Justice Services, Economic Burden and Social Costs in Canada

Annex A. Methodology

This methodological annex provides a detailed explanation of the analytical approaches and definitions underpinning the results presented in the main report. It documents how core indicators were constructed from the Canadian Legal Problems Survey (CLPS), the General Social Survey (GSS) and other sources, and how causal inference techniques were applied to isolate the effects of demographic factors and justice pathway experiences on outcomes.

Incidence

Civil issues

A respondent to the CLPS is considered to experience a justice problem if they reported a serious problem within one of the following categories: large purchase or service; employer or job; injury or serious health issue at work or a traffic accident; house, mortgage or rent; debt or money; getting government assistance; getting disability assistance; immigration to Canada; being stopped, accused or arrested; breakdown of family; child custody; executing a will or taking care of a disabled person; poor or incorrect medical treatment; and harassment (variables PRIP10A-S except D and K).

As explained in the text, the CLPS incidence results appear as outliers when compared to both older and more recent legal needs surveys, including the most recent National Justice Survey (2025 NJS). Owing to the size of its sample, however, the CLPS provides a unique source for detailed analyses of justice pathways and impacts of justice problems, particularly when the focus is placed on subgroups of population. For this reason, the report derives its levels of incidence of civil justice problems from the 2025 NJS but keeps the CLPS as its key source for other parts of the analysis.

Figure 5.1 of Chapter 5 displays the incidence of the 19 problem types against their reported severity. Incidence is the number of people reporting each problem as their most serious in the CLPS (variable SERPORB), extrapolated to the total population in accordance with the survey’s expansion factors. The severity of problems is addressed through six distinct questions in the CLPS:

• Financial Impact: respondent experienced costs exceeding USD 5 000 due to their most serious problem (variable CSTP20P)

• Social Impact A: respondent lost a job due to their most serious problem (variable SOCP10A)

• Social Impact B: respondent lost housing due to their most serious problem (variable SOCP10B)

• Health Impact A: respondent experienced extreme stress due to their most serious problem (variable HLTP20A)

• Health Impact B: respondent experienced physical health problems due to their most serious problem (variable HLTP20B)

• Health Impact C: respondent experienced mental health problems due to their most serious problem (variable HLTP20C)

The study calculates the percentage of people reporting each type of impact among those reporting the problem and averages these percentages first within social impacts and health impacts and then across financial, social and health impacts.

The joint distribution of incidence and severity helps to identify a subset of problems that can be left out of the next phases of analysis because they are both relatively uncommon and relatively unimpactful.

Criminal issues

A respondent to the GSS is considered as a victim to a crime if they reported one of the following victimisation events: assault; sexual assault; robbery; break and enter; harassment (threats, stalking, cyberbullying); theft (motor vehicle, part of a motor vehicle theft, personal property, household property); and vandalism. These problems were aggregated into two categories: ‘property crime’ for theft and vandalism and ‘violent crime’ for all other types.

Data on incidence of a third category of criminal acts labelled as ‘domestic violence’ was collated from the following two sources:

• the 2019 GSS separately collects data on spousal violence experienced in the past twelve months by respondents aged over 15 (Conroy, 2021[1])

• the 2018 SSPPS collects data on experiences of childhood victimisation reported by respondents aged over 15 (Heidinger, 2022[2]).

The SSPPS data provides information on the level of prevalence of childhood victimisation – the number of people in all age cohorts who have suffered from acts of physical (including sexual) and psychological abuse during their childhood. To derive estimates of the incidence of childhood victimisation (the number of children under the age of 15 experiencing such acts every year) from this data, the survey:

• focuses on the victimisation rate V_15-24 reported by the youngest adult cohort (15-24 years) in order to capture recent rather than older trends;

• assumes that low-frequency acts of abuse (less than 5 acts over the respondent’s childhood), which represent about two thirds of the total, happened during one of the fifteen years;

• assumes that high-frequency acts of abuse (more than 5 acts over the respondent’s childhood) , which represent about one third of the total, happened during three of the fifteen years;

• calculates the annual incidence rate of childhood victimisation as one fifteenth of the childhood victimisation rate for two thirds of the victims and one fifth for the remaining third:

I_0-15 = (1/3*1/5 + 2/3*1/15)*V_15-24 = 1/9*V_15-24

Harmonisation of incidence rates

The incidence of civil justice issues reported by the CLPS and the NJS focuses on the population aged 18 or more, while the GSS and SSPPS report criminal victimisation for the population aged 15 or more. To address this difference, the study assumed that the incidence of civil justice issues for the population aged between 15 and 18 would be similar to that of the first age cohort singled out in the NJS results, namely the 18-34 group.

The second key difference between sources is that the CLPS and the NJS consider the occurrence of problems in the three years prior to the survey while the GSS and the SSPPS focus on the last year. To address this difference, the study seeks to derive 12-month incidence rates from the CLPS and the NJS. Unfortunately, neither survey collects information on the length and repetition of problems. The literature suggests, however, that numerous problems tend to repeat themselves, extend in time or trigger new problems within the same area (Focus Consultants, 2008[3]; Currie, 2009[4])). The study conservatively assumes that 25% of problems reported in the NJS were repeated every year, while 75% were related to single events. The annual incidence rate of problems I_1Y can then be derived from the reported incidence rates over three years I_3Y as follows:

I_1Y = (0.25*1 + 0.75*1/3)*I_3Y = 0.5*I_3Y

As a result, all incidence rates presented in Table 2.1 of Chapter 2 are annual rates of incidence within the population aged 15 or more, except for the rate of domestic violence which is expressed as % of the total population.

Incidence within subgroups of population

Excess probability in this analysis is calculated through exact matching, a causal inference method that isolates the effect of covariates (here, the demographic characteristics of the respondent including gender, visible minority status, Indigenous identity, education level and income level) on legal problems. The approach works by finding individuals who are identical across gender, visible minority status, Indigenous identity, rural/urban residence, education level, and income level, except for the single demographic factor being examined - essentially creating perfectly comparable groups that differ only in the trait of interest. For example, visible minority, high education, high income, urban, women matched against non-visible minority, high education, high income, urban, women.

Once these matched groups are identified, the share of respondents within each group who experienced a given outcome (incidence, capability, legal help, etc.) is calculated.

Once the rates are calculated, differences are reported as relative risk ratios (RRR) rather than raw percentage-point differences. A relative risk ratio expresses how much more (or less) likely a matched group is to experience a given outcome compared to its counterpart. An RRR of 1.20 indicates that the group is 20% more likely to experience the outcome; an RRR of 0.80 indicates 20% less likely.

For example, if a group of matched women experience a particular legal problem at a rate of 30%, while a group of (otherwise identical) matched men experience it at 25%, then women’s relative risk is 30%/25% = 1.20, meaning women are 20% more likely to face that problem.

Justice pathways

Legal capability

Legal capability measures an individual's ability and knowledge in handling legal problems.

CLPS: A respondent is defined as having legal capability if they understood the legal implications or had knowledge of where to get information (variables SCPP10, SCPP20) when they first became aware of the problem.

GSS: Respondents are deemed legally capable if they report their problem to the police or do not report it because they do not want to deal with the police, think that reporting will not be helpful, or decided to handle the problem privately. Three reasons for not reporting were assimilated to the lack of capability: that the respondent did not know the issue could be reported, or that they thought no one was harmed, or that no harm was intended (variable cirmainc).

The sources on spousal violence (GSS) and childhood victimisation (SSPPS) do not provide information on reporting to the police or other aspects of the respondent’s legal capability. In these two areas, capability was therefore marked as not available (NA).

For each problem type, the analysis compares capability rates between matched demographic groups who experienced that specific problem. For instance, among individuals who faced employment disputes, the analysis reveals to what extent matched women are more or less likely relative to matched men to have legal capability (expressed as a relative risk ratio (RRR)). This reveals how being a man or a women independently affects one's ability to navigate legal challenges.

When analysing GSS data, nearly all respondents were classified as legally capable. This near-universal classification created insufficient variation in the outcome variable, preventing the exact matching algorithm from forming comparable demographic groups. As a result, the analysis could not produce valid matched pairs or estimate relative risk ratios for legal capability in any area of crime.

Professional assistance

Professional assistance capture whether individuals sought help from professional assistance providers.

CLPS: A respondent is defined as having received justice services if they had contacted a lawyer, court or tribunal, community organisation, government department, used a labour union to address their problem or preferred to solve the problem without a lawyer. (variables ASTP10D – ASTP10H, ASTP40D)

GSS: A respondent is defined as having received justice services if they had spoken with a lawyer, had received advice from the police or felt that the crime was minor and not worth reporting (variables TTA_150, APT_22, cirmainc).

SSPPS: The SSPPS survey does not report information on respondents’ use of justice services following childhood victimisation.

The analysis reveals how demographic characteristics affect help-seeking behaviour within each problem type. For example, if matched groups of rural and urban individuals who experienced debt problems sought legal help at rates of 22% and 15% respectively, then the relative risk ratio is 22/15 = 1.47, meaning rural individuals are 47% more likely to seek justice services than their urban counterparts.

Access to dispute resolution

Access to dispute resolution measures whether individuals used formal processes like court proceedings, tribunals, or mediation.

CLPS: a respondent is defined as having accessed dispute resolution if they had attended a court or tribunal (variable LGAP60P) or tried mediation or conciliation (variable LGAP40P).

GSS: a respondent is defined as having accessed dispute resolution if they had reported the incident to the police (variable PFO_105).

SSPPS: The SSPPS survey does not report information on respondents’ use of dispute resolution following childhood victimisation.

The analysis shows differences in the use of dispute resolution between matched demographic groups within specific problem contexts. For example, if matched groups of rural and urban residents who report facing unfair treatment used dispute resolution at rates of 18% and 12% respectively, the relative risk ratio is 18/12 = 1.50, meaning rural residents are 50% more likely to use a dispute resolution process than their urban counterparts.

Resolution of problem

Resolution of problem indicates whether individuals' justice problems were settled.

CLPS: a respondent’s problem was deemed resolved if it came to a satisfactory or unsatisfactory resolution (variable STAP10).

GSS: a respondent’s problem was deemed resolved if the police had arrested, charged or warned the perpetrator (variables AP405060, APT_30).

SSPPS: The SSPPS survey does not report information on resolution of childhood victimisation cases.

The analysis shows how demographic characteristics affect resolution success within each problem type. For example, if matched groups of visible minority and non-minority individuals who experienced harassment problems achieved resolution at rates of 28% and 35% respectively, the relative risk ratio is 28/35 = 0.80, meaning visible minority individuals are 20% less likely to achieve resolution than their non-minority counterparts.

Impacts and costs

Personal costs of dispute resolution for civil issues

The study derived the costs of dispute resolution for civil justice problem areas from the CLPS (variable CSTP20P). The costs were distributed between lawyer fees, healthcare costs, mental health costs, loss of income, childcare costs and other costs depending on which of these cost types and how many different cost types of each respondent had reported. Childcare costs were used as a proxy for the value of lost household services and aggregated with loss of income to evaluate the total impact of problem resolution on productivity. Costs for employers were derived from lost productivity by applying the marginal rate of return on labour for the Canadian economy. Loss of property was excluded in order to avoid double counting with lost property as a consequence of theft and vandalism. Lawyer fees and other costs (including court fees, fines, transportation and communications expenses) were considered as transfers to other agents and excluded. The average cost per reported problem were applied to the incidence data estimated in Section 2.

Justice costs

Justice system costs were derived from (Hoddenbagh, Zhang and McDonald, 2014[5]) and adjusted for inflation separately for criminal justice expenditures (up to 2019) and civil justice expenditures (calculated at total expenditure minus criminal court expenditures, up to 2021). Criminal justice expenditures were integrated in the costs of crime calculations (see below).

Civil court expenditures were further adjusted and the increase in the length of cases (average elapsed time from case initiation to first disposition) since 2009 and divided into costs for general civil issues and family cases based on the number of initiated cases (data from the Civil Courts Survey).

The total court costs for these two categories of civil problems were then proportionally broken down by types of legal issues. The rates for the different issues were derived from which were based on court workload indicators (Statistics Canada) and the CLPS. The rates from the CLPS were calculated by the number of individuals who contacted a lawyer or court (variables ASTP10D/ASTP10E) to address their most serious issue.

Legal aid expenditures were derived from (Justice Canada, 2024[6]) and distributed between areas of justice problems according to their overall incidence.

Costs of non-compliance in civil matters

The study separately evaluated lost productivity, intangible costs and costs for public welfare systems related to incidents involving a breach of the law with injuries at work and medical treatment issues. For that, the study collected data on the number of total cases, number of fatalities, wage loss and total healthcare costs and estimated the share of expenditures attributable to non-compliance based on data from relevant sources (AWCBC Detailed Key Statistical Measures Report 2021 for injuries at work; (Canadian Institute for Health Information, 2023[7]; Baker et al., 2004[8]; Tessier et al., 2019[9])). Intangible costs due to fatalities were calculated on the basis of the median age of victims and the value of a statistical life per year (Treasury Board of Canada, 2019[10]). Public costs encompass compensations paid to injured workers through the Canadian Workers’ Compensation System (AWCBC, Detailed Key Statistical Measures Report, 2021) and public health expenditures related to adverse health events (Tessier et al., 2019[9]), both adjusted for the share accidents attributable to non-compliance.

Aggregating these estimates with those of the previous section created a risk of redundancy between the estimates of lost productivity due to work injuries that involved non-compliance and the productivity loss due to injury at work reported in the CLPS. To avoid double counting, the study excluded the latter in the total costs presented in Table 4.4 of Chapter 4.

Costs of crime

For criminal issues, the study relied on the detailed cost estimates conducted by Zhang et al. (2012[11]) and (Hoddenbagh, Zhang and McDonald, 2014[5]) on 2009 data, which it updated to 2019 on the basis on the difference in the number of cases, as well as increases in labour productivity (for productivity loss estimates), in prices (for all expenditures) and in the value of a statistical life (for intangible costs of fatalities) (Viscusi and Aldy, 2003[12]). As the quoted studies focus on violent crime and spousal violence, the study complemented pre-existing evidence with new estimates of the value of lost property due to property crime (theft and vandalism, based on 2019 GSS data), as well as victim costs in terms of lost productivity (associated with employer losses) and intangible costs (pain and suffering) related to child victimisation. Public costs encompass the costs of support services and crisis lines for crime victims, updating Hoddenbagh, Zhang and McDonald (2014[5])’s estimates.

Child victimisation costs were based on the health impacts reported by the 2019 Canadian Incidence Study of Reported Child Abuse and Neglect (Fallon et al., 2019[13]). Victims of serious psychological harm were assumed to develop anxiety and depression syndromes at rates and for durations in line with the findings of the specialised literature (reviewed in (OECD, 2025[14])). Disability coefficients provided by the WHO Global Burden of Disease (Murray and Lopez, 1996[15]; WHO, 2020[16]) estimates were applied to each condition, yielding estimates of lost disability-adjusted life years, which were valued according to the average wage in Canada for annual productivity loss and to GDP per capita for annual intangible cost related to pain and suffering (Robinson et al., 2017[17]; OECD, 2025[14]). The total loss of productivity and intangible costs were calculated on the basis of a discounted sum of annual losses over the duration of the conditions (discount rate: 5%).

Effects of justice pathways

Table 4.5 of Chapter 4 is derived entirely using values from the CLPS.

Impact measures whether individuals experienced negative consequences from their legal problems across six dimensions: job loss, lost housing, extreme stress, physical health problems, mental health problems, and costs exceeding USD 5 000 (variables SOCP10A, SOCP10B, HLTP10A, HLTP10B, HLTP10C, CSTP20P).

The analysis reveals how access to justice factors affect problem impacts within matched demographic groups. For each combination of an access to justice measure (legal capability, professional assistance, etc.) and impact type, the analysis compares outcome rates between matched groups who differ only in whether they received that form of justice intervention. For instance, among individuals matched on all demographic factors and problem type, if those who received professional assistance experienced mental health problems at a rate of 45% compared to 52% for those without assistance, then receiving professional assistance is associated with a 7-percentage point reduction (absolute difference) in mental health impacts, representing a 13% relative decrease from the baseline rate.

References

[8] Baker, G. et al. (2004), “The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada”, Canadian Medical Association Journal, Vol. 170/11, pp. 1678-1686, https://doi.org/10.1503/cmaj.1040498.

[7] Canadian Institute for Health Information (2023), Hospital stays in Canada, 2021–2022, Canadian Institute for Health Information, https://www.cihi.ca/en/hospital-stays-in-canada-2021-2022.

[1] Conroy, S. (2021), “Spousal violence in Canada, 2019”, Juristat, No. 41, https://www150.statcan.gc.ca/n1/pub/85-002-x/2021001/article/00016-eng.htm.

[4] Currie, A. (2009), “Multiple Problems”, in The Legal Problems of Everyday Life: The Nature, Extent and Consequences of Justiciable Problems Experienced by Canadians, Department of Justice Canada, https://www.justice.gc.ca/eng/rp-pr/csj-sjc/jsp-sjp/rr07_la1-rr07_aj1/rr07_la1.pdf.

[13] Fallon, B. et al. (2019), Denouncing the Continued Overrepresentation of First Nations Children in Canadian Child Welfare: Findings from the First Nations/Canadian Incidence Study of Reported Child Abuse and Neglect, First Nations Canadian Incidence Study, https://cwrp.ca/sites/default/files/publications/FNCIS-2019%20-%20Denouncing%20the%20Continued%20Overrepresentation%20of%20First%20Nations%20Children%20in%20Canadian%20Child%20Welfare%20-%20Final_1%20%282%29.pdf.

[3] Focus Consultants (2008), Dispute Resolution Longitudinal Study Phase 3: Final report, Focus Consultants, https://www2.gov.bc.ca/assets/gov/law-crime-and-justice/about-bc-justice-system/justice-services-branch/fjsd/longitudinal-final-report.pdf.

[2] Heidinger, L. (2022), “Profile of Canadians who experienced victimization during childhood, 2018”, Juristat, Vol. 42, https://publications.gc.ca/collections/collection_2024/statcan/85-002-x/CS85-002-42-16-eng.pdf.

[5] Hoddenbagh, J., T. Zhang and S. McDonald (2014), An Estimation of the Economic Impact of Violent Victimization in Canada, 2009, Department of Justice Canada, https://justice.gc.ca/eng/rp-pr/cj-jp/victim/rr14_01/rr14_01.pdf.

[6] Justice Canada (2024), Legal Aid in Canada 2022-2023, Department of Justice Canada, https://www.justice.gc.ca/eng/rp-pr/jr/aid-aide/2023/docs/2024-legal-aid-in-canada-2022-23-eng.pdf.

[15] Murray, C. and A. Lopez (eds.) (1996), The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases Injuries and Risk Factors in 1990 and Projected to 2020, Harvard School of Public Health/World Health Organization/World Bank, https://iris.who.int/server/api/core/bitstreams/c5373052-8bac-47bf-a599-45c14cbf4745/content.

[14] OECD (2025), “Access to justice and the economic implications of crime and violence in Thailand”, OECD Working Papers on Public Governance, No. 86, OECD Publishing, Paris, https://doi.org/10.1787/5cc2bb9c-en.

[17] Robinson, L. et al. (2017), “Understanding and improving the one and three times GDP per capita cost-effectiveness thresholds”, Health Policy and Planning, Vol. 32/1, pp. 141-145, https://doi.org/10.1093/heapol/czw096.

[9] Tessier, L. et al. (2019), “The impact of hospital harm on length of stay, costs of care and length of person-centred episodes of care: a retrospective cohort study”, Canadian Medical Association Journal, Vol. 191/32, https://doi.org/10.1503/cmaj.181621.

[10] Treasury Board of Canada (2019), Canada’s Cost-Benefit Analysis Guide for Regulatory Proposals, Treasury Board of Canada, https://publications.gc.ca/collections/collection_2022/sct-tbs/BT58-5-2022-eng.pdf.

[12] Viscusi, W. and J. Aldy (2003), “The Value of a Statistical Life: A Critical Review of Market Estimates Throughout the World”, Journal of Risk and Uncertainty, Vol. 27/1, pp. 5-76, https://doi.org/10.1023/a:1025598106257.

[16] WHO (2020), WHO methods and data sources for global burden of disease estimates 2000-2019, World Health Organization, https://www.who.int/docs/default-source/gho-documents/global-health-estimates/ghe2019_daly-methods.pdf.

[11] Zhang, T. et al. (2012), An Estimation of the Economic Impact of Spousal Violence in Canada 2009, Department of Justice Canada, https://www.justice.gc.ca/eng/rp-pr/cj-jp/fv-vf/rr12_7/rr12_7.pdf.