Comparing integration and inclusion between Canadians and Americans with disabilities: Evidence from national surveys of time use

Clarke Wilson, Ph.D, Queen's University, Centre for Health Services and Policy Research; Queen’s University, School of Geography and Urban and Regional Planning

wcw [at] queensu [dot] ca

Mary Ann McColl, Ph.D. Queen’s University, School of Rehabilitation Therapy;

Queen's University, Centre for Health Services and Policy Research

mccollm [at] queensu [dot] ca

Abstract: As Canada moves toward the passage of a federal statute assuring access and inclusion for disabled persons, inevitable comparisons arise between the statutory environments for people with disabilities in Canada and the USA. In previous research, we have used daily time use as a macro indicator of the degree of integration of people with disabilities into the wider society. If statutory protection of disabled persons is effective, activity participation should be similar between persons with and without disabilities in jurisdictions that are favorable to full participation. This paper provides the analysis of national survey data on time use in the United States and Canada for 2010. It shows that the dissimilarity of time use by persons with and without disabilities is smaller for Canadians than for Americans. This finding shows that disabled Canadians are more integrated into their wider society than disabled Americans. Paid work is one activity where Canadians and Americans with and without disabilities are most dissimilar. Regression analysis of time spent in paid work indicates that, with demographic and economic descriptors held constant, the American residency does not promote an advantage in paid work which is a key indicator of integration. This casts doubt on the effectiveness of statutory protections for persons with disabilities.

Keywords:

Acknowledgements: This work was supported by Queen’s University through the Senate Advisory Research Committee. The authors have no financial interest or benefit from the direct application of this research.

Comparing integration and inclusion between Canadians and Americans with disabilities: Evidence from national surveys of time use

Clarke Wilson, Ph.D, Queen's University, Centre for Health Services and Policy Research; Queen’s University, School of Geography and Urban and Regional Planning

wcw [at] queensu [dot] ca

Mary Ann McColl, Ph.D. Queen’s University, School of Rehabilitation Therapy;

Queen's University, Centre for Health Services and Policy Research

mccollm [at] queensu [dot] ca

1. Introduction

The Government of Canada has recently undertaken extensive public consultations toward the development of Bill C-81: An act to ensure a barrier-free Canada. The bill was designed to “promote equality of opportunity and increase inclusion and participation of Canadians who have disabilities or functional limitations.” (Employment and Social Development Canada, 2016). Although the Minister for Public Services, Procurement and Accessibility, the Honourable Carla Qualtrough, has been clear that the new bill is not simply a Canadian version of the Americans with Disabilities Act (ADA; 1990; 2008), there are inevitable comparisons with the situation in the U.S.A. The ADA makes it illegal to engage in civic or corporate discrimination on the basis of disability in the areas of employment, public services, public accommodations, telecommunications and miscellaneous. It characterizes disabled people as a discrete minority group, and takes a human rights approach. Other Western countries have followed suit with similar disability discrimination protections, particularly the UK and Australia.

In Canada, multiple levels of rights protections are already in place - the Charter of Rights and Freedoms,1982; Canadian Human Rights Act, 1977; Employment Equity Act,1995; and provincial/ territorial human rights acts and labour codes. In addition, Canada has committed to signing the Optional Protocol of the United Nations Convention on the Rights of Persons with Disabilities. The Protocol recognizes the authority of the UN Committee on the Rights of Persons with Disabilities to hear complaints against states parties. Instead of adding further anti-discrimination legislation, Bill C-81 proposes a structural approach to accessibility that dovetails with the existing statutory environment. It responds to the issues and barriers enumerated by Canadians with disabilities in the national consultations that have taken place of the past year.

The idea of federal disability legislation has by no means been universally supported. According to Prince (2010), there are three camps of responses to the idea. One group supports the proposal whole-heartedly, believing that for both real and symbolic reasons, the enactment of federal disability legislation will galvanize the disability community, and provide the impetus to correct some of the slippage that has been perceived in disability policy in recent years (Boyce et al., 2001; McColl & Jongbloed, 2006).

A second group expresses ambivalence toward the idea of federal disability legislation. They recognize the potential benefits, but also the possible pitfalls of an overarching legislative response to the multi-dimensional, multi-sectoral problems experienced by the heterogeneous community of disabled people. They advocate for a highly consultative process to ensure appropriate considerations and representation.

A third group opposes a federal disability act, believing that the effects will at best be negligible, and at worst detrimental. Some believe that the existing legislative framework provides all the safeguards and provisions necessary. Others fear that such an initiative would be nothing but window-dressing, and would distract attention from a programmatic approach to the persistent and pressing problems of the most disadvantaged disabled people.

A persistent challenge for disability policy is the lack of empirical evidence to form a sound evidence base upon which to build. In its absence, ideological tensions have made it difficult for policy-makers to relate to the disability community and to achieve consensus on the needs of people with disabilities (Joiner, 2006; Prince, 2004, 2006). Furthermore, debate is often highly polarized, and inflamed by the rhetoric of rights (Bickenbach, 2006). A number of areas exist where there are strong disagreements about how disabled citizens should be viewed, what they need, and how they can be best served by governments in Canada (McColl & Jongbloed, 2006). Any fractiousness within the community permits the government to do nothing until a clear policy direction emerges with some support and momentum.

Several authors have called for an empirical approach to disability policy, using a macro-level composite index of social inclusion (Prince, 2009; Simplican et al, 2015). Such an index would operate at the population level, reflect the participation of all disabled people, and encompass variations in their needs and circumstances. It would enhance the ability of government and advocates to track and evaluate the impact of policy and program activity.

Our previous research has proposed a time use dissimilarity index as such an indicator (Wilson et al, 2017). The dissimilarity index represents the proportion of total time that is spent differently between two populations; alternately, it is the proportion of time that would have to be reallocated in order to produce identical time use profiles. The index reflects the extent to which time allocated to various activities differs between disabled and non-disabled sectors of the population. It is based on the proposition that similarity of time use reflects similarity in access to resources and opportunities. To the extent that public policy is designed to create equal opportunity and access to goods and services, the outcomes of disability policy can be judged on the extent to which time use is similar between disabled and non-disabled sectors of the population. Greater similarity between disabled and non-disabled populations is representative of greater integration. Given the similarity of many aspects of culture, geography and economic development of the United States and Canada, we argue that differences in time use of disabled people in the two countries can be at least partially attributed to the differences in the policy environment.

We have shown that time use currently differs significantly between disabled and non-disabled adults in Canada for a number of important activities: paid work (disabled average 131 minutes per day vs. non-disabled 210 minutes per day), family responsibilities (19 vs. 30 min.), education (16 vs. 36 min.) and TV/computer time (199 vs. 145 min.). On a more hopeful note, disabled and non-disabled samples have begun to converge in time use in the decades between 1992 and 2010 (Wilson et al, 2017).

In this study, we use the time use dissimilarity index to explore the (dis)similarity in time use between disabled people in Canada and the US and as an indicator of the relative effects of the policy environments in both countries. As Canadians contemplate the form and content of national disability legislation, this study compares the experience of people with disabilities in Canada (under the existing statutory infrastructure) with their counterparts in the United States, 30 years after the institution of the ADA.

The specific objectives are:

  1. to compare time use among disabled and non-disabled populations in Canada and the United States, with respect to 18 activities of daily living;
  2. to assess the mediating impact of gender on time use;
  3. to examine the relative effect of country (Canada vs USA) on time allocation to paid work which is one of the key indicators of social inclusion and opportunity.

2. Method

Design

This study is a retrospective secondary analysis of national survey data. It is a cross-sectional comparison of disabled vs. non-disabled sectors of the population on time use. National time use survey data in both countries in 2010 provides an opportunity to undertake this comparison.

Data

The data for the present study came from two 2010 time use surveys:

(a) Disability

Statistics Canada has employed a number of definitions of disability, the most recent being based on an assessment of impairment type and activity limitations (Grondin, 2016). This procedure was employed in the 2012 Canadian Survey on Disability (CSD) and subsequently in the GSS and other survey programs. Impairments include hearing, seeing, walking or climbing stairs, reaching or grasping or bending, communicating, and dealing with cognitive and psychological conditions. Activity limitations refer to the degree of restriction at home, at work, or at other places. The American ATUS defines disability as difficulty dressing, bathing, hearing, seeing, doing outside errands alone, walking, concentrating or remembering. The prevalence of disability in the Canada based on the GSS public file is estimated at 20%, which is higher than the reported rate from the 2012 CSD of 13.7 percent (Statistics Canada, 2013). Disability prevalence in the ATUS was 9.3 percent.

Grondin describes a number of validity tests that Statistics Canada performed on various disability definitions from 2008 to 2012, including one based on work of the Washington Group on Disability Statistics (2001). She concludes that hearing, seeing, walking and dexterity impairments were relatively well understood by respondents and were conceptually consistent between Canadian and American surveys. Limiting the disability definition to any of these four conditions results in prevalence rates of 15.2% in Canada and 7.8% in the U.S. This definition gives results more consistent with the prevalence rate of the Canadian Survey on Disability (see Table 1).

Table 1. Sample size and population characteristics, 2010, by disability status
Canada United States
Sample counts No disability Disability No disability Disability
Total 12,494 2,896 11,981 1,279
Female 6,911 1,778 6,663 784
Male 5,583 1,118 5,318 495
Population distribution by characteristic (weighted sample)
Estimated population 23,437,289 4,204,006 222,948,562 18,964,918
Sex
Female 49.4 % 57.1 % 51.3 % 54.5 %
Male 50.6% 42.9% 48.7% 45.5%
Age group
15 to 44 54.9% 21.9% 53.9% 15.1%
45 to 64 32.7 41.9 32.8 37.1
>= 65 12.4 36.2 13.3 47.8
Main activity
Labour force 59.7% 33.9% 72.9% 22.6%
Student 10.8 2.4 na na
Other not in LF 29.5 63.8 na na
Some post-secondary education 70.8% 57.3% 53.7% 40.0%
Mean income (CAD 2010) $ 89,054 $ 65,577 $ 67,857 $ 39,857

Currently, no indicator of severity is available on the data file. Statistics Canada expended considerable effort to refine the screening questions for disability, so that type of impairment and severity of disability are available on the 2012 Canadian Survey on Disability, as well as the 2015 GSS time use cycle and other surveys. These indicators will permit future research to examine groups within the disabled community that are being served most effectively by current programs and those which are not. Improved indicators may also permit more ambitious international comparisons of daily activities.

(b) Time use

Time use is measured using a time diary approach; that is, minutes per person per day allocated to a pre-determined set of 18 activities (see Table 2) used in previous research (Wilson et al, 2017). The total number of minutes in a day is 1440 and their distribution among activities is called a time budget.

(c) Demographic variables

In the regression analysis of paid work time which follows, we employ a number of descriptive variables to isolate the effects of disability and country:

Sample demographics

Table 1 gave the estimated Canadian and American populations by disability status and demographic characteristics. The ratios of women to men in Canada and the U.S. are within 1%, but in Canada women constitute almost 3% more of the disabled population than in the U.S. The age structure of disabled population is markedly different. In Canada 64% of disabled persons are under 65 years old and 36% are over 65. In the United States, only 52% are under 65 and 48% are over 65. In other words, the disabled population in the US is older on average than in Canada.

Labour force participation rates are similar for all persons at about 65%, but are notably higher among disabled persons in Canada (34% vs. 23%). Post-secondary educational attainment was also notably higher in Canada for both disabled persons and those without disabilities (71% in Canada vs 54% in the US for non-disabled; 57% in Canada vs 40% in the U.S for disabled).

Analysis

Objective #1 & 2: Comparison of time use between disabled and non-disabled, Canada and US, controlling for the effect of gender.

After weighting the two samples to account for sampling variations, population means for time use were compared between disabled and non-disabled sectors of the population, and between countries. Sample sizes in national surveys are so large that all but very small time differences are statistically significantly in most difference of means tests. This poses reporting problems of avoiding trivial differences and of extracting substantive conclusions from the data.

Dissimilarity indices offer a macro-measure of distributional differences. Stewart (2006) examined several indices applicable to time use data and concluded that the weighted absolute deviation index was both robust and readily interpretable. It reports the proportion of total available time that would have to be reassigned to equate two time budgets. The formula for the dissimilarity index between time budgets a and b is:

T = ∑i abs(ai – bi ) / 2880

where abs( ) is the absolute value of the expression in parentheses and the summation is over all activities, i = 1 … n. A value of, for example, 0.2 indicates that 20% of the total time of the two samples (2880 minutes) would have to be reallocated in order to equalize the two time budgets.

Objective #3: Factors affecting time allocated to work:

Paid work is one of the most important dimensions of integration and is a clear measure of participation; as such, it is also a public policy objective. To explore the effect of disability and country of residence on paid work, we regressed time spent on work against the descriptive variables discussed above. An interaction term was created for country and disability status, namely Canada-Disabled, Canada-Not Disabled, US-Disabled. US-Not Disabled was used as the comparator term.

Published weights inflate the sample sizes and exaggerate statistical significance. Our solution was to normalize the two national samples separately by the national mean weight (Thompson, 2008). The Canada and U.S. sample counts are then correct in a pooled data file and significance tests are based on actual sample sizes.

3. Results

Comparing time use between disabled and non-disabled people in Canada and the US

Table 2 gives the time budgets for Canada and for the United States in 2010 by disability status. Eight activities comprise the bulk of time use for both Canadians and Americans: sleep, screen time (TV or computer), paid work, light housework, personal care, eating, social leisure, and travel. These 8 activities account for 1250 minutes per day (20.8 hours, or 12.2 excluding sleep) among Canadians with disabilities, and 1253.7 minutes (20.9 hours, or 11.6 excluding sleep) among Americans with disabilities.

While the major time-using activities are the same for disabled people in Canada and the U.S., differences of more than 20 minutes occur for a number of activities (see shaded entries):

The Dissimilarity Index between disabled and non- disabled was 10 % for Canada and 15.8 % for the U.S. The daily routines of disabled Canadians resemble those of non-disabled Canadians more closely than the same comparison in the United States.

Table 2. Time budgets, 2010, for Canada and the United States, by disability status (minutes)
Activity Canada United States (Canada–US)
Not disabled Disabled Not disabled Disabled
Screen (TV, computer) 146.5 198.3 170.5 276.8 -78.5
Sleeping 501.5 516.9 516.9 556.0 -38.9
Unreported time 1 1.1 18.1 23 -21.9
Personal care 71.3 97.5 71 109.8 -12.3
Passive leisure 21.7 34.9 22.2 45.6 -10.7
Waiting 1.6 1.6 2.9 2.7 -1.1
Light housework 84.1 102 86.4 98.7 3.3
Adult family care 2.5 4.4 1.2 1 3.4
Shopping & services 30.9 32.6 27.7 28.1 4.5
Civic, voluntary 16.1 19.5 17.3 15 4.5
Child care 27.4 14.5 25 9.6 4.9
Active leisure 38.5 36.6 33.1 31.5 5.1
Heavy housework 27.2 32.4 23.8 25.3 7.1
Education 34.8 12.4 28.6 4.6 7.8
Eating 73.2 80.2 67.5 68.0 12.2
Travel 76.9 60.5 73.6 46.8 13.7
Social leisure 78.4 73.8 49.1 46 27.8
Paid work 206.3 120.9 205 51.6 69.3
Note. Activity ordered by disabled difference, (Canada - U.S.). Differences of 20 minutes or more shown in bold.

Comparing disabled men in Canada and the US

Table 3 gives the time budgets for men by country and disability status.

Disabled Canadian men differ by more than 20 minutes from non-disabled Canadian men on only four activities:

Disabled American men differ from non-disabled by more than 20 minutes on 7 activities:

Directly comparing disabled men in Canada and the US,

Table 3. Men’s Time budgets, 2010, by country and disability status (minutes)
Canada USA Disabled
Activity Not disabled Disabled (D – ND)* Not disabled Disabled (D – ND) (Can – US)
Men
TV, computer 165.7 226.5 60.8 182.8 302.9 120.1 - 76.4
Personal care 64.9 96.7 31.8 63.1 99.2 36.1 - 2.5
Sleeping 495.6 511.8 16.2 510.9 550.5 39.6 - 38.7
Passive leisure 19.8 33.5 13.7 20.6 45.6 25.0 - 12.1
Social 73.9 66.2 - 7.7 48 43.7 - 4.3 22.5
Travel 79.7 62.8 - 16.9 75.4 49.2 - 26.2 13.6
Education 32.8 11.4 - 21.4 28.1 4.9 - 23.2 6.5
Paid work 239.3 136.3 - 103.0 237.2 64.7 - 172.5 71.6
* Activity ordered by Canada disabled/not disabled difference. Only differences of 20 minutes or more shown, in bold.

Using the Dissimilarity Index, Canadian disabled men were 11.2 % dissimilar from men without disabilities, while American disabled men were 15 % dissimilar from American non-disabled men. In direct comparison between disabled men in both countries, Canadian disabled men were 10.6% dissimilar from American disabled men.

Comparing disabled women in Canada and the US

Table 4 gives time budgets for women.

Disabled Canadian women differ by more than 20 minutes from non-disabled Canadian women on only three activities:

Disabled American women differ from non-disabled by more than 20 minutes on 8 activities:

Directly comparing disabled women in Canada and the US,

Table 4. Women’s Time budgets, 2010, by country and disability status (minutes)
Canada USA (Can – US)
Activity Not disabled Disabled (D – ND)* Not disabled Disabled (D – ND) Disabled
TV, computer 127.0 177.2 50.2 158.8 255 96.2 - 77.8
Sleeping 507.5 520.7 13.2 522.5 560.5 38.0 -40.2
Passive leisure 23.7 35.9 12.2 23.7 45.7 22.0 - 9.8
Personal care 77.9 98.1 11.2 78.5 118.6 40.1 - 20.5
Unreported time 0.8 1 0.2 18.8 24.9 6.1 - 23.9
Social 83 79.5 - 3.5 50.1 47.8 - 2.3 31.7
Travel 74.1 58.7 - 15.4 71.9 44.8 - 27.1 13.9
Child care 37.4 20.2 - 17.2 33.7 12.9 - 20.8 7.3
Education 36.8 13.2 - 23.6 29 4.3 - 24.7 8.9
Paid work 172.6 109.3 - 63.3 174.5 40.6 - 123.9 68.7
* Activity ordered by Canada disabled/not disabled difference. Only differences of 20 minutes or more shown, in bold.

Using the Dissimilarity Index, Canadian women were 8.8 % dissimilar from Canadian non-disabled women, while American disabled women were 17.2 % dissimilar from American non-disabled women. Canadian disabled women were 12.1% dissimilar from American disabled women. This is almost 15 percent greater than the dissimilarity between disabled American and Canadian men.

Regression analysis of paid work

Table 5 gives the regression coefficients and standard errors for the pooled regression analyses of paid work time. The adjusted R-square statistic indicates that the equation accounts for 22 percent of variation in work time. The sample contained 19,845 respondents who supplied a complete data set.

The regression model is designed to reflect the impact of country and disability status on time spent in paid work, while controlling for known differences in age, sex, post-secondary education, socio-economic status, and labour force participation (see Table 1). Regression coefficients can be interpreted as the number of minutes per day accounted for by the variable.

Controlling for other variables, the impact of country and disability on work time appear to be as follows:

Table 5. Regression Coefficients for Paid Work Time
Independent variable Coefficient Std. error t statistic
Constant -248.9 16.9 -14.7
Disabled_US -51.0 15.5 -3.3
Disabled_Canada -33.5 7.8 -4.3
Not disabled_Canada -26.9 3.9 -6.9
Weekday 240.1 3.7 65.7
Male 50.3 3.3 15.2
Management/Administrative occupation 27.1 3.8 7.1
Post-secondary education 19.3 3.7 5.2
Age 17.4 0.7 25.2
Number children -10.4 1.7 -6.0
Living as couple 10.1 4.0 2.5
Unemployed -3.1 1.1 -2.9
Age squared -0.2 0.0 -24.8
Note. N = 19,845 R-squared coefficient = 0.22. All coefficients are significantly different from zero (p < .01)

Weekday completion of the survey is obviously influential, accounting for 6 hours of time allocation. Four additional variables account for more than 15 minutes each:

4. Discussion

This study set out to compare time use of disabled people in Canada and the United States with respect to daily activity patterns, in particular, paid work. The analyses of time budget data by country and by gender suggest greater inclusion of disabled persons in Canada than in the United States. Activity patterns of Canadian women with disabilities are more similar to those without disabilities than is the case for men. While these relative rates of integration may reflect differences in the impact of legislation in the two countries, they may also reflect differences in the demographics of the two survey samples. Accordingly, we have conducted a regression analysis of paid work time (a major indicator of economic opportunity and integration) to control for demographic effects.

We have shown that:

Overall, disability appears to have a smaller negative impact on paid work and other activities in Canada than in it does in the U.S. Thus any expectation that the Americans with Disabilities Act has created a more favourable environment for disabled people than the Canadian policy infrastructure is not supported in this empirical analysis, using highly robust national survey data. Time use surveys provide no evidence that disabled persons are more integrated or engaged in the United States than in Canada. Disabled Americans were 16% dissimilar from non-disabled Americans, versus 10% dissimilar for Canadians with and without disabilities. In particular, Americans with disabilities worked 17 minutes less than Canadians with disabilities. The regression model explained 22% of the variance in time allocated to work.

When we think of barriers to work, we often think in terms of four types of barriers:

The legislative environment – labour law, employment equity law, human rights protections -- is obviously a systemic factor affecting work for people with disabilities. Legislation can also contribute to physical accessibility, particularly if it establishes and enforces accessibility standards. It can favour informational accessibility if it ensures standards of information access, and upholds equal opportunities for education and training. Our previous research shows that post-secondary education produces substantial economic gains for workers with disabilities. Higher rates of post-secondary education were associated with greater time spent in paid work (Wilson et al., 2015, 2017). It is more challenging to draw inferences about a direct effect of legislation on attitudes.

Integration has also been studied in relation to immigration. In this research, four key dimensions of integration have been identified: status, rights, engagement, and identity (Klaver & Ode, 2009). These are broadly related to Prince’s (2009) five dimensions of citizenship: discourse, legal, democratic, fiscal-social, and economic.

The terms status and citizenship rights versus, legal and democratic rights, cover somewhat the same ground as they arise from either statutes or constitutions and, within stated parameters, apply universally to named populations (e.g. adults, citizens etc.). Prince however, points to a history in Canada of legal and demographic rights for persons with disabilities being initially ignored but later defended and extended by litigation. We take no exception to such observations, but point out that affirmation of such rights will eventually be reflected in changed behaviour such as greater participation in elections or civic activity. Over time such activity will be reflected in reported time use data.

Prince’s terms fiscal-social and economic rights broadly deal with the Klaver and Ode dimensions of engagement and identity. The term engagement has historically been applied to political and community activity. However it is now being expanded to include workforce attachment, neighbourhood roles, and access to the retail and service markets. As such it addresses much of the conceptual content of social and economic rights. Engagement is clearly germane to the definition and measurement of integration of disabled persons but is to a great extent captured by time use data. Identity relates to shared social, legal, historical, and cultural traditions. Movements such as the Paralympics and demands for physical access to public facilities attest to the determination of disabled persons to participate in social and cultural institutions. One could say that identity of disabled people with society as a whole is a driver in their quest for civic integration.

Time use is not a perfect measure of integration, but in an absolute sense, no empirical or conceptual construct could be. The question is, “are there dimensions of integration that are important to disabled people which are not reflected in time use data?” Most aspects of engagement (e.g. work force attachment, community participation, education and social interaction) and identity (participation in public institutions and culture) are well reflected by time use measurements. To the extent that litigation affirms democratic and legal rights, these will be reflected in increasing similarity of time use of persons with and without disabilities. What is missing is a subjective element. Respondents to national surveys are generally not asked whether they feel included, or whether they value the needs of others in society. It would not be difficult to make a case for a subjective measure of inclusiveness on social survey instruments.

5. Conclusion

In conclusion, this empirical analysis of time use data from Canada and the United States has shown that the policy environment in the U.S., which includes the Americans with Disabilities Act, has not produced superior conditions for integration of disabled people in that country. Disabled Americans spend less time in paid work and social leisure than disabled Canadians, and more time sleeping, using TVs/computers, and in unreported activity. Disabled Americans are 16% dissimilar from non-disabled Americans in their time use, whereas disabled Canadians are only 10% dissimilar from their non-disabled counterparts. With regard particularly to paid work, a key indicator of economic and social integration, Americans with disabilities are significantly more disadvantaged than Canadians with disabilities. This analysis suggests that any bias toward an American style policy environment as regards disability would not necessarily have salutary effects for disabled Canadians.

References