeiting, violations resulting in the deprivation of freedom

Drug Related Crimes Possession, trafficking, importation, production

6

Uniform Crime Reporting Incident Based Crime Survey

UCR Incident-based survey has 3 main files• Incident file• Accused file • Victim file

Each file can be used independently or they can be linked – depending on the research question.

Significant number of records

7

8

How many years of data are available?

5 years of data• RDC -- 2007, 2008, 2009, 2010, 2011• RTRA -- 2008, 2009, 2010, 2011, 2012

Files can be used individually or combined Increases N for small cells NO TREND analysis Population data included to produce rates

9

Potential Research Questions for UCR data

Intimate partner violence Violence against children and youth Youth crime – patterns, characteristics Adult crime – patterns, characteristic Temporal patterns of crime Crimes involving weapons Family verses non-family victimization

10

Potential Research Questions for UCR data (cont’d)

Gender differences – Offending

Gender differences – Victimization

Property crimes, CMA & non-CMA

Robbery crimes, CMA & non-CMA

11

Confidentiality Measures for the UCR RDC/RTRA files

Characteristics of individual incidents, accused or victims will not be disseminated to the public.

CCJS has developed confidentiality vetting guidelines specific to the UCR Survey to prevent the disclosure of sensitive information.

3 main steps taken to maintain confidentiality• Excluded highly sensitive variables from file• Aggregated sensitive response categories• Developed disclosure rules

12

UCR Variables EXCLUDED from RDC/RTRA Files

Variables that identify • Name of victim, accused

• Date of birth

• Fingerprint identifier of accused

• Incident file number

Variables with unknown or poor data quality• Address, postal code, geo-code

• Aboriginal origin of accused/victim

13

UCR Variables INCLUDED on the RDC/ RTRA -- Incident Files

Location school, institution, private dwelling, open area,

streets

Time and Date of Incident Occupancy

indicates whether the victim and accused were living together

14

UCR Variables INCLUDED on the RDC/ RTRA – Incident Files (cont’d)

Geography• National, province/territory, CMA and Non CMA

Most Serious Weapon Present

Clearance Status

UCR Variables INCLUDED on the RDC/ RTRA – Accused & Victim Files

ACCUSED FILE Age Sex Date charges laid

VICTIM FILE Age Sex Relationship to the

accused Injury sustained Weapon causing

injury

15

Incident Clearance Status • E.g., death of a complainant, the suicide or death of

the accused; accused to a mental hospital.

Incident Time • Researcher must collapse the time into 6-hour

intervals.

Sensitive violation codes • Other sexual violations• Terrorism• Homicide

16

Aggregated Response Categories and Violation

Disclosure Risk Scores and Rounding

Produce outputs with confidentiality risk scores that are at or below the acceptable threshold. • RDCs only

If confidentiality risk scores are too high:• Remove potentially sensitive variable(s) from the

output to reduce the overall score; and/or • Aggregate potentially sensitive variable(s) from the

output to reduce the overall score.

Apply a controlled rounding program applied to the output• Automatically applied to RTRA outputs

17

RDC or RTRA

RDC Modeling, regression

analysis Descriptive analysis of

violations with small counts

Confidentiality measures: Disclosure rules, based on scores and a cap or rounding

RTRA Descriptive statistics of

offences that are common, large geography (i.e., national and provincial , CMA and non-CMA)

Confidentiality measures: Rounding program (base 5)

18

19

Rounding

IOffence Actual Rounded

Attempted Murder 13 15Aggravated Sexual Assault 2 5Sexual Assault with a Weapon 11 10Sexual Assault 587 585Sexual Interference 54 55Aggravated Assault - Level 3 24 25Assault Against Peace-Public Officer 2 0Aggravated Assault Against Peace Officer 1 0Abduction Under 14, by Parent/Guardian 5 5Robbery 1457 1455Criminal Harassment 201 200Indecent/Harassing Telephone Calls 63 65Uttering Threat to Person 449 450Arson – Disregard for Human Life 5 5

Other Violations against the person 12 15

UCR Incident Based SurveyEnd-User Documentation

Developed from police scoring guides Extremely detailed (403 pages)

User-friendly Detailed confidentiality guidelines & vetting rules Scoring guide Rounding option Aggregations SAS coding examples Variable definitions & descriptions Many typos!!!!

20

Homicide Survey Files - RDC & RTRA

New pilot starting October 2013 Second attempt – 2009 pilot cancelled due to

insufficient resources Results from the first pilot were positive

• ease of use of data file, documentation and wealth of data/information

However data disclosure rules difficult to implement

21

Homicide Survey Pilot, 2013

Development of user-friendly disclosure rules to ensure confidentiality of data• Because of small numbers Homicide Survey presents

a unique challenge with regard to confidentiality

Approach for disclosure rules similar to UCR

22

23

Homicide Survey Disclosure Rules

Threshold table value: 7• Table scores>7 are not released

Maximum number of dimensions: 3

Variables deemed NOT sensitive (score 0):• 10-year data• National data• Sex of victim, accused

What kinds of data are available from the Homicide Survey?

Data available from 1961-2011 3 Files – Incident, Victim, Accused Detailed reporting on:

• Marital status

• Detailed location

• Drugs related homicide

• Precipitating crime

• Alcohol/ drugs consumed (accused or victim)

• Apparent motive

• Accused-victim relationship

• Cause of death

24

25

Potential Research Questions for Homicide data

Family related homicides Intimate partner homicides Temporal analysis Firearm related homicides CMA and non-CMA analysis Precipitating factors

Pilot - Hate Crime File

Additional data file which is linked to the UCR Incident-based survey

File available several months after initial data collection

Police services determine if the criminal act was motivated by hate of:• Race, age, sex, religion• Sexual orientation, physical disability • Language, color, nationality, ethnicity

26

Pilot - Integrated Criminal Courts Survey

Administrative records of cases before the courts

Provincial and territorial data Variables include age, sex of individual before

the courts Count of decisions by charge Types of decision, (guilty, probation) Sentencing patterns

27

Research QuestionsIntegrated Criminal Courts Survey

Case processing times

Elapsed time

Efficiency of the justice system

28

Questions?

29