Assisted-Transport Caregiving and Its Impact Towards Carer-Employees

ABSTRACT

Assisted-transport is the most common informal caregiving task and will be in greater demand due to an aging society. One population group that predominantly covers the demands of informal eldercare while working full time in the paid labor force are carer-employees. The developing carer-employee literature addresses: the health risks for carer-employees; employers of carer-employees, and policy/program interventions. Little research focuses on assisted-transport, which impacts health. This study begins to fill the gap by addressing the following objectives: (1) develop a socioeconomic profile of carer-employees performing assisted-transport tasks; (2) identify any gender differences based on the profile, particularly employment and caregiving traits; (3) examine behavioral factors that increase the likelihood of conducting assisted-transport caregiving, and; (4) determine whether carer-employees are more likely to be overwhelmed from assisted-transport caregiving. Descriptive statistics and logistic regression were used to analyze Statistics Canada’s General Social Survey Cycle 26: Caregiving dataset (2012). Compared to general carer-employees, assisted-transport carer-employees have higher education, household income, and caregiving hours per week and feel more tired and overwhelmed from caregiving. Gender gaps exist based on socioeconomic and caregiving characteristics. Logit results show that female carer-employees are more likely to perform assisted-transport caregiving and feel overwhelmed. Carer-employees conducting assisted-transport caregiving are more likely to be overwhelmed than those who do not.

KEYWORDS:

• Carer-employees
• quantitative analysis
• assisted-transport caregiving

Introduction

Currently, older adults (age 65+) represent 16% of Canada’s population (Bohnert, Chagnon, & Dion, 2015; Statistics Canada, 2017). By 2031, they are projected to nearly double proportionally due to the influx of baby-boomers transitioning to the older adult group (Bohnert et al., 2015; Newbold et al., 2017; Ireson, Sethi, & Williams, 2016). As people age, they are more vulnerable to develop complex chronic conditions and are more likely to utilize health-care resources (Wister et al., 2015). In Canada, 15% of the population are older adults, yet they consume about 45% of public health expenditures (Canadian Medical Association, 2015). With that, the aging population poses pressing challenges for the health-care system and the Canadian economy (Canadian Medical Association, 2016). For instance, while Canadian healthcare is advanced, with an emphasis on hospital-based acute care services, it is not designed to meet the aging population’s need for preventative medicine and prolonged chronic care (Canadian Medical Association, 2015; Silversides, 2014; Wister et al., 2015). Most of this care takes place in the community by informal caregivers, who are the backbone of the health-care system.

Most informal caregivers (75%) in Canada are carer-employees, who are defined as individuals providing unpaid care to someone (often a close relative) while working in the paid labor force (Duxbury et al., 2012; Health Canada, 2002). Often termed ‘employed carers’ in different national contexts, such as the UK, carer-employees do not refer to home health-care professionals or care-related professions, such as nurses and physiotherapists. There are 6.1 million carer-employees in Canada, most of whom provide 1 to 30 hours of weekly informal care, are predominantly female and part of the baby-boom generation (Duxbury et al., 2012; Research on Aging Policies and Practices (RAPP), 2014). Based on these characteristics, past studies have highlighted that carer-employees are becoming more vulnerable to ill health due to caregiver burden and their struggle to maintain a healthy work–life balance (Earle et al., 2011; Employer Panel for Caregivers, 2015; Williams, Wang, & Kitchen, 2014). Pearlin, Mullan, Semple, and Skaff (1990) presents a conceptual framework of caregiver stress and its processes based on four domains: 1) background and context of stress based on key characteristics of the caregiver (i.e., socioeconomic capital); 2) stressors subdivided into primary (i.e., cognitive status), secondary (work-caregiving conflict), and secondary intrapsychic strain (i.e., self-esteem); 3) the mediators of stress from the second listed domain, and; 4) the outcomes or manifestations of stress (i.e., depression, anxiety) (Pearlin et al., 1990). These domains illustrate the composite effects of caregiver burden and work-life imbalance, which often result in negative physical and mental health outcomes (Williams, Wang, & Kitchen, 2016). In the near future, younger carer-employees will experience exponential caregiving burden as baby-boomers transition to the older adult age cohort (Legare, Decarie, & Belanger, 2014).

While there is growing research interest on carer-employees, it is a relatively understudied area and has garnered little public awareness. Much of the literature focuses on caregiver burden and the negative economic impacts that carer-employees unintentionally experience. Some studies have identified socioeconomic characteristics and family relationship to be correlated with experiences of caregiver burden (Montgomery, Gonyea, & Hooyman, 1985; Robinson, 1983); while other studies have highlighted the severity level of the care-receiver’s behavioral disturbance to increase caregiver burden (Cantor, 1983; George et al., 1986; Lund, Pett, & Caserta, 1987). These findings conclude that caregiver burden adversely impacts caregivers’ health. In Canada, 44% of carer-employees report absenteeism, averaging approximately nine days of lost work annually (RAPP, 2014). Further, 15% reduced their paid work hours, thus cutting approximately 10 hr per week (RAPP, 2014). When considered in totality, there are 9.7 million days of absenteeism, 256 million fewer hours of paid work, and a loss of 560,000 carer-employees to the workforce, representing those who left the paid labor force completely (RAPP, 2014). In totality, this is equivalent to 1.2 million full-time paid workers in Canada, making up an approximate loss of $5.5 billion CAD annually due to counter-productivity and absenteeism (RAPP, 2014; Benefits Canada, 2015; Duxbury & Higgins, 2012).

Some employers can mitigate counter-productivity and absenteeism by implementing caregiver-friendly workplace programs, such as compressed work weeks or prolonged leaves of absence (Ireson, Sethi, & Williams, 2016). However, many employers do not implement caregiver-friendly workplace programs as they either lack the flexibility to do so or are unable to afford to implement such programs or simply believe their workplace already provides enough support (Lero, Spinks, Fast, Hibrecht, & Tremblay, 2012).

In addition to the economic consequences, the current literature has focused on mental and physical health consequences to carer-employees. Several studies have focused on how and why moral support is the most demanding task for them, and especially so for females (Duxbury et al., 2012; Family Caregiver Alliance, 2009; Park et al., 2013). Other studies have and continue to work on intervention programs, particularly carer-friendly workplace programs, to improve work–life balance (Ireson et al., 2016; Ramesh, Ireson, & Williams, 2017). These demonstrate that research on carer-employees continues to grow over time; however, the literature still has many gaps that need to be addressed. One area that has received little attention is assisted-transport.

Assisted-transport involves transporting ill or disabled care-recipients to conduct errands, shop, attend medical appointments or participate in social activities. Informal caregivers are the main provider of assisted-transport needs for care-recipient populations, as it is the most frequent type of caregiving task (National Alliance for Caregiving and AARP, 2004; Statistics Canada, 2015). In the United States, 82% of all informal caregivers report providing assisted-transport for their care-recipients (National Alliance for Caregiving and AARP, 2004). Similarly, 73% of Canadian carer-employees provided assisted-transport as the most frequent type of care to their care-recipient in the past 12 months (Statistics Canada, 2015). Additionally, assisted-transport caregiving is the second most demanding task after moral support (Duxbury et al., 2012). Proportionally, an equal number of male and female carer-employees conduct this task. However, twice as many females as males find this task very demanding (Duxbury et al., 2012). One plausible explanation is women’s larger responsibility for personal caregiving and domestic management (housework, cooking, laundry, etc.), thus reducing access to employment opportunities and increasing mental stress when traveling to and from their workplace (Hanson & Pratt, 1995; Kwan, 2000). Many women choose part-time jobs in female-dominated occupations to have the option of flexibility in caregiving. However, these jobs often have less autonomy with respect to employment hours and thus more travel constraints (Hanson & Pratt, 1995; Kwan, 2000; Odih, 1999). Overall, there is a lack of solid evidence of what specifically causes the gender gap with respect to assisted-transport.

Another gap in the literature is the lack of policy recommendations to improve work-transport-life balance. With the aging population, there will be a need to implement transport policies tailored towards the older adult and carer-employee populations. To accomplish this, more research is needed to better understand this informal caregiving task and how it impacts caregiver burden and overall health.

The objectives of this study are to: (1) identify socioeconomic and health differences amongst (i) non carer-employees, (ii) carer-employees performing assisted-transport tasks and (iii) carer-employees that do not perform assisted-transport tasks; (2) for carer-employees performing assisted-transport tasks, determine any gender differences based on socioeconomic and caregiving characteristics; (3) examine socioeconomic and sociodemographic characteristics that increase the likelihood of conducting assisted-transport, and; (4) determine whether carer-employees performing assisted-transport are more likely to be overwhelmed than carer-employees not conducting assisted-transport. Results, via descriptive statistics and multivariate logit regression analyses, intend to be informative for both the carer-employee population and for decision-makers in transport policy and planning.

Data

Statistics Canada’s General Social Survey (GSS Cycle 26: 2012) was used for this study, as it is the only dataset providing caregiving information at the national level (n = 23,093). We were required to fill out a clearance form, which granted us access by Statistics Canada to acquire the dataset that provides individual information on respondents not publicly available. Ethics clearance to perform the data analysis was exempt. The GSS Cycle 26 collected data over a 9-month period (March to December 2012) from Canadians aged 15 years and older by telephone interview. A “Rejective Sampling” technique was used to increase the sample size by reaching out to respondents living in isolated or small communities of unknown population threshold size. Respondents that are classified as caregivers or care-receiver are then asked to participate in a long interview, whereas those who are not are sub-sampled into one of two groups: one for long and the other for short interview. The response rate was nearly 66%. The objective of the GSS 26 is to obtain a brief overview of the caregiver and care-recipient’s lives. More specifically, the intent is to provide information on current or emerging social policy issues by observing changes in the wellbeing of Canadians over time. This dataset provides enough socioeconomic information to compare amongst the three population groups mentioned above. More specifically, it provides a specialized carer-employee profile for those performing assisted-transport tasks for the care-receiver, offering a better understanding of (1) the level of caregiving demand related to transportation (i.e., frequency and proximity to care receiver), (2) whether gender plays a role, and (3) factors that may increase the likelihood of conducting assisted-transport.

For the first objective, 19 variables were chosen from the survey for analyses and classified into two groups: (1) socioeconomic and demographic, and; (2) caregiving-related, including health impacts as a result of caregiving tasks. Socioeconomic and demographic variables include gender, household income, ethnicity, work type, and geographic setting (urban or rural). Based on the data, most of the respondents live in an urban setting; thereby, this research study is characterized as having primarily an urban perspective. Wealth and race were not incorporated in the analyses, due to the fact that they were not available in the dataset. Examples of caregiving-related variables include: ‘relationship to the care recipient’, ‘weekly hours of caregiving’, and ‘conducting assisted-transport.’ Variables describing the health impact from caregiving tasks include: ‘if they feel tired from caregiving’, and ‘feel overwhelmed.’ Table 1 provides a detailed description of each variable.

Table 1. List of variables used from GSS cycle 26.

Variable name	Definition	Values
AGEGR5	Age of Respondent	18+
SEX	Gender of Respondent	Male, Female
ETHNIC25	Ethnic Background of Respondent	26 values recoded to: European Canadian/Mixed Canadian Asian Other
EOR_Q04	Education-level of Respondent	High school or less Trade certificate, University certificate below bachelor’s level, College, CEGEP, or non-university certificate or diploma Bachelor’s Degree Post-graduate Degree
INR_Q110	Respondent’s Estimated Annual Household Income	$20k – $49.9k $50k – $74.9k $75k – $99.9k $100k +
NOCS2011 C10	National Occupation classification of the respondent – last 12 months.	White-collar – (Health occupations; Management occupations; Natural and applied sciences occupations; Occupations unique to primary industry; Occupations in art, culture, recreation, and sport; Business, finance, and administrative occupations) Blue-collar – (Sales & Services Occupation; Trades, transport and equipment operators, and related occupations; Occupations unique to processing, manufacturing, and utilities)
WHW_Q130	How many hours per week do you usually work at your main job?	35–168
GEO_UA_RA_TYPE	Geographic Setting of where Respondent resides	Urban, Rural
PRN_Q20	Age of care recipient	18+
PRG_Q10	Relationship to care recipient	Immediate Family – (Spouse, Mother, Father, Son, Daughter, Brother, Sister) Extended Family – (Grandmother, Grandfather, Grandson, Granddaughter, Mother-in-law, Father-in-law, Son-in-law, Daughter-in-law, Brother-in-law, Sister-in-law, Aunt, Uncle, Niece, Nephew, Cousin) Other – (Close Friend, Ex-Spouse, Neighbour, Co-Worker, Other)
PRD_Q10	Drive Proximity to Care Recipient	0–29 min 30–59 min 60–179 min 180+ min
PRH_Q10	Frequency Seeing Care Recipient	Daily At Least Once a Week At Least Once a Month Less than Once a Month
HAP_Q10	In an average week, how many hours of care or help did you provide with these activities?	0–168
CRH_Q10	During the past 12 months, have your caregiving responsibilities caused you to feel tired?	Yes, No
CRH_Q30	During the past 12 months, have your caregiving responsibilities caused you to feel overwhelmed?	Yes, No
ICL_Q110 – 120	In the past 12 months, have your caregiving responsibilities caused you to spend less time with your spouse or partner, and children?	Yes, No
ICL_Q135 – 140	In the past 12 months, have your caregiving responsibilities caused you to spend less time with your friends and/or activities or hobbies?	Yes, No
ART_Q10 (Assisted-transport)	During the past 12 months, have you helped the primary care-receiver with transportation to do shopping or errands, or to get to medical appointments, or social events?	Yes, No
ART_Q20	Based on ART_Q10, how often have you helped the primary care-receiver to do shopping or errands, or to get to medical appointments, or social events?	Daily At least Once a Week At least Once a Month Less than Once a Month
ART_Q40	Received Unpaid Assistance for Assisted-Transport	Yes, No
ICL_Q100	In general, how have you been coping with your caregiving responsibilities?	Very Well Generally Well Not so well Not well at all
WGHT_PER	Estimate population representing per observation	Continuous
Bootstrap 1–500	Bootstrapped Weights	Continuous

Methodology

Defining CE & assisted-transport

Descriptive statistics consisted of cross-tabulations among the three groups to create the specialized carer-employee transport profile. We identify carer-employees based on the following three GSS questions:

1. Have you provided care for a person with a long-term health condition or a physical or mental disability during the past 12 months? (If yes)

2. How old is your primary care receiver? (If above 18 years old)

3. Are you employed full time? (If yes)

Carer-employees conducting assisted-transport were identified based on the fourth question:

1. During the past 12 months, have you helped the primary care receiver with transportation to do shopping or errands, or to get to medical appointments, or social events? (If yes)

Non-carer-employees are classified as those working full time, not active in caregiving, and at least 18 years of age. Of the 23,093 observations in the original dataset, 3,867 observations met the carer-employee criteria as defined in the first three questions. This data subset was used in the logistic regression analyses. With the assisted-transport question (question #4) added, only 1,274 of the 3,867 carer-employees had conducted assisted-transport caregiving. This smaller data subset was used to create a carer-employee transport profile. RStudio was used for descriptive statistics and logistic regression modeling.

Descriptive statistics

First, we created the profile of the carer-employees who performed assisted-transport via descriptive analyses. The weighted proportion of categorical variables and the averages of continuous variables are reported in Table 2–5. This builds the specialized profile of carer-employees conducting assisted-transport, including gender differences. The two-proportion t-test was used to test for statistical significance of differences between two proportions of each variable’s value (Tables 2 and 4).

Table 2. Socioeconomic and caregiving characteristics contrast among Non-CEs, CEs, and CEs with assisted-transport.

Socioeconomic variables	(1) Non-CE n = 100	(2) CE n = 658	(3) CE (Transport) n = 1274	T-Test^a
Age of the Respondent
18–29	13.3%	22.0%	19.1%	2–3
30–44	37.7%	26.0%	27.7%	−
45–64	46.0%	50.2%	51.6%	1–3
65+	3.1%	1.7%	1.6%	−
Gender of the Respondent
Male	58.5%	63.0%	54.4%	2–3
Female	41.5%	37.0%	45.6%	2–3
Ethnicity of the Respondent
Canadian/Mixed-Canadian	9.7%	14.9%	13.5%	1–2
European	66.3%	55.8%	54.6%	1–2; 1–3
Asian	−	2.4%	2.5%	−
Other	7.8%	7.8%	7.5%	−
Highest Education Obtained
High School or Less	22.1%	34.3%	32.1%	1–2; 2–3
Some College/Trade	36.3%	37.0%	34.5%	−
Bachelor’s Degree	27.4%	20.1%	24.2%	2–3
Post-Graduate	14.2%	8.3%	9.1%	1–2; 1–3
Household Income
$20k – $49.9k	LS	4.6%	2.9%	LS
$50k – $74.9k	6.8%	8.6%	7.4%	2–3
$75k – $99.9k	8.1%	10.5%	9.9%	−
$100k+	50.8%	32.1%	39.9%	1–2; 1–3; 2–3
Type of Work
Blue Collar	27.9%	37.7%	33.3%	1–2; 2–3
White Collar	71.7%	61.4%	66.0%	2–3
Hours of Work per Week
35–50 hrs.	76.5%	86.8%	85.8%	2–3
51–70 hrs.	20.1%	11.1%	11.1%	1–2
71+ hrs.	LS	2.1%	3.1%	LS
Geographic Setting
Urban	77.1%	73.1%	79.2%	2–3
Rural	22.9%	26.9%	20.8%	2–3
Caregiving Variables
Age of Care Recipient
18–30	NA	5.6%	4.9%	−
31–44	NA	4.4%	4.9%	−
45–64	NA	19.1%	20.5%	−
65–79	NA	36.9%	35.2%	−
80+	NA	34.0%	34.4%	−
Relationship to Care Recipient
Immediate Family	NA	51.8%	59.9%	2–3
Extended Family	NA	29.6%	28.6%	−
Other	NA	18.1%	10.7%	2–3
Drive Proximity to Care Recipient
0–29 min	NA	71.2%	78.0%	2–3
30–59 min	NA	11.7%	9.4%	−
60–179 min	NA	7.3%	7.5%	−
180+ min	NA	9.8%	5.1%	2–3
Hours of Caregiver per Week
1–15 hrs.	NA	77.3%	82.9%	2–3
16–30 hrs.	NA	4.7%	7.2%	2–3
31–44 hrs.	NA	0.3%	2.0%	−
45+ hrs.	NA	1.3%	2.0%	−
Feel Tired from Caregiving
Yes	NA	18.9%	46.6%	2–3
No	NA	31.9%	31.2%	−
Feel Overwhelmed from Caregiving
Yes	NA	11.0%	27.5%	2–3
No	NA	39.8%	50.5%	2–3
Time for Spouse
Yes	NA	8.2%	21.8%	2–3
No	NA	24.8%	30.3%	−
Time for Children
Yes	NA	9.0%	27.6%	2–3
No	NA	25.1%	29.3%	2–3
Time for Friends
Yes	NA	17.5%	43.2%	2–3
No	NA	34.3%	35.8%	−
Time for Hobbies
Yes	NA	18.0%	46.9%	2–3
No	NA	33.8%	32.0%	−
Frequency of Conducting Assisted-Transport
Daily	NA	NA	9.0%	NA
At Least Once a Week	NA	NA	45.0%	NA
At Least Once a Month	NA	NA	30.0%	NA
Less than Once a Month	NA	NA	16.0%	NA

^a p-value significance (≤0.05) at 95% confidence; LS: sample size too low to report; NA: not applied; − indicates not significant; Numbers (1 to 3) correspond to the population groups: 1 is Non-CE, 2 is CE, and 3 is CE Transport.

Table 3. Amount of time spent CEs conducting assisted-transport.

Frequency of conducting assisted-transport	Time spent on assisted-transport	%
Daily (n = 87)	< 1 hr.	22.0
	1–2.9 hrs	52.0
	3–9.9 hrs	26.0
At Least Once a Week (n = 530)	< 1 hr.	5.9
	1–2.9 hrs	53.7
	3–4.9 hrs	26.9
	5–9.9 hrs	10.7
	10 hrs +	2.7
At Least Once a Month (n = 445)	< 1 hr.	6.9
	1–2.9 hrs	43.4
	3–4.9 hrs	28.0
	5–9.9 hrs	14.6
	10 hrs +	6.7
Less than Once a Month (n = 207)	< 1 hr.	6.0
	1–2.9 hrs	58.0
	3–4.9 hrs	24.0
	5–9.9 hrs	8.0
	10 hrs +	4.0

Table 4. Gender differences of assisted-transport CEs.

Variable	Value	Male (%) n = 608	Female (%) n = 666	T – Test^a
Business Sector	Business, Finance, & Admin.	13.1	25.6	√
	Health Occupations	1.8	11.5	√
	Management Occupations	12.9	8.9	−
	Natural & Applied Sciences	13.9	5.1	√
	Art, Recreation, and Sports	1.4	3.0	√
	Social Sciences, Education, & Government	7.4	22.4	√
	Primary Industry	5.6	1.1	√
	Processing, Manufacturing, & Utilities	4.6	3.1	−
	Sales & Services	15.6	15.9	−
	Trades, Transport, and Equipment Operators	23	2.5	√
Work Sector Type	Blue Collar	43.3	21.4	√
	White Collar	56.1	77.7	√
Income	$20k – $49.9k	3.2	2.6	−
	$50k – $74.9k	8.3	6.3	−
	$75k – $99.9k	10.6	9.1	−
	$100k +	40.0	37.8	−
Education	High School or Less	36.3	27.0	√
	Some College/Trade	31.7	37.8	√
	Bachelors Degree	24.0	24.5	−
	Post-Graduate Degree	7.8	10.8	√
Ethnicity	European	54.8	54.3	−
	Canadian/Mixed Canadian	13.8	13.0	−
	Asian	1.5	3.6	−
	Other	7.8	7.2	−
Weekly Hrs. Of Employment	35–50 hrs	81.7	90.8	√
	51–70 hrs	14.4	7.1	√
	71+ hrs	3.8	2.1	√
Geographic Setting	Urban	76.2	82.8	√
	Rural	23.8	17.2	√
Weekly Hrs. Of Caregiving	1–15 hrs	84.4	81.1	√
	16–30 hrs	7.1	7.3	√
	31–44 hrs	2.2	1.7	−
	45+ hrs	1.1	3.1	√
Proximity to Care Recipient	0–29 min	79.1	76.8	√
	30–59 min	8.8	10.0	−
	60–179 min	7.7	7.2	−
	180+ min	4.4	6.0	√
Frequency Seeing Care Recipient	Daily	9.5	17.8	√
	At Least Once a Week	48.5	47.9	−
	At Least Once a Month	11	12.9	−
	Less than Once a Month	4.2	3.6	−
Frequency Conduct Assisted-Transport	Daily	8.7	7.9	−
	At Least Once a Week	44.5	45.5	−
	At Least Once a Month	30.6	29.6	−
	Less than Once a Month	16	16.6	−
Receive Unpaid Assistance for Assisted-Transport	Yes	81.2	69.8	√
	No	18.1	30.1	√
Cope with Caregiving	Generally Well	44.8	47.3	√
	Very Well	30.6	28.2	√
	Not Well At All	0.8	0.5	−
	Not Very Well	1.9	3.6	−
Feel Tired from Caregiving	Yes	41.7	52.3	√
	No	35.6	26.0	√
Overwhelmed from Caregiving	Yes	19.4	37.2	√
	No	58.3	41.1	√

^a√ = p-value significance (≤0.05) at 95% confidence; − is not significant.

Table 5. Gender differences in assisted-transport CEs per income bracket – females and males.

Variable	Value	$20k – $49.9k	$50k – $74.9k	$75k – $99.9k	$100k+
Females (n = 666)
Weekly Work (Hrs)	35–50	63.4%	95.2%	94.3%	88.4%
	51+	36.6%	−	5.7%	11.6%
Weekly Caregiving (Hrs)	1–15	57.9%	96.3%	78.3%	74.8%
	16+	38.0%	−	12.7%	13.5%
Received Unpaid Assistance for Assisted-transport	No	38.6%	30.8%	28.9%	32.8%
	Yes	61.4%	69.2%	70.5%	67.2%
Males (n = 608)
Weekly Work (Hrs)	35–50	94.9%	92.4%	79.4%	78.4%
	51+	−	7.6%	20.6%	21.6%
Weekly Caregiving (Hrs)	1–15	87.6%	92.3%	83.2%	84.5%
	16+	−	7.7%	16.0%	8.2%
Received Unpaid Assistance for Assisted-transport	No	25.2%	16.8%	34.7%	11.8%
	Yes	73.6%	82.7%	65.3%	87.6%

Logistic regression

Second, we conducted two logistic regressions. The first logit model (n = 1,163; Table 6A) examines the socioeconomic difference of carer-employees ‘conducting assisted-transport’ (dependent variable). The second model (n = 1,056; Table 6B) reveals the determinants contributing to the carer-employee ‘feeling overwhelmed’ (dependent variable), particularly if they are more overwhelmed from assisted-transport caregiving. Bootstrap weights were applied to improve accuracy and estimate the standard error of coefficients in the logit models.

Table 6. A. odds ratio of model 1: CEs conducting assisted-transport (Dependent – Reference: Yes).

Variable	Reference	EXP (B)^a	95% CI^a
Income	Continuous	1.00	1.00
Advanced Degree	1 = Bachelor or Above	1.00	0.89–1.11
Immediate Family	1 = Immediate Family member	1.56	1.41–1.72
Gender	1 = Female	1.76	1.58–1.94
Age of CE	1 = 18–44	0.84	0.75–0.93
Age of Recipient	1 = 45+	0.50	0.42–0.60
Frequency Seeing Care Recipient	1 = Frequent (Daily or Weekly)	1.33	1.19–1.48
Weekly Hrs. Of Caregiving	1 = 16+ hrs	3.99	3.24–4.90
White Collar	1 = White	1.48	1.32–1.67
Urban	1 = Urban	1.31	1.13–1.52
Public Transit	1 = Available	1.51	1.31–1.73
B. Odds Ratio of Model 2: CEs Feel Overwhelmed (Dependent – Reference: Yes)
Variable	Reference	Exp (B)	95% C.I.
Income	Continuous	1.00	1.00
Advanced Degree	1 = Bachelor or Above	1.02	0.94–1.10
Gender	1 = Female	3.47	3.22–3.75
Age of CE	1 = 45+	1.42	1.32–1.52
Age of Recipient	1 = 18–44	1.91	1.68–2.17
Proximity to Care Recipient	1 = Close (< 30 min of drive)	1.16	1.07–1.27
Assisted-Transport Caregiving	1 = Yes	1.18	1.04–1.35
Weekly Hrs. Of Caregiving	1 = 16+ hrs	3.06	2.75–3.42
Weekly Hrs. Of Work	1 = 35–50 hrs	1.34	1.22–1.47
Public Transit	1 = Available	1.15	1.04–1.26
Urban	1 = Urban	0.84	0.76–0.94

^a EXP(B) – exponentiation of the B coefficient, hence the odds ratio.

^a 95% Confidence interval of odds ratio.

Approximately 20% of carer-employees’ income information was not provided. To retain enough sample size for the multivariate regression analyses, missing income was imputed by the predictive mean matching algorithm (PMM) found in the MICE package in RStudio. The following predictors were used: respondent’s age, sex, number of hours worked per week, and whether they worked in the white-collar sector. PMM ensures that imputed values are credible and possibly more suitable than standard regression methods (Horton & Lipsitz, 2001).

Results

Socioeconomic & health differences amongst population groups

The following variables relevant to the two carer-employee groups were examined: 1) age of the respondent; 2) age and relationship to the care-recipient; 3) weekly hours of caregiving, and; 4) working in the white-collar sector. Surprisingly, there were proportionally more male carer-employees than female carer-employees. Each population group has its own socioeconomic profile (Table 2).

Based on proportional t-tests, the socioeconomic profile of non-carer-employees is shown to be significantly: younger, of European background, of higher education and household income, and putting in more labor hours per week when compared to carer-employees. Comparing the socioeconomic profile of carer-employees who conducted assisted-transport to those who did not, carer-employees conducting assisted-transport are more educated (bachelor’s degree), have a somewhat higher household income, are more likely to work in the white-collar sector, and are more likely to live in urban areas. Carer-employees conducting assisted-transport are more likely to care for an immediate family member; live in closer proximity to their care-recipient; have higher caregiving hours per week; feel tired and overwhelmed from caregiving, and; have less time for their children compared to carer-employees who did not conduct assisted-transport. With respect to the frequency of conducting assisted-transport caregiving, three-quarters of them conduct this task at “least once a week” to “at least once a month.” Each value for each frequency of conducting assisted-transport is broken down into the number of hours spent (Table 3). In summary, most carer-employees spent between 1 and 4.9 hours on assisted-transport caregiving, ranging from “daily” to “less than once a month”.

Gender differences in CEs conducting assisted-transport

The following variables were used to examine gender differences: 1) employment sector and type; 2) income; 3) education level; 4) work hours per week, and; 5) variables related to caregiving (Table 4).

From the t-test scores, most females are employed in business, finance, and administrative positions, health occupations, and social sciences, education, and government positions. Proportionally more females than males work in the white-collar workforce, have some college/trade training, are more likely to have a post-graduate degree, work 35–50 hr per week, and are urban dwellers. Proportionally more females than males do “extreme” cases, reflected in weekly hours of caregiving exceeding 45, driving for more than 180 min to see the care-recipient (180+ min), and visiting their care-recipient daily. Additionally, proportionally fewer females than males receive unpaid help for assisted-transport and feel more tired and overwhelmed from caregiving tasks. There were no household income differences between the genders; however, Table 5 displays each household income value per assisted-transport carer-employee gender.

Proportionally, fewer females tend to work longer (51+ hr) and provide care fewer weekly hours (16+ hr) as they climb the household income ladder; whereas, more males work longer hours (51+ hr).

Characteristics associated with providing assisted-transport

Using the sample of carer-employees dataset (n = 3,867), the first model examined the characteristics of carer-employees, including geographic variables that are associated with the probability of providing assisted-transport caregiving. Observations were dropped due to missing values. Table 6A reports the results of the logistic regression. It indicates that carer-employees are more likely to conduct assisted-transport caregiving if they are taking care of an immediate family member [OR = 1.56], see the care-recipient on a daily/weekly basis [OR = 1.33], are female [OR = 1.76], provide at least 16 hr of caregiving per week [OR = 3.99], work in the white-collar sector [OR = 1.48], and live in an urban area [OR = 1.31].

[t] Table 6A near here [/t]

The second model tested whether carer-employees with assisted-transport caregiving are more likely to be overwhelmed while controlling for the characteristics of carer-employees. Model 2 (Table 6B) lost 2,811 observations due to many, “Not Asked” values in assisted-transport. The odds of being overwhelmed for the carer-employees conducting assisted-transport caregiving are 18% more than for carer-employees not conducting this task. The highlight of Model 2 is that carer-employees are more likely to be overwhelmed from assisted-transport caregiving [OR = 1.18], which validates whether carer-employees are more likely to be overwhelmed from assisted-transport caregiving. Other highlights are carer-employees more likely to be overwhelmed from caregiving if they are female [OR = 3.47] and provided at least 16 hr of caregiving per week [OR = 3.06] (Table 6B).

Discussion

Little is known about assisted-transport caregiving, particularly the socioeconomic profiles of carer-employees, behavioral factors that increase the likelihood of conducting the task, and whether carer-employees are more likely to feel overwhelmed from it.

Assisted-transport CE profile

The assisted-transport carer-employee profile is similar to those found in previous studies (Duxbury et al., 2012; Employer Panel for Caregivers, 2015; RAPP, 2014), especially the predominant characteristics of carer-employees being baby-boomers, providing caregiving for 1 to 15 hr per week, and being related to their care-recipient as an immediate family member. What differentiates the assisted-transport carer-employee profile are: gender, education, household income, work sector, location, feeling tired and overwhelmed from caregiving, driving time to their dependent, frequency of conducting the task, and receiving additional informal assistance for the task. Compared to the general carer-employee, the carer-employee assisted-transport population are: 1) more gender equal; 2) have higher education and income; 3) more representative of the white-collar sector; 4) more representative of urban dwellers; 5) provide more weekly caregiving hours, and; 6) feel more tired and overwhelmed from caregiving. Of all the findings in this paper, the assisted-transport carer-employee profile is noteworthy, despite its simplicity. The profile can be set as an establishment for future interdisciplinary research, particularly in the fields of demography, public health, and transport geography, specifically as a channel to cultivate appropriate assisted-transport policies for carer-employees. Additionally, it can be set as the first domain (key characteristics of the caregiver) of Pearlin’s framework for modeling purposes.

Gender differences in CEs conducting assisted-transport

Both descriptive and model results highlighted several gaps between male and female carer-employees conducting assisted-transport, which addressed the second objective. Proportionally more assisted-transport female carer-employees are overwhelmed and feel tired from general caregiving tasks. One possible explanation for this gender gap may be the multiple roles female carer-employees often occupy, including caregiver, employee, mother, domestic manager, cook, and spouse. This uneven division of labor is reflected in who conducts the assisted-transport: female carer-employees are 76% more likely to conduct assisted-transport than male carer-employees (as shown in Table 6A).

Gender roles continue to gradually change, as males become more involved with household tasks while females become more engaged in the workforce; however, gender inequality continues to remain problematic. Findings from Table 4 and from past studies (Beede et al., 2011; Canadian Women’s Foundation, 2017), indicate that women continue to be concentrated in the service sectors; teaching, nursing, health care, office, and administrative work, while men are more represented in numerous STEM fields. As mentioned in the literature, many of these female-dominated occupations are often lower in the employment hierarchy, resulting in less freedom over employment times and less travel flexibility. This explanation may be represented as a secondary stressor (i.e., work-caregiving conflict) from Pearlin’s framework. Lastly, it is possible that assisted-transport female carer-employees feel tired and overwhelmed due to fewer receiving unpaid assistance for their transport tasks, together with more likely seeing their care-recipient daily (as shown in Table 4).

Behavioral characteristics

The logit models constitute the outcomes of stress in Pearlin’s framework. The first logit model predicted the likelihood of carer-employees conducting assisted-transport based on their socio-characteristics, such as proximity and age of their recipient, and thus, addressed the third research objective. Based on descriptive statistics findings, some of the results in the model were expected. For instance, it was expected that carer-employees were more likely [OR = 1.56] to provide assisted-transport to an immediate family member (Table 6A). Strong indicators of the likelihood of conducting assisted-transport are frequency of seeing the care-recipient and weekly hours of caregiving per week. Interestingly, carer-employees were 50% less likely to provide assisted-transport to care-recipients aged 45+. This may indicate that an increase in age does not reflect on the demand of assisted-transport. One question to explore in the future is to see if carer-employees are more likely to conduct assisted-transport based on the poor health status of the care-recipient. Carer-employees working in the white-collar sector are 48% more likely to provide assisted-transport than those working in blue-collar (Table 6A). This may be attributed to the traits of work flexibility, such as co-workers covering the carer-employee’s work tasks while temporarily unavailable; telecommuting, and/or; having compressed work weeks (Duxbury et al., 2012).

Assisted-transport CEs feeling overwhelmed

The second logit model addressed whether carer-employees performing assisted-transport were more overwhelmed than those who do not perform this task, thus fulfilling the fourth objective. Expected results are similar to the first logit model, particularly carer-employees’ gender and age, proximity to the care-recipient, and weekly hours of caregiving. Carer-employees aged 45 years, or more were 42% more likely to be overwhelmed from caregiving (Table 6B). This age group, sometimes referred to as the sandwich generation, make up most of the carer-employee population. The sandwich generation are mainly baby-boomers and Gen Xers, which take care of both their children and parents at the same time (Rubin and White-Means, 2009). This group has the most work experience and greatest family responsibility; tend to struggle with work–life balance; are more likely to be overwhelmed (Duxbury et al., 2012), and; are more likely to conduct assisted-transport daily (Nichols et al., 1997; Rubin and White-Means, 2009). Carer-employees are more likely to be overwhelmed from assisted-transport. This is likely due to the frequency of conducting assisted-transport and the amount of time spent doing it. Results show most carer-employees carrying out doing assisted-transport on a weekly basis, averaging 1 to 4.9 hr per week (Table 2). Though this may not seem much, this accumulates together with other caregiving tasks – some of which are indirectly related, such as running errands to the bank and visiting the recipient for emotional support. Additionally, about 10% of carer-employees conduct assisted-transport daily.

One additional explanatory factor may be related to what comprises carer-employees activity-travel behavior, such as stopping in-between their commute, and/or taking detours for their dependent’s needs; these activities would increase their commute time and potentially cause them to become more stressed. This transit concept is known as trip-chaining (McGuckin et al., 1995). Several studies have noted that sociodemographic and socioeconomic characteristics are one of the primary drivers for people to conduct trip-chains, which make up their activity-travel behavior (Islam et al., 2012; Wang, 2015). Transport policies, such as the development of mass transit systems, also impact activity-travel behavior. In Canada, the current mass transit systems are not well designed for the needs of the older adult population (Mercado, Paez, Scott, Newbold, & Kanaroglou, 2007); thus, many resort to informal caregivers as their main provider of assisted-transport. Transit policies will soon need to be revolutionized to mitigate assisted-transport demand. Fortunately, Transport Canada has proposed an investment of approximately $77 million CND for the next 5 years to modernize transportation systems, including the development of automated vehicles (Government of Canada, 2017). Automated vehicles may be a solution to the need for assisted-transport; however, it will take a few decades to implement after factoring in regulations, ethics, beta-testing, and activity-travel behavior research. Mass transit and automated vehicles would only be feasible in urban locations. More data would need to be collected from rural locations in order to suggest more tailored mobility solutions for rural areas. Another significant discipline in transport policy is accessibility to essential services (i.e., clinics, grocery stores). The availability of transit services is one of the main factors that affects accessibility, which in turn can impact travel demand. Several Canadian cities recognize the importance of investing in new transit infrastructure to improve their economy, environment, and residents’ health (SNC-Lavalin, 2015). The Canada Line in Vancouver opened in 2009 and carries more than 122,000 people per day (SNC-Lavalin, 2015). In Calgary, the West Light Rail Transit extension was implemented in 2012, which has increased accessibility to 44,000 residents (SNC-Lavalin, 2015). In addition, an improvement in accessible transit can save $390 million CND in annual public costs (Canadian Urban Transit Association, 2013). These estimates signify that an improvement in transit systems augments accessibility for the residential population. Thereby, potentially reducing the older adults’ dependency upon the automobile including the demands of assisted-transport from carer-employees. More research will be required for the fields of accessibility and activity-travel behavior as it has not been addressed thus far in the caregiving context. Nonetheless, this research paper sets as a steppingstone towards the implementation of transport policies related to assisted-transport and activity-travel behavior.

The significance of both models validates the results from the descriptive statistics and enables decision-makers to have a better understanding of assisted-transport caregiving. Therefore, decision-makers may use predictive models as a prevention tool to monitor their carer-employees, particularly at the workplace, whilst implementing a caregiver-friendly workplace program tailored towards carer-employee conducting assisted-transport. One example of an assisted-transport caregiver-friendly workplace program would be the opportunity to conduct assisted transport during low-activity business hours, allowing the carer-employee to make up these hours outside of business hours.

Limitations

The main strength of this study is that it addresses the assisted-transport gap in the carer-employee literature via descriptive statistics and logistic regression models. Both methods revealed useful information for decision-makers. However, this study has encountered limitations as well. One limitation is the nature of the data, especially classifying amongst the three population groups of concern. The dataset is tailored to caregiver respondents, which is why there were low observations of non-carer-employee (n = 100). Additionally, the survey had a variable similar to the assisted-transport one, which asked if the respondent generally provided transportation assistance, whereas the other is only if conducted to the primary care-recipient, which was used for the carer-employee assisted-transport profile. For descriptive analyses, this similar assisted-transport variable was incorporated to classify carer-employees strictly not conducting assisted-transport; hence, relatively low observations (n = 658). As for modeling, household income was used rather than individual income due to far more missing observations (~50%) in the individual category rendering it not suitable for imputation. The second model indicates that carer-employees conducting assisted-transport are overwhelmed. There is no concrete explanation as to why or what causes this feeling (despite the first model). The first model includes carer-employees that conduct assisted-transport, regardless if they are overwhelmed or not. This model was not further refined due to small sample size. After imputing and subsetting to the definition of the carer-employee and those conducting assisted-transport, the dataset accounted for only 5.5% of the total observations to perform logistic regression. This was the result of a high number of values denoting “Don’t Know” or “Not Asked” for assisted-transport, which were treated as NA values. “Not Asked” was the most common value and may have to do with the question of not being applicable to the respondent. These limitations can be addressed, for instance, by developing a new survey instrument.

Conclusion

This paper is the first in the CE literature to develop a socioeconomic profile of CEs performing assisted-transport tasks, differentiating it from the other two groups of concern. It examines whether gender differences exist based on the profile, identifies behavioral factors that increase the likelihood of conducting assisted-transport, and whether CEs are more likely to be overwhelmed as a result. Gender gaps have been identified based on employment features and caregiving characteristics as plausible explanations. Weekly hours of caregiving, gender, age of the recipient and white-collar employment are the primary variables that strongly associate the likelihood of CEs conducting assisted-transport. Lastly, CEs conducting assisted-transport are more likely to feel overwhelmed than those not conducting this task. This paper further addresses the assisted-transport gap; however, there is a need for more research to address the limitations of this study. Future research will require creating a dataset with enough sample size to perform a new model. The dataset would consist of all variables used for the three population groups of concern, incorporating a work-commute scale. The work-commute scale would measure the activity-travel behavior of carer-employees conducting assisted-transport while traveling to work or working on-site. Examples of variables in this proposed scale include: “Do you make stops related to assisted-transport while commuting to work?”, “Does your current work schedule make it harder for you to perform assisted-transport?”, and “Does the assisted-transport task require you to leave during work hours because of limited opening hours (i.e., physician office)?”

The new model would identify characteristics that cause assisted-transport carer-employees to be overwhelmed. These suggested improvements would provide the foundation for decision-makers to begin implementing caregiver-friendly workplace programs related to assisted-transport, as no such transit policy exists to improve the carer-employee’s work–life balance. An example of a caregiver-friendly workplace policy related to assisted-transport is to allow the carer-employee to take one day off every two weeks (without using their personal days) to conduct their caregiving-related tasks. In exchange, the carer-employee would have to devote an extra hour per day for the week. A more comprehensive caregiver-friendly workplace policy related to assisted-transport would require a deeper profile of the carer-employee’s activity travel behavior. This could be accomplished by collecting GPS and trip diary data from carer-employees, which can then be analyzed using geographic information systems (GIS) (Kwan et al., 2015). GIS is an effective tool to measure the demands of assisted-transport caregiving by analyzing accessibility through changes in service areas of mass transit systems. Additional analyses would be identifying locations that’ll require an immediate need to implement services to improve accessibility, which may reduce assisted-transport burden amongst carer-employees. Consequently, this area of research has the potential to facilitate the decision-making processes for both, the carer-employee and care-recipient populations.

Acknowledgments

This research is supported by the Canadian Institutes of Health Research (CIHR) program, and specifically Dr. Allison Williams (primary author’s supervisor) Chair in Gender, Work & Health, which examines Gender, Health and Caregiver Friendly Workplaces (CGI 126585).

Additional information

Funding

This work was supported by the CIHR Chair in Gender, Work, and Health [CG1 126585].